## tarea2026037

## Source details

**Canonical URL:** [tarea2026037](https://www.imf.org/-/media/files/publications/tar/2026/english/tarea2026037.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/tar/2026/english/tarea2026037.pdf.md)
- [Structured JSON version](/-/media/files/publications/tar/2026/english/tarea2026037.pdf.json)

---

### Mission, scope, and meetings
- IMF MCM mission visited Montevideo from November 24 to December 5, 2025, at the request of the Central Bank of Uruguay (BCU) to provide technical assistance on monetary policy implementation and liquidity management.
- Mission reviewed operational framework: open market operations, reserve requirements, standing facilities, central bank bills (LRM), liquidity forecasting, and institutional arrangements.
- Meetings included BCU management (Mr. Guillermo Tolosa; Ms. Isidra Delfino; Mr. Gerardo Licandro; Mr. Leonardo Vicente; Mr. Fabio Malacrida), other senior officials, technical staff, banks, non-bank financial institutions, journalists, unions, consultancies, academia, and media.

### Executive summary — core findings
- Interest-rate framework:
  - Uruguay’s interest-rate-based framework is credible and effective; the BCU has kept the overnight interbank rate (T1D/TMM) within a few basis points of the Monetary Policy Rate (TPM).
- Short-end transmission:
  - Policy-rate changes pass through strongly to short-tenor LRMs and key peso deposit rates.
  - Short-end term premia are economically small; premia rise at longer maturities and are explained by country risk and medium-term inflation expectations.
- Structural and market constraints:
  - Approximately three-quarters of bank deposits are foreign‑currency denominated.
  - Unsecured peso interbank activity is thin beyond overnight; structural excess liquidity is substantial.
  - Routine liquidity redistribution mainly channels through the central bank’s balance sheet.
- Operational stance:
  - BCU relies on frequent fine-tuning operations creating a very narrow “mini-corridor” around TPM within a much wider formal standing facility range, keeping T1D close to TPM but constraining interbank price discovery.

### Recommended operating-framework changes (high-level)
- Core changes:
  - Introduce a weekly fixed-rate, full-allotment 7-day main operation at the TPM as the core liquidity-management tool.
  - Recalibrate standing facilities to a corridor of ±50 basis points around the policy rate.
  - Concentrate LRM issuance in three standardized maturities: 3 months, 6 months, and 12 months; discontinue the 30-day bill.
  - During transition, continue fine‑tuning operations with rates around TPM until new setup fully anchors short-term rates.
- Forecasting and institutional improvements:
  - Strengthen liquidity forecasting and data infrastructure; adopt IMF MCMCO liquidity-forecasting toolkit.
  - Strengthen information sharing with the Treasury, including currency split of fiscal flows and structured auction-calendar coordination.
  - Consolidate liquidity forecasting and operations execution in a dedicated Monetary Operations and Market Development area with specialist units.

### Detailed recommendations (selected)
- Monetary Policy Framework:
  - Weekly fixed-rate, full-allotment 7-day main operation at policy rate. Priority: High. Timeframe: Short-term. Paragraph: 82.
  - Calibrate corridor to at least ±50 bps around policy rate. Priority: High. Timeframe: Short-term. Paragraph: 109.
  - Make standing facilities accessible to all eligible counterparties meeting collateral and risk-management requirements; remove time-based caps. Priority: High. Timeframe: Short-term. Paragraph: 119.
  - Remunerate peso reserve requirements at the policy rate. Priority: High. Timeframe: Immediate. Paragraph: 96.
  - Unify peso reserve requirement ratios across tenors. Priority: Medium. Timeframe: Short-term. Paragraph: 103.
  - Move MRR from contemporaneous averaging to a backward-looking calibration. Priority: Medium. Timeframe: Medium-term. Paragraph: 107.
  - Lower effective reserve requirement ratios in pesos to reduce liquidity parked in non-marketable assets. Priority: Medium. Timeframe: Short-term. Paragraph: 106.
  - De-stigmatize the standing lending facility by clearly separating it from ELA. Priority: Medium. Timeframe: Medium-term. Paragraph: 122.
  - Strengthen organizational capacity of monetary policy operations area. Priority: Medium. Timeframe: Medium-term. Paragraph: 125.
- Liquidity Forecasting:
  - Reframe forecasts around a standard liquidity table focused on reserves; treat reserve requirement as a constraint. Priority: High. Timeframe: Immediate. Paragraph: 43.
  - Distinguish forecasting horizons and preserve immutable forecast vintages. Priority: High. Timeframe: Immediate. Paragraph: 44.
  - Produce and use forecast uncertainty (prediction intervals) as operational input. Priority: Medium. Timeframe: Short-term. Paragraph: 49.
  - Evaluate forecast errors regularly and transparently. Priority: High. Timeframe: Immediate. Paragraph: 50.
  - Strengthen information sharing with the Treasury via a formal MoU with fiscal reports disaggregated by currency. Priority: Medium. Timeframe: Medium-term. Paragraph: 53.
- Money Market Development and Managing Liquidity Surplus:
  - Concentrate LRM issuance in 3-, 6-, and 12-month tenors. Priority: High. Timeframe: Short-term. Paragraph: 137.
  - Standardize LRM auction design and pre-announce fixed issuance amounts. Priority: Medium. Timeframe: Short-term. Paragraph: 143.
  - Reduce auction frequency and adopt maturity concentration to enhance liquidity. Priority: Medium. Timeframe: Short-term. Paragraph: 147.
  - Maintain split-curve in short term (BCU LRMs up to 12 months; Treasury beyond); assess medium-term transition to Treasury-only model. Priority: Medium. Timeframe: Medium-term. Paragraph: 155.

### Liquidity forecasting and data/institutional improvements
- Current shortcomings:
  - Forecasts produce usable point estimates but rely on complex Excel files, tacit knowledge, and incomplete government cash-flow information.
  - Forecast vintages are often overwritten; uncertainty not systematically quantified.
- Recommended steps:
  - Adopt a standard liquidity table where autonomous factors (CiC, GovD) feed projections of aggregate reserves relative to requirement plus buffer.
  - Implement IMF MCMCO toolkit for automated model estimation, forecast evaluation, prediction intervals, and immutable vintages.
  - Strengthen information sharing with Treasury via formal MoU, fiscal reports disaggregated by currency, and auction-calendar coordination.

### LRM market role and issuance strategy
- Dual purpose:
  - LRMs absorb structural liquidity surpluses and anchor the short end of the peso yield curve.
- Empirical findings:
  - Short-dated LRMs move closely with the policy rate; 30-, 90-, and 180-day pass‑through broadly in the ~0.7–0.9 range.
  - 360-day node pass‑through smaller and trending downward; 12-month increasingly reflects expectations and risk premia.
  - Auction results: BCU acts as price maker at short tenors and more as price taker at 6‑ and 12‑month maturities.
- Operational recommendations:
  - Concentrate issuance in 3-, 6-, and 12-month standardized maturities with clustered maturities; discontinue 30-day bill.
  - Standardize auction rules, cut-off conventions, and pre‑announce fixed issuance amounts.
  - Maintain split-curve in near term; coordinate cash- and debt-management closely; consider medium-term move to single-instrument model.

### Market functioning, depth, and BCU footprint
- Market depth:
  - Peso money market functional but shallow beyond overnight; unsecured interbank trades beyond overnight negligible.
  - By November 2025: overnight “Call pesos garantizado” turnover UYU 626 billion; ~41 percent originated from BCU fine-tuning operations; unsecured call operations ≈ 3 percent of turnover.
- Market concentration and turnover (July 2024–June 2025):
  - Total market operations ≈ 3,340.312 (UYU mi).
  - Primary issuance ≈ 2.5 trillion pesos (≈ three-quarters); secondary trading ≈ 0.8 trillion pesos (≈ one-quarter).
  - Market shares (percent): LRMs primary 45.1; LRMs secondary 88.5; Treasury notes primary 2.5; Treasury secondary 3.9; Private CDs primary 50.6; Private CDs secondary 0.0; International issuances secondary 7.0; Other securities secondary 0.6.
  - Private-sector secondary market effectively negligible (1.7 billion out of 1.3 trillion; ~0.1 percent).
- BCU counterparty role:
  - Outside overnight, BCU is counterparty in around 60–80 percent of peso PG transactions across most maturities.
- Mini-corridor statistics:
  - Observed TMM–TPM spread: 7.3 bp (median 5 bp; P90 20 bp), narrowing from 16.3 bp in 2020 to 2.9–3.3 bp in 2024–2025.
  - Formal corridor spans approximately 1,400 bp (1–15 percent).

### Pass-through, term premia, and retail rates
- Short-run pass-through model:
  - ARDL estimated tenor-by-tenor in rolling 36-month windows; cumulative short-term pass-through presented with 90 percent HAC confidence bands.
- Key empirical findings:
  - LRM pass‑through: 30‑, 90‑, 180‑day cut-offs estimated broadly in the ~0.7–0.9 range.
  - Deposit pass‑through: firms >0.6; households very short (~15 days) often indistinguishable from zero; longer household buckets around 0.6.
  - Lending pass‑through: firm loan rates moderate, slower and asymmetric; household overdraft rates volatile and frequently indistinguishable from zero.
- Term-premium decomposition:
  - Term premia near zero at one month; moderate and time‑varying at 6–12 months; overall term premium lower than 10 percent.
  - Dominant drivers at 6–12 months: Uruguay USD sovereign yield (country‑risk proxy) and medium‑term inflation expectations.
  - FX changes not independently systematic once USD sovereign yield and 24‑month inflation survey included.
- Implications:
  - Anchoring 24‑month inflation expectations within the 3–6 percent target range lowers inflation‑risk component of 6–12‑month yields and improves pass‑through.
  - Pass‑through to retail lending rates depends on competition, bank risk appetite, capital positions, and excess liquidity beyond liquidity instruments alone.

### Reserve requirements, remuneration, and calibration
- Structure and magnitudes:
  - Peso/UI reserve schedule: Demand & <30-day time 15%; 30–90 days 3%; 91–180 days 2%; 181–366 days 1%; >1 year 0%.
  - FX resident deposits: ≤180 days 28%; >180 days 20%; FX non-residents net 28%.
  - Effective MRR about 12.5–13 percent since 2022.
  - Reserve balances at BCU ≈ 9–10 percent of GDP; excess reserves amounted to UYU 431 billion in November 2025 (≈ 12 percent of GDP).
- Remuneration and recent change:
  - Required peso reserves were remunerated at 60 percent of TPM; new Circular No. 2491 (Nov 27, 2025) updates remuneration:
    - Effective March 1, 2026: required reserves in local currency earn 100 percent of policy rate (up from 60 percent).
    - Excess balances in local currency remunerated at policy rate − 2 percentage points (zero floor).
  - Foreign-currency reserve remuneration:
    - Required reserves on euro‑ and dollar‑denominated deposits <30 days earn policy rate − 1 percentage point.
    - Other required FX reserves earn policy rate − 0.5 percentage point (never below zero).
    - Excess FX balances remunerated at policy rate − 1 percentage point, with zero floor.
- Recommendations:
  - Remunerate peso reserve requirements at the policy rate to eliminate reserve-tax wedge and tighten link between policy rate and transfer prices.
  - Unify peso reserve ratios across tenors.
  - Lower effective MRR toward international norms (staff indicative effective ratio benchmark: 7 percent).
  - Move MRR from contemporaneous averaging to backward-looking calibration to fix targets at start of maintenance period.

### Recommended calibration and steady-state split (staff indicative)
- Effective peso reserve ratio benchmark: 7 percent (covers roughly 95 percent of observed daily deposit outflows).
- Applying 7 percent effective ratio to reserve-requirement base as of September 2025 → Required reserves: UYU 29,398 million.
- Recommended stock of weekly 7-day OMO absorption: UYU 73,155 million.
- Residual structural surplus assigned to LRMs.

### Weekly 7‑day FRFA main operation — design, evidence, and simulation
- Main recommendation:
  - Weekly fixed-rate, full-allotment 7‑day operation at TPM as core liquidity-management tool; LRMs remain structural tool; standing facilities form corridor; ISF/PG-Call repositioned as exceptional backstops.
- Rationale and expected effects:
  - Replace near-daily micro-management with predictable routine; improve banks’ liquidity planning, interbank trading, reserve averaging use, and transmission of TPM to marginal funding costs.
- Evidence on current intensity:
  - From 2020 to mid-2025: 2,380 ISF interventions across 1,019 days; average 38 operations per month across 16 ISF days.
  - Mini-corridor width averaged 7.3 bp; median 5 bp; P90 20 bp; narrowed from 16.3 bp in 2020 to 3.3 bp in 2025.
- Annex III simulation scenarios:
  - Scenario 1 — mop-up to zero on intervention day.
  - Scenario 2 — mop-up of minimum projected liquidity in the week.
  - Scenario 3 — mop-up of average projected liquidity in the week (preferred; consistent with reserve averaging).
  - Key metric: Ratio_m(s) = 100 × Scenario volume_m(s) / Baseline volume_m(s); values < 100 percent indicate lower intervention volume than current regime.
  - Preferred calibration: Scenario 3 minimizes persistent reliance on standing facilities and aligns with reserve-averaging.

### Corridor strategy and standing facilities
- International evidence:
  - Corridor widths across central banks range from approximately 50 basis points to over 200 basis points.
  - Many inflation targeters operate in 50–150 bps range.
- BCU recommendations:
  - Reset lending and deposit facility rates to a narrow, symmetric band around TPM approximating current mini-corridor but at slightly wider spreads; pragmatic starting point: corridor of at least 50 basis points around the policy rate.
  - Replace frequent fine‑tuning gradually with corridor-guided operations; widen corridor over time contingent on stable anchoring and improved forecast performance.
  - Ensure standing facilities accessible to all eligible counterparties meeting collateral and risk-management requirements; remove time-based caps.
  - De-stigmatize standing lending facility by clearly separating it from ELA; explicitly communicate facility availability and eligibility.

### Institutional and organizational recommendations
- Current fragmentation:
  - Liquidity forecasting within Monetary Policy Management; execution of LRMs and fine-tuning in Asset and Liability Management; small Monetary Operations Division handles day-to-day implementation.
- Recommended reorganization:
  - Create a dedicated Monetary Operations and Market Development area (initially under Asset and Liability Management) with specialist units:
    - Monetary Operations Execution Unit
    - Short-Term Liquidity Forecasting Unit
    - Market Analysis and Intelligence Unit
    - Market Development Unit
    - Back-Office Operations Unit
  - Strengthen staffing: Monetary Operations Division currently one manager and three staff; scale-up recommended.
  - Mandate: keep overnight rate close to TPM and develop local markets to support transmission and de-dollarization.

### Capacity building, governance, and future TA
- Institutional capacity needs:
  - Staff training in market operations; robust internal governance; transparency of operational procedures to Board.
  - Strong coordination between monetary policy and operations via systematic sharing of liquidity projections.
- Future TA priorities:
  - Leverage advanced AI tools and advanced quantitative modeling.
  - Support for statistical liquidity forecasting, reserve-demand estimation, and market-development initiatives.
  - Support for developing an ELA framework and building institutional capacity to align operational structure with interest-rate targeting objectives.

### Balance-sheet and liquidity-management facts (Annex IV)
- BCU balance-sheet snapshot (as of end-October 2025, mn UYU and percent of GDP):
  - Total Assets: 576100 17.8 %
  - Total Liabilities: 576100 17.8 %
  - Net Foreign Assets: 783883 24.2 %
  - Net Domestic Credit: -207784 -6.4 %
  - Net Claims on Central Government: 126554 3.9 %
  - Net Claims on financial sector: -332253 -10.3 %
  - Net Claims on non-financial sector: 1154 0 %
  - OIN: -3239 -0.1 %
  - Equity: 24879 0.8 %
  - Base Money: 154951 4.8 %
    - Currency in circulation: 109418 3.4 %
    - Banks' reserves: 43544 1.3 %
    - Other deposits: 1989 0.1 %
  - CB Bills: 396270 12.2 %
- LRMs stock:
  - Rose from UYU 333 billion at end-2023 (~10 percent of GDP) to UYU 425 billion in November 2025 (~12 percent of GDP).
  - Composition of LRMs outstanding: 30d 7%; 90d 27%; 180d 23%; 360d 43%.

### Operational diagnostics and publication recommendations
- Forecast publication and transparency:
  - Recommendation: publish liquidity forecasts (with uncertainty measures) after seven‑day FRFA adopted and forecasting framework validated.
  - Current practice: internal daily planning; forecasts not published; forecast errors not published; vintages overwritten.
- Data and toolkit:
  - Develop structured database integrating liquidity data, balance-sheet information, and forecast vintages; operationalize R-based toolkit with immutable vintages, model selection, reconciliation, fan charts, and exception logs.
- Forecast evaluation:
  - Use expanding rolling-origin out-of-sample evaluation; report ME, MAE, RMSE, interval scores, and empirical coverage rates; re-evaluate models at least every "six months" or when performance deteriorates.

_Italic source: IMF Technical Assistance Report — Preface, Executive Summary, Recommendation Summary, and selected Annexes (tarea2026037)._

### Preface ................................................................................................................

### tarea2026037 - Preface

### Mission, scope, and meetings
- An MCM from the IMF visited Montevideo from November 24 to December 5, 2025, at the request of the Central Bank of Uruguay (BCU) to provide technical assistance on monetary policy implementation and liquidity management.
- The mission reviewed the BCU’s operational framework, including open market operations, reserve requirements, standing facilities, central bank bills, liquidity forecasting, and institutional arrangements for monetary operations.
- The mission met with BCU management, including:
  - Mr. Guillermo Tolosa, President of the BCU
  - Ms. Isidra Delfino, Manager of Institutional Communications
  - Mr. Gerardo Licandro, Manager of Economic Advisory
  - Mr. Leonardo Vicente, Manager of Monetary Policy
  - Mr. Fabio Malacrida, Manager of the Asset and Liability Management Area
  - Other senior officials and technical staff at BCU
- The mission also met with representatives from banks, non-bank financial institutions, journalists, unions, consultancy companies, academia, and the media.

### Executive summary — core findings
- Uruguay’s interest-rate-based framework is credible and effective in ensuring monetary policy transmission; the BCU has maintained the overnight interbank rate (T1D) within a few basis points of the Monetary Policy Rate (Tasa de Política Monetaria, TPM).
- Short-end transmission:
  - Policy rate changes pass through strongly to short-tenor Monetary Regulation Bills (Letras de Regulación Monetaria, LRM) yields and key peso deposit rates.
  - Term premia are economically small at the front end of the curve; they rise at longer maturities but remain moderate and are primarily explained by the country's risk and medium-term inflation expectations.
- Structural and market constraints:
  - Approximately three-quarters of bank deposits remain denominated in foreign currency, constraining peso intermediation and limiting market-based monetary policy transmission.
  - Unsecured peso interbank activity is thin beyond overnight; structural excess liquidity remains substantial.
  - Routine liquidity redistribution is channeled mainly through the central bank’s balance sheet.
- Operational stance to date:
  - The BCU has relied on frequent—but not always entirely predictable—fine-tuning operations that establish a very narrow “mini-corridor” around the TPM within a much wider formal standing facility range.
  - This keeps T1D close to the policy rate but constrains interbank price discovery and market development.

### Recommended operating framework changes (high-level)
- Introduce a weekly fixed-rate, full-allotment seven-day main operation at the TPM as the core liquidity-management tool to reinforce transmission and simplify allocation.
- Recalibrate standing facilities to provide a narrower corridor of ±50 basis points around the policy rate to support short-term liquidity management.
- Continue issuing LRM to absorb structural liquidity and shape the short end of the curve, with issuance concentrated in three standardized maturities: 3 months, 6 months, and 12 months.
- Discontinue the 30-day bill, which duplicates the operational target and fragments liquidity.
- During transition, continue relying on fine-tuning operations with rates around the TPM until the new operational setup fully anchors short-term rates.
- Strengthen liquidity forecasting and data infrastructure; reframe forecasts around a standard liquidity table and adopt the IMF MCMCO liquidity forecasting toolkit to automate model estimation, forecast evaluation, and prediction intervals.
- Strengthen information sharing with the Treasury, including a currency split of fiscal flows and structured coordination on auction calendars, to improve forecasts of government deposits (GovD).
- Consolidate liquidity forecasting and operations execution in a dedicated Monetary Operations and Market Development area within the BCU, with specialist units for execution, short-term liquidity forecasting, market intelligence, market-development initiatives, and back-office support.

### Detailed recommendations (selected items from Recommendation Summary)
- Monetary Policy Framework:
  - Introduce a weekly fixed-rate, full-allotment 7-day main operation at the policy rate as the core liquidity-management tool. Priority: High. Timeframe: Short-term. Paragraph: 82.
  - Calibrate the width of the corridor for deposit and lending standing facilities to at least ±50 bps around the policy rate. Priority: High. Timeframe: Short-term. Paragraph: 109.
  - Ensure standing facilities are accessible to all eligible counterparties meeting collateral and risk management requirements; remove time-based caps and similar quantitative constraints. Priority: High. Timeframe: Short-term. Paragraph: 119.
  - Remunerate peso reserve requirements at the policy rate to improve transmission. Priority: High. Timeframe: Immediate. Paragraph: 96.
  - Unify reserve requirement ratios in pesos across tenors. Priority: Medium. Timeframe: Short-term. Paragraph: 103.
  - Move MRR from contemporaneous averaging to a backward-looking calibration. Priority: Medium. Timeframe: Medium-term. Paragraph: 107.
  - Lower effective reserve requirement ratios in pesos to reduce liquidity parked in non-marketable assets. Priority: Medium. Timeframe: Short-term. Paragraph: 106.
  - De-stigmatize the standing lending facility by clearly separating it from emergency liquidity assistance (ELA). Priority: Medium. Timeframe: Medium-term. Paragraph: 122.
  - Strengthen the organizational capacity of the monetary policy operations area. Priority: Medium. Timeframe: Medium-term. Paragraph: 125.
- Liquidity Forecasting:
  - Reframe liquidity forecasts around a standard liquidity table focused on reserves and treating the reserve requirement as a constraint rather than a forecast object. Priority: High. Timeframe: Immediate. Paragraph: 43.
  - Distinguish forecasting horizons and preserve immutable forecast vintages. Priority: High. Timeframe: Immediate. Paragraph: 44.
  - Systematically produce and use forecast uncertainty (prediction intervals) as operational input for calibrating liquidity buffers and operation sizes. Priority: Medium. Timeframe: Short-term. Paragraph: 49.
  - Evaluate forecast errors regularly and transparently and conduct structured reviews before new monthly forecasts. Priority: High. Timeframe: Immediate. Paragraph: 50.
  - Strengthen information sharing with the Treasury through a formal MoU that provides fiscal reports disaggregated by currency for better GovD forecast. Priority: Medium. Timeframe: Medium-term. Paragraph: 53.
- Money Market Development and Managing Liquidity Surplus:
  - Concentrate LRM issuance in three standardized tenors—3 months, 6 months, and 12 months—to deepen liquidity, support secondary-market trading, and strengthen the reference-rate role of LRMs. Priority: High. Timeframe: Short-term. Paragraph: 137.
  - Standardize LRM auction design and cut-off conventions, and pre-announce fixed issuance amounts. Priority: Medium. Timeframe: Short-term. Paragraph: 143.
  - Adopt LRM maturity concentration and reduced auction frequency to enhance market liquidity and improve price formation. Priority: Medium. Timeframe: Short-term. Paragraph: 147.
  - Maintain, in the short term, a split-curve arrangement—BCU LRMs up to 12 months and Treasury securities beyond—while strengthening cash- and debt-management coordination and assessing over the medium term the merits of a transition toward a single-instrument (Treasury-only) model. Priority: Medium. Timeframe: Medium-term. Paragraph: 155.

### Liquidity forecasting and data/institutional improvements
- Current shortcomings:
  - Forecasts deliver usable point estimates but rely heavily on complex Excel files, tacit knowledge, and incomplete information on government cash flows.
  - Forecast vintages are often overwritten and uncertainty is not systematically quantified.
- Recommended steps:
  - Adopt a standard liquidity table where autonomous factors (e.g., currency in circulation, GovD) feed directly into projections of aggregate reserves relative to the requirement, plus a small buffer.
  - Implement the IMF MCMCO liquidity forecasting toolkit for automated model estimation, forecast evaluation, prediction intervals, and immutable forecast vintages.
  - Strengthen information sharing and coordination with the Treasury (formal MoU, fiscal reports disaggregated by currency, auction-calendar coordination).

### LRM market role and issuance strategy
- LRM dual purpose: absorb structural liquidity surpluses and anchor the short end of the peso yield curve.
- Empirical findings summary:
  - Short-dated LRMs move closely with the policy rate.
  - The 12-month node increasingly reflects expectations and risk premia.
  - Auction results indicate the BCU acts as a price maker at short tenors and more as a price taker at 6- and 12-month maturities.
- Operational recommendations:
  - Concentrate issuance in 3-, 6-, and 12-month standardized maturities with clustered maturities.
  - Discontinue the 30-day bill to avoid duplication of the operational target and fragmentation of liquidity.
  - Standardize auction rules, cutoff conventions, and pre-announce fixed issuance amounts to deepen benchmark lines and foster secondary trading.
  - Maintain a split-curve arrangement with BCU LRMs up to 12 months and Treasury issuance beyond, while improving coordination on calendars and volumes.

### Institutional/organizational recommendations
- Current structure:
  - Liquidity forecasting is within Monetary Policy Management.
  - Execution of LRMs and fine-tuning operations is placed in Asset and Liability Management.
  - A small Monetary Operations Division handles day-to-day implementation.
- Recommended reorganization:
  - Consolidate functions into a dedicated Monetary Operations and Market Development area with mandate to keep the overnight rate close to the TPM.
  - Create specialist units for execution, short-term liquidity forecasting, market intelligence, market-development initiatives, and back-office support.
  - These adjustments are intended to sustain strong rate control as peso markets deepen and de-dollarization advances.

*Source: IMF Technical Assistance Report — Preface, Executive Summary, and Recommendation Summary (tarea2026037).*

### Introduction

### Introduction

### Background and operational framework
- The BCU has conducted monetary policy with a primary mandate of price stability within an inflation-targeting framework since 2005.
- The explicit inflation target range is currently 3–6 percent.
- The BCU controls the marginal cost of peso funding by targeting the overnight interbank peso rate (TMM/T1D), which has served as the operational target since 2020 and is kept closely aligned with the TPM set by the Monetary Policy Committee (COPOM).
- The operational set-up aims to ensure short-term money-market rates reflect the policy stance, providing banks with a clear overnight price signal and a stable anchor for the term structure of interest rates.

### Mission objectives and scope
- Objective: Support the BCU in refining its operational framework so that policy-rate changes transmit to money markets, credit, funding conditions, and the real economy.
- At the authorities’ request, an IMF Monetary and Capital Markets Department (MCM) technical assistance mission reviewed monetary policy implementation, liquidity management practices, and instrument calibration.
- Focus areas:
  - Strengthening the design and use of the operating toolkit—open market operations (OMOs), central bank bills (LRM), standing facilities, and reserve requirements.
  - Modernizing liquidity forecasting so that the overnight rate remains aligned with the TPM under varying conditions.
  - Reducing reliance on ad hoc fine-tuning and creating stronger incentives for market-based liquidity management.

### Institutional reforms since 2020
- Measures undertaken by the BCU:
  - Adoption of the overnight interbank rate as the primary operational target (since 2020).
  - Narrowing the inflation tolerance band from 3–7 percent to 3–6 percent, with greater emphasis on the midpoint.
  - Increasing the frequency of COPOM meetings.
  - Strengthening transparency through improved communication of policy decisions.
- These measures have strengthened credibility and enhanced inflation-control effectiveness.

### Key structural challenges
- High financial dollarization and a shallow peso money market complicate implementation:
  - Approximately three-quarters of bank deposits are denominated in U.S. dollars.
  - A significant share of lending is dollar-denominated.
- Consequences:
  - Many financial contracts respond more to foreign interest rates and exchange-rate expectations than to domestic peso rates.
  - Unsecured interbank trading in pesos is limited; liquidity is often redistributed through central bank operations.
  - Peer-to-peer trades account for only a small share of turnover.
- Operational implications:
  - Given a relatively wide interest-rate corridor around the TPM and structural excess liquidity, the overnight rate would tend to fluctuate more widely—and gravitate toward the floor—without frequent BCU intervention through fine-tuning operations.
  - The BCU has had to intermediate a large share of routine liquidity shocks through its balance sheet to keep the overnight rate on target, limiting market-based price discovery and secondary-market depth for BCU instruments.
  - A more predictable liquidity regime and deeper benchmark instruments can reduce premia embedded in peso assets and support a gradual shift to peso portfolios.

### Roadmap and report structure
- The report outlines a roadmap to modernize the operational framework and enhance liquidity management:
  - Section I: Reviews current monetary policy implementation and proposes adjustments to the operational toolkit—design and use of OMOs, the interest-rate corridor, and reserve requirements—to support more predictable, market-friendly transmission.
  - Section II: Focuses on liquidity forecasting, assessing current practices, and presenting a streamlined, model-based framework to forecast key autonomous factors and system-wide liquidity over relevant horizons.
- Aim: Provide recommendations to align instruments, liquidity forecasts, and market incentives with the BCU’s price-stability mandate while considering high dollarization and the need to deepen the peso money market.

### I. Performance assessment — overview of tools and metrics
- Instruments relied upon:
  - Fine-tuning operations (intensively used for short-term liquidity fluctuations).
  - Issuance of LRM (sterilization tool to absorb excess structural liquidity).
  - Reserve requirement system based on maintenance-period averaging.
  - Interest rate corridor.
- Operational note: Fine-tuning operations are intensively used to manage short-term liquidity and ensure market rates remain aligned with the policy rate.
- Interest-rate corridor: Large spread between lending and deposit rates, ranging from 1 percent to 15 percent, reduces its role in guiding market rates.

### A. Marginal cost of funding
- TPM is set to steer the T1D, the primary proxy for the marginal cost of funding in domestic currency; the framework aims to keep T1D close to TPM.
- T1D design and publication practices:
  - The BCU publishes T1D every business day, with monthly averages and detailed transaction volumes disaggregated into interbank trades and operations with the BCU.
  - Overnight call-peso instruments account for most peso money-market operations; T1D is effectively a volume-weighted average of the main pool of overnight peso funding.
  - Against IOSCO principles, T1D reflects data sufficiency, high quality, and a reasonable degree of representativeness: based on actual arm’s-length overnight transactions, referenced in private contracts and legal provisions, and calculated using a transaction-weighted average methodology.
- Caveat: Need to monitor market depth and concentration over time.

### B. Pass-through to overnight rates
- Pass-through findings:
  - Changes in TPM are quickly reflected in T1D; pass-through is rapid and nearly complete, comparing favorably with international peers.
  - Rolling Root Mean Squared Error (RMSE) analysis: Uruguay’s average deviation of the operational target from the policy rate declines markedly after the adoption of the current operating framework and now fluctuates within or below 16 basis points.
  - Methodology note: Rolling RMSE defined with N=60 daily observations.
- Operational mechanism:
  - Strong steering of overnight rates is achieved through intensive use of the operational toolkit rather than deep market intermediation.
  - With a wide standing facility corridor and persistent structural excess liquidity, maintaining the operational target close to the policy rate requires intensive fine-tuning operations.
  - Routine liquidity redistribution is channeled through the BCU’s balance sheet, constraining price discovery, market development, and broader transmission.
  - Absent structural measures to reduce excess liquidity and foster interbank activity, day-to-day micro-management by the central bank will remain necessary.

### C. Pass-through to term rates and retail rates
- Role of LRM:
  - LRM yield curve serves as the primary building block of the short-end risk-free peso term structure.
  - BCU issues LRMs at different maturities (30, 90, 180, 360 days, and occasionally longer) to absorb structural excess liquidity and provide reference yields along the curve.
- Observed behavior:
  - LRM auction cut-off rates across maturities generally track the policy rate and the interbank overnight rate, indicating investors respond systematically to realized decisions rather than strongly anticipating them.
  - Exception: At the start of the latest easing cycle in mid-2025, markets did not price in cuts ahead of time; 30-day LRM yields adjusted only once the policy rate began to decline.
  - Term-structure premia are asymmetric:
    - During the mid-2021 tightening, the entire LRM curve traded persistently above the overnight rate, with widened spreads at 90–360 days well before policy reached those levels.
    - Around December 2022 and in the most recent period, pockets of negative premia appeared at intermediate maturities as markets anticipated rate cuts ahead of actual easing.
  - Implication: Asymmetry suggests higher ex-ante funding costs in tightening phases without a clearer, systematic forward-looking narrative on the reaction function and likely policy corridor.
- Retail deposit and lending rates:
  - Adopting a clear policy rate target compressed spreads and risk premia across instruments and maturities (LRM, bank deposits, lending rates), aligning them with the interbank market rate within approximately six months after introduction.
  - Tenor-matched comparisons:
    - LRM yields, retail peso deposit rates, and lending rates generally move with the policy rate.
    - Co-movement is stronger for funding (deposits) than for credit (corporate lending).
    - Corporate loan rates respond more slowly and less symmetrically: they incorporate tightening rapidly and remain elevated even as policy and LRM rates fall, reflecting borrower credit risk, collateralization, and product-specific pricing.
    - Loan spreads over policy and LRM rates widened notably during the 2021–23 tightening episode and compressed only gradually during the recent easing cycle, indicating state-dependent and persistent transmission to lending conditions.

*Source: IMF staff calculation using BCU data.*

### 19. While the visual analysis provides an overview of how market rates co-move with the

### tarea2026037 - 19. While the visual analysis provides an overview of how market rates co-move with the

### Short‑run pass‑through model and estimation
- Model structure and purpose:
  - Estimated using an autoregressive distributed‑lag (ARDL) structure (Pesaran, Shin, and Smith, 2001).
  - Relates monthly changes in a given market rate to current and lagged changes in the policy rate, allowing for inertia in the market rate itself.
  - The cumulative short‑term pass‑through is the summary measure: ∑훽푡(푑)푝푖=0 (plotted line); the band is a 90 percent confidence interval computed using HAC (Newey–West) standard errors.
- Estimation approach:
  - Estimated tenor‑by‑tenor in rolling time windows to obtain time‑varying pass‑through coefficients.
  - Maximum lags (p,q) selected by global grid search p∈[0,6], q∈[0,6] using the BIC metric on OLS fit and model complexity.
  - Chosen (p,q) held fixed; equations estimated using fixed‑length rolling windows of 36 months (with a minimum window length of 24 months).
- Model equation (as specified):
  - Δ푟t(푑) = α(푑) + ∑훽t(푑) p i=0 ΔPolicyt−i + ∑γt(푑) q j=1 Δ푟t−j(푑) + εt(푑),
  - Where Δ푟t(푑) is the percentage point change in the market rate for tenor 푑 at year‑month t; ΔPolicyt is the policy rate percentage point change from t−1 to t.

### Key empirical findings on pass‑through to LRMs, deposits, and lending
- LRM (auction cut‑off yields) pass‑through:
  - Short‑end LRM pass‑through is high and has strengthened since 2025.
  - 30‑, 90‑, and 180‑day LRM cut‑offs: estimated pass‑through now lies broadly in the ~0.7–0.9 range and is statistically significant across most windows.
    - Interpretation: a 100 bp policy change is typically associated with a 70–90 bp move in these maturities.
  - 360‑day node: coefficient is smaller and trending downward; 720d included when present.
  - Interpretation: strong operational anchoring at the short end; longer‑dated yields influenced more by expectations, term premia, and limited market depth beyond six months.
- Deposit pass‑through:
  - For firms:
    - Deposit pass‑throughs generally exceed 0.6 and are statistically significant.
    - The ~46‑ and ~76‑day buckets show the highest responsiveness.
  - For households:
    - Very short deposits (~15 days) are often not statistically different from zero.
    - Longer buckets (~136 and ~274 days) are significant, averaging around 0.6.
- Lending pass‑through:
  - Firm loan rates exhibit moderate pass‑throughs, with greater adjustment at longer tenors and less at very short tenors.
  - Household overdraft rates are volatile and frequently indistinguishable from zero — consistent with revolving nature and borrower‑specific risk and fee reliance in pricing.
- Overall transmission pattern:
  - Monetary transmission beyond the overnight segment is effective but slower and heterogeneous across instruments and tenors.
  - Strong short‑run pass‑through to short‑end LRMs and key peso deposit buckets; weaker and more variable effects at longer maturities and in some retail segments.

### Survey evidence and structural drivers of attenuated transmission
- Common patterns across banks:
  - High share of sight and very short‑term deposits in the peso funding base.
  - Higher reserve requirements applied to these liabilities compared to longer‑term deposits.
  - Result: higher effective cost of funding associated with overnight and demand balances; average marginal funding cost in pesos adjusts primarily as maturing term deposits are repriced.
  - Reserve requirements and their remuneration are explicitly incorporated into internal transfer prices and funding‑cost metrics, feeding into loan and deposit pricing.
- Institution‑specific idiosyncrasies:
  - Frequency and aggressiveness of price revisions vary:
    - Many institutions run regular (often monthly) ALM and pricing cycles reacting to policy, yield curves, and competitor rates.
    - Others adopt more gradual adjustments shaped by customer behavior and a preference to avoid frequent changes in posted retail rates.
- Market reference curves:
  - Banks report peso funding costs and transfer prices are closely linked to the sovereign peso yield curve—particularly the sovereign normative yield curve (ITLUP) and LRM nodes—which are adjusted following policy moves.

### Term premium decomposition and determinants
- General pattern:
  - Term premia near zero at one month and moderate—but time‑varying—at 6–12 months.
  - Term premium higher at 180–360 days than at 30 days; overall term premium is lower than 10 percent.
  - Term premia peaked during the 2020–21 COVID‑19 period and compressed thereafter.
- Estimation approach:
  - Monthly term premia for selected LRM tenors (30d, 90d, 180d, 360d) expressed in percentage points.
  - Term premium for tenor n defined as observed LRM yield minus model‑implied average path of the policy rate over the next n months.
  - Expected path obtained from a VAR estimated on monthly policy rate and LRM spreads; BIC used to select lag length.
- Dominant drivers at 6–12 months:
  - USD sovereign yield (Uruguay USD sovereign yield) — a country‑risk proxy — shows the strongest positive association with 90–360‑day LRM premia.
    - Interpretation: local‑currency bills embed a credit component co‑moving with external spreads; when the USD curve widens, investors demand more compensation on peso paper beyond expected short rates.
  - Medium‑term inflation expectations increase premia, with influence rising with tenor.
- Role of exchange‑rate fluctuations:
  - Once USD sovereign yield and 24‑month inflation survey are included, monthly foreign exchange (FX) changes do not show an independent, systematic effect on premia.
- Magnitude by tenor:
  - Statistically insignificant intercept at 30 days; statistically significant positive intercepts at 6–12 months.
  - Conclusion: average premia negligible at one month, present and meaningfully larger at 180–360 days.
- Policy implication:
  - Anchoring inflation expectations materially reduces the term premium; keeping 24‑month expectations inside the 3–6 percent objective range lowers the inflation‑risk component of 6–12‑month yields and improves pass‑through.

### Implications for operational framework and policy recommendations
- Transmission and operational anchor:
  - Strong pass‑through to short‑end LRMs and core peso deposits consistent with tight control of the overnight operational target and centrality of ITLUP‑ and LRM‑based references in banks’ pricing.
- Frictions that slow or unevenly distribute pass‑through:
  - Prominence of short‑maturity deposits subject to higher reserve requirements.
  - Limited depth of the long end of the peso curve.
  - Product‑level rigidities in retail pricing.
- Recommended focal areas:
  - Further deepen term peso markets.
  - Refine the design and role of reserve requirements.
  - Strengthen the informational content of the peso yield curve so changes in the TPM translate more smoothly and uniformly into banks’ overall marginal cost of funding in local currency.
- Limits to pass‑through:
  - Pass‑through to retail lending rates not determined by liquidity instruments alone; competition in peso lending and deposit markets, banks’ risk appetite, capital positions, and excess liquidity also matter.
  - Proposed operating reforms improve the risk‑free peso reference curve and predictability of marginal funding conditions, but completeness of pass‑through to credit conditions will continue to depend on competitive conditions and borrower risks.

### Calibration of open market operations and liquidity forecasting
- Ideal benchmark:
  - With precise liquidity forecasts and a well‑functioning interbank market, the central bank could control the marginal cost of funding with a single OMO per period by offsetting the system’s net liquidity need.
- Practical constraints:
  - Forecast errors and market frictions require the BCU to calibrate both quantity of liquidity absorbed/supplied and the mix of instruments.
  - When market is less developed or segmented, heterogeneous access to funding and willingness to trade lead to unstable demand for excess reserves; central bank interventions become larger, more frequent, and more instrument‑intensive.
- Two pillars for effective calibration:
  - Liquidity forecasting: combines projections of autonomous factors (CiC, government balances, FX operations) with banking system demand for required and excess reserves to determine aggregate supply–demand for central bank money.
  - Market functioning: depth and reliability of interbank activity determine how aggregate positions translate into operational choices.
- Current BCU forecasting practice:
  - Distinct liquidity forecasting unit produces autonomous factor forecasts across horizons.
  - Main forecast conducted towards the end of the month to predict autonomous factor movements for each business day of the following month; day‑by‑day adjustments made for deviations.
  - Weekly (Friday) forecasts for the next two weeks refine initial monthly forecasts; a medium‑term forecast anticipates autonomous factors over several months.
  - Process delivers usable point forecasts for operations but currently relies on complex, Excel‑based procedures, tacit knowledge, and incomplete Treasury information, and does not systematically quantify and use uncertainty.

*IMF staff calculation using BCU data.*

### 36. The liquidity forecast informs LRM issuance and day-to-day operations. The forecasting

### 36. The liquidity forecast informs LRM issuance and day-to-day operations.

### Forecasting process and operational linkage
- The forecasting team proposes a daily issuance path for the next "two to four weeks"; the monetary policy department contributes tenor and volume views based on counterparty outreach, and decisions are reconciled at the senior level.
- LRM auction calendars are fixed "two weeks" in advance, creating a sharp deadline for the two-week forecasts on each week’s Fridays.
- Objective of the projection: estimate evolution of system-wide liquidity to identify mismatches requiring BCU action and translate these into a consistent plan for LRM issuance and the use of fine-tuning instruments (ISF).
- Adjustments to the issuance plan by tenor directly impact liquidity and add deviations from forecasted liquidity on top of forecast errors; such deviations are offset with fine-tuning operations to stabilize money-market conditions.
- Operational objective at the start of the period is near-zero excess reserves; in practice, a small surplus often remains.
- Detailed implementation or rate setting for the daily OMO/FTO is outside the remit of the forecasting unit.

### Cash-flow projection and institutional blocks
- The monetary policy department embeds forecasts in a cash-flow-based projection by broad institutional blocks using daily spreadsheets that combine historical and projected balances.
- Step 1: estimate liquidity position and investment capacity of non-bank investors (notably pension funds and other institutional investors), whose LRM placements tend to flow back into the banking system through deposit certificates within pre-defined limits.
- Step 2: analyze the banking system with the same daily framework: historical days record actual reserve holdings and are used to infer a target reserve level for the remainder of the month, while autonomous factors projections are received from the liquidity forecasting unit.
- The BCU chooses the mix of LRM issuance and fine-tuning operations so that, by month-end, system reserves are consistent with the estimated target (within a tolerance band reflecting the volatility of minimum and voluntary reserves) and the overall liquidity balance remains positive and in line with historical practice, with ISF operations used as a residual timing-adjustment tool.

### Key autonomous liquidity drivers and forecast performance
- Principal autonomous factors driving domestic liquidity: CiC and government deposits (GovD).
- Other autonomous items, such as the net foreign asset (NFA) position, are not forecasted, reflecting the central bank’s minimal participation in the FX market in recent years; larger NFA swings were mainly observed before 2021, when the BCU purchased FX.
- Liquidity Management Deposits by NBFIs, including pension funds, are considered autonomous factors when material but are currently not a significant source of liquidity variation and are not forecast.
- Government cash flows are the largest and most volatile driver of liquidity and form the most challenging component of the forecast.
- Available evidence suggests BCU’s CiC forecasts are broadly accurate, whereas GovD projections are more mixed.
- A comprehensive assessment of BCU’s current forecast performance is not entirely possible in the absence of preserved, unadjusted vintage forecasts.

### Findings on intramonth patterns and bank behaviour
- Evidence from "November 2024" onward: most institutions hold fulfillment ratios below "110 percent" of their monthly requirements, and the vast majority of bank-month observations lie below "120 percent".
- Typical excess reserves are in the order of only "0–10 percent" above the minimum, with very few cases of materially larger precautionary balances.
- Intramonth fulfilment paths: cumulative positions often deviate by more than "±30 percent" from the eventual requirement during the period; many banks move into a cumulative deficit relative to the final requirement after around "day 25".
- Banks actively use averaging; pattern supports BCU practice of supplying a modest liquidity buffer early in the maintenance period and gradually withdrawing it later.
- Implication: small system-wide excess-reserve cushions imply forecasts must be very precise or operations must be adjusted frequently; relying on a single monthly forecast concentrates realignment toward the end of the period.

### Forecast uncertainty, evaluation, and recommended diagnostics
- Recommendation to systematically produce and use forecast uncertainty (prediction intervals) as an explicit operational input for calibrating liquidity buffers and operation sizes.
- The R toolkit provides measures of forecast variance and prediction intervals; forecasts can be accompanied by prediction intervals (for example, "80 and 95 percent") that widen with horizon and reflect model-specific forecast-error variance.
- Coverage diagnostics should be monitored routinely: persistent under-coverage triggers review of model variance estimates or event design; over-coverage signals scope to sharpen the bands.
- Interval communication is especially critical around month-end, where larger swings are typical.
- Recommendation to evaluate forecast errors regularly and transparently, including structured reviews before new monthly forecasts and periodic model re-evaluation to detect structural changes early.
- Model evaluation protocol: out-of-sample, expanding rolling-origin scheme; for each origin, models are re-estimated using all information available at that point, H-step forecasts are produced for fixed horizons (e.g., "5, 10, 20 business days"), and errors are collected before advancing the origin.
- Performance metrics to report: Mean Error (ME) for bias; MAE and RMSE for absolute accuracy; interval-oriented scores (e.g., mean interval score) and empirical coverage rates for predictive distributions.
- Current gap: forecasts are sometimes updated and overwritten mid-month, preventing ex-post error evaluation.

### Forecast governance, data sharing, and operational recommendations
- Recommendation 1: Reframe liquidity forecasts around a standard liquidity table focused on reserves, avoiding references to “money supply” or “money demand”, and treat the reserve requirement as a constraint rather than a forecast object.
- Recommendation 2: Distinguish forecasting horizons and purposes (short-term, one-month-ahead, and medium-term) and preserve immutable forecast vintages to improve operational use and separate statistical forecasts from judgment.
  - Maintain an Excel (or equivalent) file recording production date, target date, point forecast, and variance; separate statistical forecast and expert judgment.
  - Short-term intra-month forecasts covering a few days should be stored and compared with previous forecasts for the same target date rather than overwritten.
  - Medium-term forecasts spanning several months should be produced as monthly averages using temporal hierarchies rather than detailed day-by-day paths.
  - The proposed R-based toolkit operationalizes saving immutable forecast vintages with production date, target date, model set, and interval parameters.
- Recommendation 3: Increase robustness and frequency of forecasts once the new R-based toolkit is in place and calibrated; the forecasting unit should produce forecasts at weekly frequencies once automated.
  - Automation reduces workload for a small team by automating data ingestion, estimation, evaluation, and reporting.
  - More frequent forecasts improve calibration of liquidity-absorbing operations and alignment between structural issuance and short-term fine-tuning.
- Recommendation 4: Systematically produce and use forecast uncertainty (prediction intervals) as an explicit operational input for calibrating liquidity buffers and operation sizes.
- Recommendation 5: Evaluate forecast errors regularly and transparently; hold brief discussions before producing new monthly forecasts and record written comments on primary sources of recent forecast errors; re-evaluate models at least every "six months" or when statistical performance deteriorates.
- Recommendation 6: Strengthen information sharing with the Treasury through a formal MoU that provides fiscal reports disaggregated by currency for better GovD forecast.
  - Current fiscal reports combine local and foreign currency flows; disaggregation by currency is essential in a highly dollarized economy.
  - Government cash flows: spending and most tax collection are reflected in the government’s account at BROU; debt issuance, redemptions, and interest are settled in the government account at the BCU.
  - Daily fiscal result at BROU is first projected with a statistical model, then expert judgment layers calendar effects (wages, pensions, tax deadlines, June/December extra-salary payments).
  - BCU↔BROU transfers are modeled; these transfers are reserve-neutral because government deposits at BROU are subject to a "100 percent" reserve requirement.
  - For policy analysis, focus is on the “restricted monetary base,” excluding the government’s BROU balance to capture liquidity available to the private sector.
- Additional operational note: government bond auction behavior creates uncertainty (government may deviate from planned auction amounts; LRM holders can pay with them in government bond auctions), making immediate liquidity impact uncertain.
- Time-series characteristic: government deposits show a strong intramonth pattern—deposits begin to rise "8 business days" before month-end and sharply decline on the first day of the month; distance to month-end is the relevant seasonal factor.

### Forecasting methods and model governance
- IMF-designed forecasting framework: compact, transparent set of time-series models for each target and horizon; either selects the top-performing model or combines several strong alternatives.
- Model menu includes: Naïve and seasonal-Naïve benchmarks; ETS; ARIMA/SARIMA with optional regressors and targeted intervention terms to capture end-of-month and start-of-month effects; supervised machine-learning methods (Random Forests) as candidate models.

*Source: tarea2026037 - 36. The liquidity forecast informs LRM issuance and day-to-day operations. The forecasting*

### 58. Random Forests (RFs) are well-suited to financial time series with complex, nonlinear

### 58. Random Forests (RFs) are well-suited to financial time series with complex, nonlinear patterns and rich calendar effects

### Random Forests model design and features
- RFs are ensemble models combining many simple decision trees into a single predictor without imposing strong distributional form assumptions.
- Models naturally capture interactions between calendar variables, lagged values, spike indicators, and trend measures.
- Two specifications described:
  - One forecasts a monthly mean and daily deviation around that monthly mean.
  - The other focuses directly on day-to-day changes anchored to the latest observations, exploiting both level and pattern information.
- Engineered features include:
  - day of week and month,
  - business days to month end,
  - quarter end flags,
  - lagged levels and changes,
  - moving averages,
  - indicators for spike windows.
- Model complexity controlled through number and depth of trees and minimum node size to avoid overfitting amid volatile observations.
- Prediction intervals derived from the empirical distribution of past forecast errors at each horizon.

### Calendar structure and seasonality handling
- Main seasonal components for both CiC and GovD are monthly.
- Seasonal features handled via:
  - regressors (end-/start-of-month and business-day counters),
  - moving-holiday flags,
  - short-memory dynamics for fast reversion.
- Sparse trigonometric (Fourier) terms configurable for longer cycles where appropriate.
- Candidate calendar terms may be pre-screened to reduce overfitting.
- Reports include seasonal diagnostics and event-day overlays to verify capture of salary payments, tax deadlines, and other events.

### Forecast outputs, horizons, and coherence
- Outputs produced at needed operational frequencies:
  - daily paths at 7- and 30-day horizons,
  - monthly averages at a four-month horizon.
- Multi-month forecasts with differing horizons/aggregations are made coherent by construction so daily paths align with monthly averages.
- Standard deliverables generated: tables, fan charts, exception logs.
- Statistical forecast is separated from any judgmental overlay incorporating one-off fiscal or cash-management information.
- Core tasks mirrored: model suite selection, out-of-sample evaluation, reconciliation, interval estimation, and production deployment.

### Reconciliation across horizons
- Coherence enforced through reconciliation for long-term forecasts: daily trajectories aggregate exactly to corresponding monthly means.
- Available reconciliation schemes calibrated on a rolling validation window:
  - simple “bottom-up” variant adjusts aggregate to match sum of components,
  - formal approaches (e.g., regression-based weights or MinT) minimize forecast-error covariance.
- Empirical finding: bottom-up and MinT tend to outperform naïve base forecasts and yield similar gains at short horizons.
- Reconciliation results reported alongside unreconciled metrics to document value-add and avoid over-fitting to a single period.

### Model evaluation and accuracy metrics
- Evaluation follows Recommendation 5 principles:
  - expanding rolling-origin scheme yields horizon-specific performance profiles,
  - fixed-horizon evaluation avoids mixing different prediction problems.
- Reported metrics include ME, MAE, RMSE alongside interval-based scores and empirical coverage rates.
- Event-conditioned diagnostics for month-end and first-business-day episodes are included.
- Structure produces stable, comparable accuracy estimates across model families and supports data-driven selection or combination without relying on in-sample fit.

### Forecast uncertainty and diagnostics
- Prediction intervals (for example, 80 and 95 percent) generated for all horizons; coverage diagnostics tracked over time.
- Fan charts and tables of interval widths by horizon are part of standard outputs.
- Systematic deviation of coverage from target triggers review of variance estimation or event design.

### Implementation, governance, and operationalization
- Implementation automated through an R toolkit:
  - pipeline ingests source data from an Excel file,
  - runs estimation and rolling evaluation,
  - performs selection/combination and reconciliation,
  - exports standardized outputs (data tables, HTML reports, and a machine-readable “exception” log).
- Immutable forecast vintages saved with production date, target date, model set, and interval parameters for ex-post evaluation and transparent change control.
- A brief user guide and parameter catalog are maintained in Annex IV.
- Primary operational outputs for decision-makers: point forecasts, prediction intervals, and concise accuracy diagnostics.
- Approach addresses operational risks from manual, Excel-based processes relying on tacit knowledge and visual checks.

### CiC (Currency in Circulation) empirical findings
- CiC exhibits a stable intramonth profile and a stable upward-moving trend (Figures referenced).
- Past two years: stable increasing trend with a minor weekday effect—slightly higher CiC on Thursdays and Fridays.
- Salary-related peak at the turn of the month followed by a gradual decline of roughly 10 percent towards the month-end.
- Dollarization is a minor factor; no significant stock of USD banknotes circulating domestically.
- Payment habits: uneven card acceptance; cash prevalent in household services; some households prefer to hold cash balances.
- Operational lens shifted to CiC held by the public rather than total notes outstanding because vault cash counts toward minimum reserve requirements (MRR).

### BCU’s current CiC modeling and benchmarking
- BCU’s approach estimates monthly averages using an ARIMA model and applies back-tested day-of-month factors.
- BCU tested additional regressors without material improvement in performance.
- Ongoing vintage forecast adjustments at BCU prevent precise historical benchmarking against IMF recommended models.
- Forecast performance reported sufficiently accurate for operational calibration; day-of-month factors found well calibrated.
- Extra salary payments in June and December addressed based on expert judgment.
- Latest BCU and IMF forecasts over the same 40-day horizon show a close overlap.
- Forecast cut-off date referenced: October 13, 2025 (in Figure 12 notes).

### GovD (Government Deposits) patterns and RF performance
- Government deposits exhibit a pronounced intramonth pattern: deposits rise ahead of month-end and drop sharply on the first day of the month.
- Variance and exact timing of the month-end spike are major sources of uncertainty and principal drivers of forecast errors at BCU.
- RF specification including distance to month-end as a feature performs better than time-series models.
- Time-series models with seasonal components and regularization are unsuitable for GovD due to varying business days per month and sharp spikes.
- Greatest potential to improve GovD forecasts lies in closer information exchange with the government.

### Market functioning and depth (peso money market)
- Peso money market is generally functional but shallow beyond overnight maturities; banks hoard liquidity and rely on BCU fine-tuning operations and reserve requirement averaging.
- BCU acts as a dominant counterparty and performs daily fine-tuning through ISF operations to absorb short-term liquidity shocks.
- By November 2025:
  - total turnover in overnight “Call pesos garantizado” operations reached UYU 626 billion,
  - approximately 41 percent originated from BCU fine-tuning operations.
- Call pesos garantizado is a short-term secured instrument mainly collateralized by Treasury and BCU-issued securities.
- Unsecured call operations accounted for roughly 3 percent of total turnover.
- Term unsecured trades beyond overnight were negligible.
- Interbank market for maturities longer than overnight remains shallow, constraining price discovery and broader money market development.
- Survey evidence: market functioning for most maturities still relies on BCU’s active footprint; depth is described as “not deep”; liquidity concentrated within the week and thins markedly beyond one month.
- Practical development agenda suggested:
  - deepen call garantizado or repo activity beyond overnight,
  - broaden the peso investor base,
  - strengthen short-end reference points,
  - complement with daily-average reserve-requirement flexibility for intramonth smoothing.

### Market concentration, turnover, and implications for secondary markets
- Trading in primary and secondary markets concentrated in a narrow set of instruments.
- Primary issuance: almost all split between private-sector CDs and LRMs (about 96 percent of volumes).
- Secondary market: trading overwhelmingly in LRMs (nearly 90 percent); marginal activity in Treasury notes, international issuances, and other securities; effectively no trading in private CDs.
- Table 3 market shares (percent):
  - LRMs: Primary market share 45.1; Secondary market share 88.5
  - Treasury notes: Primary market share 2.5; Secondary market share 3.9
  - Private sector (CDs): Primary market share 50.6; Secondary market share 0.0
  - International issuances: Primary market share 0.0; Secondary market share 7.0
  - Other securities: Primary market share 1.9; Secondary market share 0.6
- Turnover by issuer sector and trading venue (UYU mi) summarized in Table 4; total all segments: 3,340.312
- July 2024 to June 2025 totals:
  - total market operations approximately 3.3 trillion pesos,
  - roughly three-quarters (approximately 2.5 trillion pesos) primary issuance,
  - about one-quarter (around 0.8 trillion pesos) secondary trading.
- Private sector instruments: secondary market trades account for only about 0.1 percent of total private-sector volumes (1.7 billion out of 1.3 trillion pesos), implying a secondary-to-primary ratio close to zero.
- Public sector: secondary operations represent about 40 percent of total public sector volumes.
- Low turnover indicates buy-and-hold structure that limits price discovery and market liquidity, hampers reliable market prices, constrains portfolio adjustment and risk transfer, and reduces usefulness of securities as collateral.
- Shallow secondary markets can weaken monetary transmission through asset prices and yields and make domestic capital markets a less flexible, market-based source of finance.

*IMF Technical Assistance Report*

### 76. Outside the overnight segment, the BCU is the counterparty in around 60–80 percent of all

### Outside the overnight segment, the BCU is the counterparty in around 60–80 percent of all peso PG transactions across most maturities and over long periods.

### BCU footprint in the peso money market
- Finding: Outside the overnight segment, the BCU is the counterparty in around 60–80 percent of all peso PG transactions across most maturities and over long periods.
- Finding: The market does not routinely clear on a purely interbank basis; only the overnight tenor exhibits some degree of depth without continuous BCU presence.
- Interpretation: Periods of temporarily lower BCU share are largely attributable to more flexible use of LRM auction calendars, not to a structural increase in private interbank activity.
- Observation: A sound market infrastructure exists via the electronic exchange BEVSA, but the dominant central-bank footprint implies the system remains reliant on the BCU for redistribution and is not yet a deep, liquid peso money market in its own right.

### Interaction between ISF fine‑tuning and reserve averaging
- Finding: Daily ISF fine-tuning, rather than reserve averaging and market trading, currently absorbs most short-term liquidity shocks.
- Empirical pattern (Figure 15 analysis):
  - Daily deviations are plotted: x-axis = day’s required-reserve balance minus month’s average; y-axis = day’s ISF stock minus month’s average (both in UYU millions).
  - A sizeable share of observations falls in the “offsetting” quadrants where banks’ use of averaging and BCU ISF actions move in opposite directions.
  - Many days show banks running reserves below the monthly mean while ISF operations simultaneously absorb more liquidity than usual; conversely, days exist when banks hold reserves above average while the BCU absorbs less or injects.
- Implication: The system behaves as if daily OMOs are the first line of defense, with averaging residual; this weakens incentives for active liquidity management by banks and makes the overnight rate heavily dependent on day-to-day central bank operations.
- Recommendation (implied): Move toward a framework where a regular weekly main operation combined with reserve averaging absorbs most predictable shocks, leaving ISF windows as a genuine backstop.

### Proposed instruments and overall objective
- Objective: Reinforce monetary policy transmission by strengthening the link between the policy rate and banks’ marginal cost of peso intermediation over horizons relevant for pricing decisions.
- Current toolkit in Uruguay: (i) short-term OMOs, including frequent fine-tuning operations (ISF); (ii) reserve requirements; and (iii) a wide standing-facility corridor that serves as a formal backstop but is overshadowed by daily OMO interventions.
- Reform focus: Reduce the central bank’s operational footprint and promote market development through clearer instrument hierarchy and predictable routines.

### Main recommendation — weekly 7-day FRFA main operation
- Recommendation 1: Introduce a weekly fixed-rate, full-allotment 7-day main operation at the policy rate (TPM) as the core liquidity-management tool to reinforce transmission and simplify allocation.
- Rationale and expected effects:
  - Preserve Uruguay’s excellent short-term rate control while replacing near-daily micromanagement with a predictable routine.
  - Under a weekly 7-day FRFA at the TPM, banks plan liquidity over the maintenance period, redistribute idiosyncratic shocks via the market, and internalize opportunity costs between operations.
  - Promote more interbank trading, better internal liquidity forecasting and planning, more effective use of reserve averaging, and improved transmission between the policy rate and banks’ marginal peso funding costs.
- Operational design features:
  - Weekly 7-day FRFA at the policy rate with full allotment against eligible collateral for all eligible counterparties (e.g., a fixed weekday tender).
  - LRMs remain the main tool for structural sterilization and for shaping the short end of the yield curve.
  - Standing facilities provide the corridor; ISF/PG-Call re-positioned as exceptional backstop instruments.
  - Policy rate changes could become effective at the next scheduled seven-day FRFA operation date.
  - Supplementary shorter-maturity operations may be used only for extraordinary short-term liquidity fluctuations, assessed via short-term liquidity forecasts and corridor capacity.
- Calibrated steady-state split (staff indicative):
  - Required reserves: UYU 29,398 million.
  - Outstanding 7-day operations: around UYU 73,155 million.
  - Remaining structural surplus absorbed through LRMs.
- Simulation evidence:
  - Simulations (Annex III) reconstruct historical liquidity and compare three illustrative weekly rules; results indicate total short-term OMO volume would be lower than under the current regime.
- Operational benefits:
  - Weekly tender “resets” system liquidity once per week at the policy rate, while reserve averaging and the corridor absorb day-to-day noise.
  - Maintains close alignment of overnight rate to TPM with far fewer transactions and a clearer division of labor between instruments.

### Evidence on current BCU intervention intensity and mini-corridor
- From 2020 to mid-2025:
  - The BCU conducted 2,380 ISF interventions across 1,019 days.
  - Average of 38 operations per month, spread over 16 ISF days.
- Mini-corridor statistics inside the formal corridor:
  - Mini-corridor width averaged 7.3 bp, median 5 bp, P90 20 bp.
  - Mini-corridor narrowed over time from 16.3 bp in 2020 to 3.3 bp in 2025.
  - Formal standing-facility corridor is 1–15 percent.
- Interpretation: Intensive, irregular central bank footprint in the overnight segment creates a de facto narrow mini-corridor close to TPM through frequent ISF fine-tuning.

### Role and re-positioning of fine‑tuning operations
- Current role: ISF/PG-Call used frequently as near-daily fine-tuning at or near the policy rate.
- Recommendation: Re-position ISF/PG-Call as exceptional instruments, reserved for clearly defined events (large unforeseen fiscal or cash-flow shocks, technical settlement problems, other events causing temporary deviations that cannot wait until the next weekly tender).
- Intended effect: Clarify hierarchy of instruments—weekly 7-day FRFA primary, LRMs structural, standing facilities corridor, fine-tuning marginal.

### Reserve requirements and averaging profile
- Empirical reserve behavior:
  - Banks tend to start the maintenance period with reserve balances about 30 percent above the required minimum and run them down toward month end.
  - Substantial heterogeneity exists across institutions and over the maintenance period.
  - Toward the end of the period, especially after around Day 25, an increasing share of banks moves into a cumulative deficit relative to the final requirement.
- Interaction with ISF:
  - Scatter plot of daily deviations shows the BCU often adjusts fine-tuning operations in a way that offsets, rather than complements, banks’ use of averaging (many days with reserves below monthly average coincide with ISF absorption above average, and vice versa).
- Implication: Better integration of the reserve-averaging profile with OMO calibration and the proposed weekly FRFA would support intended use of averaging and reduce reliance on daily central-bank interventions.

*Source: IMF staff analysis as presented in the provided content.*

### 95. The banks surveyed report that they actively manage their BCU account balances against

### tarea2026037 - 95. The banks surveyed report that they actively manage their BCU account balances against

### Banks' operational management of BCU reserves
- Banks target a small positive buffer (often around one day of required reserves) to avoid compliance risk while minimizing opportunity cost of holding low-remunerated excess reserves.
- Volatility in large corporate sight balances leads some institutions to hold a technical reserve buffer above the minimum.
- Excess balances beyond the operational cushion are usually redeployed into money-market instruments, indicating demand for excess reserves in pesos is limited and driven mainly by operational, not precautionary, motives.

### Reserve requirement roles and scale
- Reserve requirements differentiate between peso, inflation-indexed (UI), and foreign-currency deposits because these liabilities serve different policy objectives.
- Peso reserve requirements: primarily a monetary policy and liquidity management tool.
- UI reserve requirements: preserve the real value of prudential buffers on inflation-linked liabilities (adequate safeguard in real terms rather than nominal liquidity tool).
- Foreign-currency reserve requirements: mainly macroprudential, to contain currency-induced credit risk and ensure liquid FX buffers.
- Reserve balances at the BCU account for around 9–10 percent of the country's GDP.
- Compliance measured on a monthly-average basis (“período de cómputo”); vault cash in pesos contributes to the peso reserve base.

### Current remuneration and its effects
- Required reserves in pesos are remunerated at 60 percent of the TPM.
- Wedge between policy rate and reserve remuneration implies an implicit tax: return on required portion of peso deposits is structurally below the policy rate.
- Banks incorporate reserve requirements and remuneration into internal transfer-pricing frameworks; treasurers view the reserve requirement as a non-negligible wedge affecting short-maturity peso deposits.
- Two monetary-implementation effects of the wedge:
  - Raises funding costs relative to the policy rate, passed to depositors (via lower deposit rates) and borrowers (via higher lending rates).
  - Blurs link between policy rate and benchmark pricing; internal transfer rates and retail pricing may adjust less than one-for-one to policy moves.

### International practice and recommendation on remuneration
- International practice: remunerate required reserves at or near the policy rate in interest-rate-based frameworks.
- Recommendation 2: Remunerate peso reserve requirements at the policy rate to improve transmission.
- Mission recommends raising remuneration on peso reserve requirements to the policy rate (or a narrow spread below it) over a clearly communicated transition path to:
  - Eliminate the reserve requirement tax on peso intermediation.
  - Tighten the link between the policy rate and banks’ transfer prices, reinforcing the TPM as the unique anchor for the marginal cost of funds in local currency.
  - Support development of the peso yield curve by making term peso funding more attractive.

### Recent regulatory change and operational details
- BCU issued Circular No. 2491, published on November 27, 2025, updating remuneration framework.
- Under the new rules, effective March 1, 2026:
  - Required reserves in local currency will earn interest equal to 100 percent of the policy rate (up from 60 percent).
  - Excess balances in local currency will be remunerated at the policy rate minus 2 percentage points (with a zero-floor rate).
- Floors and nominal cutoffs imply incentive effects are not linear in the policy rate; mission recommends minimizing absolute cutoffs and anchoring remuneration to the policy rate.

### Foreign-currency reserve remuneration and guidance
- For foreign currency reserves under new rules:
  - Required reserves on euro- and dollar-denominated deposits with maturities under 30 days will earn the policy rate minus 1 percentage point.
  - Other required reserves will earn the policy rate minus 0.5 percentage point (never below zero).
  - Excess balances in foreign currency accounts will be remunerated at the policy rate minus 1 percentage point, with a zero floor.
- Applicable policy rate for foreign currency is that of the currency issuing authority; if no target rate exists, the one-day market reference rate applies; in a band, the upper limit is used.
- Any changes to FX reserve requirement levels should be assessed with prudential authorities, considering banks’ FX liquidity profiles, availability of high-quality liquid FX assets, and incentives for dollar deposit intermediation.
- Mission suggests regular reassessment of remuneration of excess balances in foreign currency as part of dollarization reduction strategy.

### Recommendations on reserve requirement ratios and levels
- Recommendation 3: Unify reserve requirement ratios in pesos across tenors to better align liquidity-management incentives with the operating framework, reduce distortion, and simplify application.
- Current peso/UI reserve requirement schedule:
  - 15 percent for demand and time deposits with maturities of less than 30 days
  - 3 percent for maturities of 30–90 days
  - 2 percent for maturities of 91–180 days
  - 1 percent for maturities of 181–366 days
  - 0 percent for deposits with maturities exceeding one year
- Foreign currency: 28 percent for deposits with maturities of up to 180 days and 20 percent for maturities exceeding 180 days; 28 percent applies to non-resident FX liabilities net of eligible FX assets.
- Rationale for unification:
  - Effective reserve requirement ratio has remained stable at about 13 percent since 2022 despite cuts for longer maturities.
  - More than 80 percent of the reserve requirement base consists of deposits with maturity of less than 30 days; share of longer-term deposits increased modestly from 11 to 18 percent after reductions.
  - Maturity-differentiated schedule interacts awkwardly with monthly averaging and weekly main operation; aggregate demand over maintenance period matters more than original deposit maturity.
  - Banks manage liquidity mainly through internal ALM limits, wholesale funding plans, LRMs, and government securities rather than exploiting reserve differentials.

- Recommendation 4: Lower the effective reserve requirement ratios in pesos to reduce liquidity parked in non-marketable assets.
  - Current effective MRR about 12.5–13 percent is high relative to typical deposit volatility.
  - Across all banks and business days in the sample:
    - A day-to-day fall in the RR base in excess of 12 percent occurs in only about 2.4 percent of observations.
    - Declines exceeding 10 percent occur in about 3.1 percent.
    - Changes larger than 8 percent are observed on roughly 4.3 percent of days, concentrated in a small number of institutions.
  - Reducing the effective MRR by several percentage points would still cover tail of observed daily outflows for most banks and free structural liquidity to support market-based liquidity redistribution.
  - Most inflation-targeting central banks set reserve requirement ratios below 7 percent; an effective ratio above 12 percent places BCU far above international means.

- Table: Increase in share of bank-day observations with deposit outflow exceeding requirement ratio, assuming starting point effective reserve requirement ratio of 13 percent:
  - Decrease in the effective reserve requirement ratio in percentage points → Increase in the share of bank-day observations with deposit outflow exceeding requirement ratio
    - 1 → 0.20
    - 2 → 0.58
    - 3 → 0.96
    - 4 → 1.62
    - 5 → 2.17
    - 6 → 3.03
    - 7 → 3.99
    - 8 → 5.6

### Recommendation on accounting and averaging
- Recommendation 5: Move MRR from contemporaneous averaging to a backward-looking calibration to reduce operational complexity.
  - Current contemporaneous monthly average means banks only know exact requirement at the end of the maintenance period.
  - Bank-level data since November 2024 show the monthly average deposit base is relatively stable; using a lagged—backward-looking—measure (last month’s average RR base) would deliver a required reserve level very close to the ex-post contemporaneous requirement for most institutions.
  - Backward-looking calibration would fix the target at the start of the maintenance period, simplifying liquidity planning and reducing late-period adjustments.
  - Backward-looking reserve accounting is used internationally (Eurosystem, Bank of Japan, Brazil).

### Interest-rate corridor and standing facilities
- Recommendation 6: Calibrate the width of the corridor for deposit and lending standing facilities to at least ±50 bps around the policy rate, ensuring it remains wider than typical market transaction costs to avoid excessive use of the facility and encourage market transactions.
- Current standing-facility corridor is very wide; overnight deposit and credit facilities are currently set at 1 and 15 percent.
- TPM has fluctuated within a much narrower range due to frequent ISF operations creating a de facto “mini-corridor” around the policy rate.
- Consequences:
  - Standing facilities perceived as backstops rather than routine reference points.
  - Mini-corridor via fine-tuning gives excellent short-term rate control but increases central-bank footprint and reduces guidance from the formal corridor.
- In modern frameworks, corridors are measured in tens of basis points around the policy rate; corridor should (i) provide hard bounds for overnight rate, (ii) create incentive for interbank trading by being wider than all-in transaction costs, and (iii) support operational target by discouraging systematic recourse to facilities.
- BCU’s 1–15 percent corridor is unusually wide and disconnected from money-market pricing; frequent ISF/PG-Call operations at or near the TPM are the de facto channel for short-term liquidity.

*IMF Technical Assistance Report — extracted content unit*

### 113. Across central banks, the width of the corridor ranges from approximately 50 basis points

### Across central banks, the width of the corridor ranges from approximately 50 basis points to over 200 basis points

### Interest rate corridor international evidence and implications
- Across central banks, the width of the corridor ranges from approximately 50 basis points to over 200 basis points.
- Some central bank practitioners view corridors of around 200 basis points as striking a reasonable balance between rate control and market development.
- Distribution concentrated between 50 and 200 basis points:
  - Many inflation targeters operate in the 50–150 basis point range.
  - A smaller group utilizes very wide corridors of 200 basis points or more.
- Implication for the BCU:
  - Re-align the BCU corridor with ranges sufficiently wide to foster market discipline, but not so wide as to render the corridor irrelevant for day-to-day pricing.
  - A corridor of at least 50 basis points around the policy rate can be a pragmatic starting point, with scope to adjust over time.

### Table 6 — Interest Rate Corridor Widths in Selected Countries (2013 and 2018)
- Corridor width 50 bps and less
  - 2013: Australia, Canada, Chile, Colombia, Czech Republic, Israel, Kazakhstan, Malaysia, New Zealand
  - 2018: Albania, Australia, Brazil, Canada, Chile, Israel, Jordan, Malaysia, New Zealand, Saudi Arabia, United Kingdom
- Corridor width 75 bps
  - 2013: Jordan, United Kingdom
  - 2018: Denmark, European Central Bank
- Corridor width 100–149 bps
  - 2013: Afghanistan, Egypt, Fiji, Norway, Thailand, Tunisia
  - 2018: Czech Republic, Egypt, Fiji, Hungary, Norway, Paraguay, Philippines, Sri Lanka, Thailand
- Corridor width 150–199 bps
  - 2013: Euro Area, Sweden, Peru
  - 2018: Indonesia, Peru, Sweden
- Corridor width 200 bps
  - 2013: Colombia, DRC, Costa Rica, Dominican Republic, Guatemala, Hungary, Indonesia, Korea, Morocco, South Africa
  - 2018: Colombia, Costa Rica, Georgia, Guatemala, Iceland, Kazakhstan, Korea, Morocco, Pakistan, Poland, Romania, Russian Federation, South Africa, Tunisia
- Corridor width 200 bps and more
  - 2013: Albania, Bhutan, Croatia, Honduras, Iraq, Mauritius, Moldova, Mongolia, Mozambique, Nigeria, Romania, Rwanda, Serbia
  - 2018: Afghanistan, Algeria, Argentina, Armenia, Belarus, Croatia, Dominican Republic, Honduras, Jamaica, Kyrgyz Republic, Moldova, Nigeria, Qatar, Republic of North Macedonia, Serbia, Seychelles, Tajikistan, Turkey, Ukraine
- Source note from the table: Cross-country evidence based on IMF ISIMP (2013) and Monetary Operations and Instruments Database – MOID (2018).

### Operational design recommendations for the BCU corridor and operations
- Phase 1: Reset lending and deposit facility rates to a narrow, symmetric band around the TPM that broadly mimics the current effective mini corridor as perceived by the market, but at slightly wider spreads than those implied by typical ISF operations.
  - Choose facility rates far closer to the TPM than the current 1–15 percent bounds, yet still at a spread large enough that banks are not indifferent between using the facilities and trading in the market.
  - Since the initial configuration approximates the corridor banks already face through ISF operations (formalized and marginally higher), immediate funding-structure impacts would be limited.
  - A corridor of at least 50 basis points around the policy rate suggested as a pragmatic starting point.
  - An even narrower corridor risks disincentivizing the already limited interbank trading.
- Replace fine-tuning operations gradually with a functional standard interest rate corridor to improve predictability, transparency, and effectiveness in anchoring short-term market rates.
  - Corridor framework benefits:
    - Predefine and publicly disclose key parameters (overnight maturity, interest rates for liquidity provision and absorption).
    - Reinforce confidence in capacity to offset short-term liquidity shocks by ensuring facility can inject or absorb liquidity at predefined rates, in amounts up to the collateral available to eligible institutions.
- Fine-tuning operations may remain available but use should be progressively reduced:
  - Gradually widen the spread between rates applied in fine-tuning operations so that at a certain point fine-tuning rates become less attractive than corridor rates.
  - When that point is reached, the corridor should assume the role of guiding market rates.
- Over time, widen the corridor to foster more active liquidity management as short-term liquidity forecasts become more reliable and banks adapt to the weekly 7-day FRFA operation at the policy rate.
  - Widening steps should be contingent on observable progress:
    - Stable anchoring of the overnight rate around the TPM under the weekly framework.
    - Improved forecast performance.
    - Evidence from surveys and market data of increasing interbank volumes and use of reserve averaging.
- Communication strategy:
  - Accompany reforms with dedicated communication covering objectives, operational mechanics, timeline, and expected market impact.
  - Support communications with targeted technical sessions with market participants.
  - Referenced companion report: “Uruguay TAR Modernizing Central Bank Communication” (December 2025) for detailed guidance and templates.

### Standing facilities — access, stigma, and recommendations
- Current practice and problem:
  - Regulations impose time and usage limits on the standing lending facility (e.g., capping consecutive use to a limited number of business days), diminishing its role as an “always-available” backstop.
  - When access is known to be capped, banks cannot rely on the facility as a stable safety valve and may engage in alternative funding at rates higher than the lending facility.
  - Survey respondents perceive the standing lending facility as akin to an emergency tool, closer to ELA than to a routine monetary policy instrument, creating stigma.
- Recommendation 7: Ensure that standing facilities are accessible to all eligible counterparties that meet collateral and risk management requirements.
  - Remove time-based caps and similar quantitative restrictions on the use of standing facilities while retaining robust safeguards through eligibility and collateral requirements.
  - Practical rule: any institution that (i) is an eligible monetary policy counterparty and (ii) has sufficient eligible collateral should be able to access deposit and lending facilities on any business day, without constraints on the number of days in a maintenance period.
  - Risk management continues via haircuts, eligibility criteria, and related factors.
  - Streamlining access aligns BCU practice with standard practice in modern inflation-targeting regimes where standing facilities are non-discretionary.
- Recommendation 8: De-stigmatize the standing lending facility by clearly separating it from ELA.
  - Differentiate standing facilities (rule-based, available on demand to eligible counterparties with sufficient collateral) from ELA (exceptional, discretionary, granted at central bank judgment, subject to stricter conditions and separate legal/governance framework).
  - Action: explicitly state in regulations and public communication that the standing lending facility is available on demand to all eligible counterparties meeting collateral and risk-management requirements.
  - Ensure operating framework (rulebook, circulars, website) and internal/external documents consistently distinguish standing facilities from ELA.
  - Peer example: Reserve Bank of Australia notes counterparties may obtain overnight funding at the standing facility rate against eligible collateral.
  - Expected outcome: normalize facility use, encourage banks to use standing facility when appropriate, contain short-term interest rate volatility, and support smooth transmission of the policy rate.

### BCU organizational structure and capacity for monetary operations
- Context and legacy issues:
  - Despite progress toward interest rate targeting, BCU organizational structure retains elements focused on monetary aggregate management.
  - Manifestations:
    - Limited resource allocation to monetary operations and supporting functions, even five years after adoption of interest rate targeting.
    - Fragmented governance for market operations: liquidity forecasting and settlement dispersed across management areas.
    - Operational practices retaining aspects of prior regime, e.g., defining LRM issuance volumes based on projections of monetary aggregates.
  - Need: a more integrated and policy-oriented operational framework aligning instruments, procedures, and governance with liquidity management and interest rate control objectives.
- Governance structure (per BCU Administrative Regulations, December 23, 2024):
  - BCU Board of Directors has three members—the President, Vice President, and Director—appointed by the President of Uruguay.
  - Managerial units reporting directly to the Board include:
    - General Secretariat
    - Economic Policy and Markets Management
    - Superintendency of Financial Services
    - Institutional Services Management
    - Legal Advisory
    - Economic Advisory
    - Internal Audit
    - Planning and Strategic Management
    - Payments System Management
- Recent reorganization and current division of responsibilities:
  - Former Economic Policy and Markets Management reorganized into Monetary Policy Management and Asset and Liability Management.
  - Execution of money market operations assigned to Asset and Liability Management, within the Monetary Operations Division.
  - Monitoring monetary aggregates and liquidity forecasting remain under Monetary Policy Management, within Monetary Analysis.
  - Decision-making on LRM issuance volumes and auction calendar rests with head of Asset and Liability Management, relying on liquidity forecasts from Monetary Policy Management/Monetary Analysis and Monetary Operations Division.
  - Settlement of monetary operations carried out by subunits within Payment System Management.
- Staffing and capacity gaps:
  - Monetary Operations Division currently comprises one manager and three staff members responsible for executing fine-tuning operations and issuing LRMs.
  - Recommendation: significantly strengthen capacity in core and supporting functions to ensure timely and efficient implementation of monetary operations and mitigate operational risks.
- Recommended specialized teams within a dedicated Monetary Operations unit (organize into subareas):
  - Monetary Operations Execution Unit:
    - Conduct all monetary operations, including main monetary operation, issuance of LRM, fine-tuning operations, and managing standard facilities and interest rate corridor.
  - Short-Term Liquidity Forecasting Unit:
    - Monitor and deliver accurate short-term liquidity projections to support timely decisions on monetary policy instruments.
    - Integrate liquidity projections provided by Monetary Analysis into comprehensive reports and actionable insights.
  - Market Analysis and Intelligence Unit:
    - Evaluate domestic asset price dynamics and generate insights for monetary policy decisions and instrument selection.
    - Gather and analyze market intelligence via direct engagement with dealers and market participants.
  - Market Development Unit:
    - Assess initiatives for improving local market functioning, including interbank and LRM markets.
    - Monitor performance of operational tools and the interest rate corridor.
  - Back-Office Operations Unit:
    - Responsible for settlement, reconciliation, and operational risk management related to monetary transactions.
- Strategic alignment:
  - Leverage synergies to incorporate market development initiatives into the division’s mandate, supporting domestic market infrastructure improvements and enhancing interbank and secondary market functioning for LRM.

*Source: IMF Technical Assistance Report (excerpts provided).*

### 134.  In the process of strengthening the monetary operations framework, institutional capacity

### tarea2026037 - 134.  In the process of strengthening the monetary operations framework, institutional capacity

### Institutional capacity and governance
- Findings:
  - Institutional capacity building is essential and includes:
    - comprehensive staff training in market operations and related areas;
    - establishment of robust internal governance mechanisms to support effective decision-making;
    - promotion of transparency through clear disclosure of operational procedures to the Board of Directors.
  - Strong coordination between the monetary policy department and the monetary operations area must be ensured, particularly through the systematic sharing of liquidity projections and analytical insights.
  - These measures will foster operational efficiency, improve policy implementation, and enhance institutional credibility.
- Core mandate:
  - The core mandate of the monetary operations area is to maintain the overnight interest rate close to the target and ensure sound liquidity management.

### Alignment of operational framework with monetary policy
- Findings:
  - Priority: ensure operational framework is fully aligned with the monetary policy objective.
  - Instruments to be calibrated include OMOs, standing facilities, and reserve requirements to maintain short-term interest rates close to the policy target.
  - Operating procedures and liquidity management strategy must be coherent with the interest rate target objective.
  - Instrument selection and sizes should be guided by monetary policy objectives rather than by investment decisions typical of asset and liability management areas.
  - Any limits on the size of monetary instruments should align with policy implementation goals, prioritizing effectiveness and consistency over investment-oriented considerations.

### Organizational placement and naming
- Recommendation / finding:
  - Suggested name: Monetary Operations and Market Development.
  - Rationale: execution of monetary operations by the Central Bank acts as a catalyst for developing the local market—ensuring liquidity, depth, and efficiency—and market development reinforces price discovery, yield curve formation, and transmission mechanism.
- Implementation sequencing:
  - Initial placement: the new Monetary Operations and Market Development unit should initially remain under the Asset and Liability Management area while ensuring governance arrangements remain focused on monetary policy objectives.
  - Revisit placement once the unit achieves operational maturity and demonstrates measurable performance improvements to enable greater integration and capture synergies.

### LRMs, structural surplus, and market-development role
- Key statistics and context:
  - Effective reserve requirement ratio: around 13 percent.
  - Excess reserves: amounted to 431 billion pesos in November 2025 (approximately 12 percent of GDP).
  - From a narrow implementation perspective, the entire surplus could in principle be absorbed through a 7-day operation, but this would require frequent rollovers of very large amounts, concentrating rollover and execution risk at the short end.
- Functional roles:
  - LRMs play a dual role—supporting monetary policy implementation and contributing to money-market development.
  - Central bank responsibility should remain focused on the short segment of the curve; longer-dated benchmarks are left to government issuance (“splitting the yield curve”).
- Recommendation 1:
  - Concentrate LRM issuance in three standardized tenors—3 months, 6 months and 12 months—to deepen liquidity, support secondary-market trading, and strengthen the reference-rate role of LRMs.

### Design and functional distinction of LRM maturities
- Findings:
  - From a liquidity-management perspective:
    - LRMs should absorb most of the structural surplus.
    - The 3-month LRM serves primarily as a monetary policy implementation tool.
    - The 6-month and 12-month LRMs act as benchmarks for market development.
    - The weekly 7-day FRFA operation plays a residual role.
  - Functional differences:
    - 7-day operation: main instrument for keeping the marginal cost of funding close to the BCU’s policy rate; central bank acts as a price maker.
    - 3-month LRM: extends implementation horizon; short-term reference rate closely linked to expected policy/interbank rate; used for internal transfer pricing and loan pricing.
    - 6- and 12-month LRMs: primarily market-development instruments; central bank more price-taker, issuing stable amounts and allowing market demand and risk assessments to set yields.
- LRM instrument features and historical stocks:
  - LRMs are issued as zero-coupon bills—typically at 30, 90, 180, and 360 days (the BCU can issue up to 24 months, though longer tenors have not been used since November 2023).
  - Stock rose from about UYU 333 billion at end-2023 (~10 percent of GDP) to UYU 425 billion in November 2025 (~12 percent of GDP), larger than the monetary base.
  - Composition of outstanding stock: the 30-, 90-, 180-, and 360-day LRMs account for about 7, 27, 23, and 43 percent of the outstanding stock, respectively, with the 720-day line negligible.
  - Market behavior: institutional investors emphasize LRMs as their main outlet for short-term liquidity, with a marked preference for the 6–12-month segment where they tend to buy and hold.
- Empirical pass-through evidence:
  - Short-dated LRMs (30, 90, and 180 days) historically exhibited high and statistically significant pass-through coefficients (typically 0.7–0.9).
  - Interpretation: a 100-basis-point change in the policy rate/overnight interbank market rate is associated with a 70–90-basis-point move in these yields over the short run.
  - The 360-day node exhibits weaker and more time-varying pass-through.

### Auction microstructure and cut-off behavior
- Findings:
  - Bid-to-cover versus premium curves:
    - Short-dated LRMs: curves almost flat and low, indicating cut-off yield remains tightly anchored to the policy/interbank rate; consistent with a price-maker role.
    - Longer tenors: curves become more downward-sloping; higher bid-to-cover ratios associated with smaller premia over the policy rate, indicating greater role for market forces and price-taking behavior.
- Recommendation 2:
  - Standardize LRM auction design and cut-off conventions, and pre-announce fixed issuance amounts to support transparent price discovery.
- Auction rule recommendations:
  - Adopt and communicate a clear, rule-based set of auction procedures, including a pre-announced calendar with fixed LRM issuance amounts.
  - Define auction format, treatment of competitive and non-competitive bids, and the principle that the BCU will behave predominantly as a price taker in LRM auctions while adhering to a pre-announced calendar of fixed-amount auctions.
  - Reinforce adherence to announced issuance volumes; deviations could create market distortions and undermine investor confidence.
- Auction mechanics:
  - Cut-off yield or price: determined by ranking competitive bids from most favorable to least favorable until total amount offered equals auction size; bids at or better than the cut-off are accepted.
  - Allotment rule: adherence to the financial volume offered for each security should be reinforced; issuance volumes are announced in advance and should be strictly observed.

### Maturity concentration, auction frequency, and scheduling
- Recommendation 3:
  - Adopt LRM maturity concentration and reduce auction frequency to enhance market liquidity and improve price formation.
- Rationale and expected effects:
  - Concentrating maturities and lowering auction frequency create predictable points of cash flow and reinvestment, stimulating trading activity and deepening secondary markets.
  - Concentrated maturities generate larger volumes of securities that reach redemption simultaneously, encouraging institutional investors to rebalance and reinvest, increasing demand for new issuances and existing instruments.
  - Benefits include increased turnover, reduced bid-ask spreads, stronger benchmark development, improved valuation accuracy, and facilitation of hedging and peso-denominated collateral use—supporting market dedollarization over time.
- Implementation specifics:
  - As a first step, each LRM issuance should be adapted to allow maturity concentration on a single day of each month (first workday, for example).
  - Current fragmentation:
    - LRMs are scheduled to mature on roughly 100 different business days, with about half of those days seeing two or more individual maturities.
    - Maturities are dispersed across all days of the calendar month, increasing the likelihood that sizeable redemptions on idiosyncratic days generate temporary swings in liquidity conditions and short-term rates.
  - Strategic rationalization:
    - Reduce primary auction frequency and concentrate issuance into fewer, larger auctions to foster deeper secondary markets, improve price formation, increase order book depth, narrow bid-ask spreads, and strengthen price discovery.
    - Fewer auctions reduce administrative burdens and allow market participants to concentrate liquidity around key issuance dates.

*Source: IMF Technical Assistance Report*

### 155. The evidence presented in Error! Reference source not found. underscores the importance of

### tarea2026037 - 155. The evidence presented in Error! Reference source not found. underscores the importance of 

### Auction participation, term premiums, and short- vs long-maturity LRMs
- Cut-off rates for longer-term LRMs generally carry higher risk premiums than shorter-term LRMs.
- An increase in the bid-to-cover ratio can exert significant downward pressure on cut-off rates for longer-term LRMs.
- Under heightened competition, cut-off rates for longer-term LRMs may converge toward levels observed in short-term auctions, which typically reflect lower risk premiums.
- Caution: excessive concentration of maturities can create liquidity clusters, complicating the Central Bank’s management of aggregate liquidity and potentially exerting pressure on overnight interest rates.
- Operational implication: the BCU should be prepared to use its monetary policy instruments to minimize undesirable volatility in liquidity conditions and to smooth short-term rates.

### Coordination with the Debt Management Office — recommended institutional arrangement
- Recommendation 1: Maintain, in the short term, a split-curve arrangement—BCU LRMs up to 12 months and Treasury securities beyond—while strengthening cash- and debt-management coordination with the Ministry of Economy and Finance and assessing, over the medium term, the merits of a transition toward a single-instrument (Treasury-only) model for structural liquidity absorption.
- Rationale for structural operations in large/persistent surplus-liquidity environments:
  - Short-term tools (e.g., 7-day operations and fine-tuning repos) steer overnight/interbank rates and absorb day-to-day forecast errors.
  - When excess reserves are very large, relying solely on short-term operations requires very high outstanding volumes and frequent rollovers.
  - Structural operations (reserve-requirement calibration and/or issuance of securities) park surplus at longer maturities and stabilize regular short-term operations.
- Two implementation options:
  - Split-curve model (BCU issues LRMs up to 12 months; Treasury issues longer maturities):
    - Advantages: (i) strong operational autonomy for the central bank; (ii) ability to provide easily marketable assets tailored to implementation needs; (iii) institutional separation between fiscal funding and monetary operations.
    - Drawbacks: fragmentation of demand, reduced liquidity of each line, complication of constructing a single benchmark curve, and explicit sterilization cost on the central bank’s balance sheet (risk of sustained central bank losses and pressures around recapitalization or profit transfers).
  - Single-instrument (Treasury-securities) model (Treasury issues T-bills/bonds used for funding and liquidity management; proceeds placed in a blocked account at the central bank):
    - Advantages: concentrated liquidity and benchmark status in one instrument, supports development of a unified government-securities market, avoids fragmentation, and is more transparent for public-debt reporting.
    - Requirements/risks: (i) strong legal and operational framework (typically a detailed MoU) to ensure monetary-policy needs are accommodated without undermining central bank independence; (ii) careful communication to markets about issuance motives; (iii) mutual trust and Treasury willingness/capacity to issue additional securities when required.

### Uruguay: sequenced approach and market views
- Mission view: maintain split-curve in short term; keep option to transition toward greater use of Treasury securities for structural absorption over the medium term.
- Near term: continue BCU LRMs up to 12 months combined with strong issuance coordination and clear maturity segmentation to preserve operational control over structural sterilization.
- Medium term: consider gradually increasing role of Treasury bills for long-dated absorption as coordination, cash-flow forecasting, and communication frameworks with MEF are strengthened.
- Transition criteria (three):
  - (i) safeguarding monetary-policy transmission and operational independence;
  - (ii) supporting development of a liquid, well-structured domestic debt market;
  - (iii) ensuring a sustainable and transparent allocation of costs between the BCU and the sovereign.
- Market/survey evidence:
  - Participants indicated bills with maturities beyond three months should progressively migrate to the Ministry of Economy and Finance.
  - Recommendation to codify LRMs limited to 3, 6 and 12 months and rely on the sovereign for longer maturities to align with market perception and usage.

### Governance, MoU, and reporting implications
- Governance recommendation: revisit and strengthen the mandate of the Public Debt Coordination Committee (Comité de Coordinación de Deuda Pública) to operationalize structured coordination between BCU and MEF.
- MoU elements and operational practices to include:
  - Regular meetings (e.g., monthly) to: (i) review short- and medium-term liquidity forecast; (ii) agree auction calendars and indicative issuance ranges by tenor consistent with split-curve arrangement; (iii) plan use of Treasury securities in central bank operations (including add-on issuance for monetary purposes or gradual shift toward Treasury bills for structural absorption); (iv) discuss forthcoming fiscal operations that could materially affect system liquidity.
  - Define mandates, data-sharing arrangements, confidentiality safeguards, dispute-resolution mechanism, principles for cost-sharing, and treatment of blocked accounts.
  - Preserve BCU operational independence for monetary policy implementation and Treasury responsibility for public-debt management.

- Reporting implication of moving LRMs to Treasury securities:
  - Reallocation of sterilization liabilities from BCU to sovereign issuer should be assessed against the relevant debt perimeter used by the authorities in published statistics.
  - Baseline approach: LRMs run off at maturity while Treasury increases issuance in the relevant tenor, with auction proceeds held as government deposits at the BCU in a sterilization account.
  - Greater Treasury issuance for monetary purposes would raise reported gross central-government debt, even if consolidated public-sector liabilities are unchanged once the BCU position is included.
  - The MoU should include a communication approach for reporting the transition.

### Operational sizing and calibration of instruments
- Proposed steady-state split between three instruments:
  1. Required reserves
  2. The outstanding stock of a weekly 7-day main operation
  3. LRMs
- Calibration anchor: September 2025 configuration of the balance sheet as representative for steady-state liquidity conditions.
- Key quantitative calibration inputs and outputs:
  - Effective peso reserve ratio benchmark: 7 percent (covers roughly 95 percent of observed daily deposit outflows; ratios above this mainly insure against rare, idiosyncratic events).
  - Applying 7 percent effective ratio to the peso reserve-requirement base as of September 2025 yields implied structural absorption via required reserves of UYU 29,398 million (the MRR component).
  - Design rule for weekly 7-day operation: a 90th-percentile weekly shock in absolute value should correspond to at most a fraction of the outstanding 7-day operation; with set to 0.3 implies a typical large weekly shock requires an adjustment of at most about one-third of the operation’s volume.
  - Using historical distribution and September 2025 balance-sheet, recommended stock of weekly 7-day OMO absorption: UYU 73,155 million.
  - Residual structural surplus at September 2025 is assigned to LRMs (consistent with LRMs carrying primary burden of structural sterilization).
- Time-series decomposition method:
  - For each month t: compute structural surplus from analytical balance sheet; MRR component = 7 percent of reserve-requirement base in month t; 7-day OMO component = constant UYU 73,155 million; LRM component = residual. Result: reserve requirements provide stable structural drain, weekly operation covers medium-frequency variability, LRMs absorb residual structural surplus.

### Conceptual framework for LRM yield decomposition and empirical method
- LRM yield decomposition: observed yield y_t(n) = 1/h(n) * sum_{k=1}^{h(n)} E_t[r_{t+k}] + Term Premium_t(n).
  - First term: market’s expectations of the average policy rate over LRM life.
  - Term Premium_t(n): risk compensation, liquidity/auction effects, convenience value, and other technical wedges.
- Diagnostic use:
  - If LRM curve moves because expected path rises → markets repricing COPOM’s future stance (monetary policy transmission).
  - If term premium widens with stable expected path → operational story (uncertainty or liquidity frictions).
- Empirical approach:
  - Define spreads s_t(n) ≡ y_t(n) − r_t and state vector x_t ≡ [r_t, s_t(30), s_t(60), s_t(180), s_t(360)].
  - Estimate reduced-form VAR(p) in levels with intercept by least squares; choose p by Schwarz (Bayesian) information criterion.
  - Use expanding-window, real-time VAR re-estimation each month to produce policy-rate forecast path {r̂_{t+1|t}, r̂_{t+2|t}, ...}.
  - Expectations component: discrete average of forecasted path over bill life (explicit formulas for 30, 60, 180, 360 days preserved in the framework).
  - Term premium = observed LRM yield − model-implied expectation (explicit tenor-by-tenor residual formulas preserved).
  - Monthly recursion continues up to sample end minus necessary forecast horizon to avoid look-ahead; method is free of look-ahead bias.

### Future technical assistance (TA) and capacity-building needs
- Follow-up TA and capacity building likely required to operationalize recommendations, including:
  - Leveraging advanced AI tools to enhance productivity and quality.
  - Improving statistical liquidity forecasting.
  - Strengthening monetary policy frameworks.
- Specific capacity-building and TA priorities:
  - Capacitation of BCU staff via seminars and training on central bank operations and best practices.
  - Seek Fund assistance for implementation of advanced quantitative modeling tools on monetary policy implementation (e.g., reserve demand estimation).
  - Leverage Fund expertise in local market development: communication strategies and design of domestic financial instruments to support market deepening and reduce dollarization.
  - Provide support for development of a robust ELA framework to safeguard financial stability and reinforce lender-of-last-resort credibility.
  - Dedicated support mission to build institutional capacity to strengthen the interest rate targeting framework within the inflation-targeting regime and align organizational structure with strategic monetary policy objectives.

*IMF Technical Assistance Report | content unit tarea2026037*

### Annex III. Simulation of Fixed Weekly Operations

### Annex III. Simulation of Fixed Weekly Operations

### Overview
- Purpose: Evaluate how moving to a fixed weekly tender framework—while respecting the existing within-month reserve-averaging mechanism—would affect short-term liquidity management in Uruguay.
- Focus: Implementation efficiency—whether a predictable rule-based framework can deliver similar liquidity control with fewer interventions and more systematic use of averaging provisions.
- Weekly design assumptions common to counterfactuals:
  - a single weekly tender every Tuesday,
  - a fixed seven-day maturity (rolled week to week),
  - the same policy rate as in the baseline.

### Measuring the Liquidity Position
- Baseline liquidity need:
  - Historical stock of outstanding short-term OMOs is reconstructed daily using each operation’s issuance date, maturity date, and amount.
  - For any given day, all active operations are summed to obtain the stock of short-term OMOs outstanding. This stock measures the liquidity absorbed from the banking system through short-term sterilization instruments, in addition to the outstanding LRMs.
- Required reserves and averaging:
  - Daily required reserves are taken from the BCU’s reserve-requirement database and averaged within each maintenance month.
  - Banks are conceptually assumed to manage current-account balances so the average balance equals the required level; short-term OMOs and standing facilities provide an adjustment margin around this averaged path.
- Isolating role of short-term OMOs:
  - Long-term securities (LRMs), government deposits, and other autonomous balance-sheet factors are held at their observed levels so differences arise only from frequency and calibration of short-term OMOs and implied recourse to standing facilities.
- Caveat:
  - The approach abstracts from behavioral and second-round effects; potential real-world predictability gains are intentionally excluded, so estimates are conservative.

### Empirical Strategy and Simulation Scenarios
- Baseline:
  - Reflects actual operations with irregular short-term OMOs and reserve averaging; provides benchmark for liquidity volatility and total intervention volume.
- Counterfactual weekly tender framework:
  - Three calibration rules differ only in weekly tender size:
    - Scenario 1 — Mop-up to zero on the intervention day:
      - Each Tuesday, the weekly tender equals the observed liquidity deviation on that day, bringing the system’s liquidity position to zero at the start of the week.
      - Subsequent within-week shocks are absorbed entirely through standing facilities.
      - Characterized as strictly reactive; within-week variability evident in alternating use of deposit and lending facilities.
    - Scenario 2 — Mop-up of the minimum projected liquidity in the week:
      - Weekly tender calibrated to the minimum liquidity level expected between Tuesday and the following Monday.
      - Ensures the system remains structurally long over the week so they never need to borrow from the standing lending facility; residual liquidity is parked in the deposit facility.
      - Operationally cautious, consistent with an early transition phase.
    - Scenario 3 — Mop-up of the average projected liquidity in the week:
      - Weekly tender set equal to the average liquidity need projected for the week ahead, considering expected autonomous-factor movements and reserve-averaging behavior.
      - Aims to align the weekly operation with typical liquidity conditions so residual daily shocks are small and standing-facility usage is symmetric.
      - Corresponds to a more mature framework using reliable short-term liquidity forecasts.

### Measuring Intervention Volume
- Baseline monthly intervention volume definition:
  - Baseline volume_m = ∑ |ISF_t| over days in m.
  - Captures how much the central bank intervened via fine-tune operations during the month, irrespective of the sign of the flow.
- Scenario monthly intervention volume definition:
  - Scenario volume_m = ∑ |op_k,d| over weeks in m (sum of absolute weekly operation sizes, the simulated 7-day operations).
- Alignment:
  - Both series are aligned to “anchor weeks” running from Tuesday to Monday; volumes summed within anchor weeks then aggregated to calendar months.
- Key metric:
  - Ratio_m(s) = 100 × Scenario volume_m(s) / Baseline volume_m(s), where s ∈ {1,2,3}.
  - Interpretation:
    - Values below 100 percent indicate the scenario requires less intervention volume than the current framework (fewer fingerprints).
    - Values above 100 percent indicate more intervention volume.

### Standing Facilities and Reserve Averaging: Comparing Outcomes
- Mechanism under weekly tenders:
  - Weekly 7-day operation sets starting position each Tuesday; autonomous factors and reserve-averaging behavior move system away from that point and banks use deposit or lending facilities.
- Diagnostic:
  - Compare net standing-facility position (deposits minus lending; black line in Figure) with monthly average required reserves (dashed grey line).
  - Standing-facility amounts computed as if reserve averaging were not used to smooth shocks—day-to-day residuals after weekly operations assumed to go through deposit and lending facilities.
- Simulation findings:
  - Scenario 0 (observed regime): most adjustment carried out through short-term OMOs rather than fully exploiting reserve averaging.
  - Scenarios 1 and 2: net standing-facility line is large and persistent, indicating substantial corridor use would still be required even if reserve averaging were available.
  - Scenario 3: net standing-facility line is relatively modest and close to the reserve-requirement path, with more balanced two-way use of deposits and lending.
- Interpretation:
  - Only Scenario 3 is consistent with the existing reserve-averaging arrangement absorbing most typical shocks, leaving standing facilities as a residual buffer.
  - In Scenarios 1 and 2, the current reserve-averaging parameter would not be sufficient to internalize typical shocks implied by weekly tenders, forcing reliance on the corridor.

### Moving From a Narrow to a Wider Corridor — Effective Costs for Institutions
- Rationale:
  - Fixed weekly main operation paired with a narrow interest-rate corridor preserves effective short-term rate control while removing continuous fine-tuning.
  - Current fine-tuning behaves like a “mini-corridor”: weighted-average rates cluster with max–min spreads around a few tens of basis points (≈15 bps in sample), muting the formal corridor (1–15 percent).
  - Recommended transition: start with a very narrow formal corridor (close to the mini-corridor) and widen gradually as forecasting and interbank trading deepen.
- Cost simulation setup:
  - Weekly 7-day allotment priced at the observed policy rate.
  - Standing facilities remunerated at policy rate plus or minus the corridor radius (ceiling and floor).
  - When surplus, deposit facility pays policy rate minus corridor radius; when deficit, lending facility charges policy rate plus corridor radius.
  - From central bank viewpoint: paying interest on deposits is an expense; charging interest on lending generates income.
- Corridor-width findings:
  - Moving from a narrow corridor (±25 bps) to a wider one (±100 bps) raises the role of standing facilities and increases the cost change from the viewpoint of financial institutions.
  - As the corridor widens, irrespective of scenario, the remuneration gap between facilities and the policy rate increases, raising incentives/costs for institutions.
  - Recommendation: start with a very narrow corridor and amplify as the market develops.
- Cost decomposition reported in simulations:
  - Cumulative implementation costs decomposed into weekly operations (blue), deposit-facility costs (green), and lending-facility income (orange; negative by convention).
  - Grey line shows baseline cumulative cost under current irregular fine-tuning framework.
  - All costs shown from the central bank’s perspective: positive values represent outlays, negative values denote income.

### Key Findings and Policy Recommendations
- Predictability and operational efficiency:
  - A regular weekly calendar reduces need for ad hoc fine-tuning, clarifies timing of central bank actions, and provides banks a stable reference for liquidity planning.
  - Combined with reserve averaging, a weekly framework allows smoothing of short-term shocks with fewer transactions while keeping overnight rates anchored to the policy rate.
- Intervention volume:
  - Overall reduction in volume of ISF operations across the three weekly scenarios, especially Scenario 2 (absorb minimum projected liquidity).
  - Any scenario with Ratio_m(s) persistently below 100 percent achieves similar liquidity control with less intervention volume.
- Preferred calibration:
  - Scenario 3 (mop-up of the average projected liquidity in the week) is most consistent with reserve averaging and minimizes persistent reliance on standing facilities.
  - Scenarios 1 and 2 would leave sizeable, persistent standing-facility positions relative to reserve requirements.
- Corridor management:
  - Begin transition with a narrow corridor (close to current mini-corridor behavior, e.g., ≈15 bps clustering observed) and widen gradually to avoid materially increasing effective liquidity-management costs for institutions.
  - Wider corridors (e.g., ±100 bps versus ±25 bps) increase the role and cost implications of standing facilities for institutions.

*Annex III. Simulation of Fixed Weekly Operations — IMF Technical Assistance Report*

### Annex IV. Liquidity Management

### Annex IV. Liquidity Management

### A. BCU’s Balance Sheet
- High-level structure and size
  - Total Assets: 576100 17.8 %
  - Total Liabilities: 576100 17.8 %
  - Net Foreign Assets: 783883 24.2 %
  - Net Domestic Credit: -207784 -6.4 %
  - Net Claims on Central Government: 126554 3.9 %
  - Net Claims on financial sector: -332253 -10.3 %
  - Net Claims on non-financial sector: 1154 0 %
  - OIN: -3239 -0.1 %
  - Equity: 24879 0.8 %
  - Base Money: 154951 4.8 %
    - Currency in circulation: 109418 3.4 %
    - Banks' reserves: 43544 1.3 %
    - Other deposits: 1989 0.1 %
  - CB Bills: 396270 12.2 %
  - Notes: As of end of October 2025 and in mn UYU and in  percent of GDP.
- Summary observations from the annex
  - The balance sheet is dominated by foreign reserve assets; these comprise the largest share of assets.
  - Currency in circulation corresponds to 3.6 percent of GDP, below its regional peers.
  - Capital strengthened in 2024, lifting equity to around 2 percent of GDP and improving buffers.
  - The balance-sheet expansion was driven by accumulation of international reserves that resulted in a system-wide liquidity surplus.
  - Stable stock of net foreign assets (NFA) is being sterilized through issuance of Letras de Regulación Monetaria (LRM).
  - Required reserves provide a stable drain under the monthly-averaging rule; the deposit facility and 30-day OMOs (ISF) are used mainly for fine-tuning.
  - Because the BCU has refrained from FX interventions since 2021 and NFA is shown at constant FX with a three-month average, month-to-month NFA movements are small; short-term oscillations mainly arise from seasonality in currency circulation and fluctuations in government deposits.
  - The increase in outstanding LRMs during 2024–25 reflects ongoing need to absorb a persistent liquidity surplus.
  - BCU is prioritizing the use of 90-days, 180-days and 360-days LRMs.
    - Medium-term LRMs reduce rollover risk and support banks’ liquid-asset holdings, but maintaining large stocks can constrain rapid response to significant liquidity shocks and can affect demand for government securities.
  - ISF volumes are small and episodic relative to LRMs; LRMs carry the weight of sterilization and play the primary role in interest rate transmission across tenors.
    - LRMs are the main instrument for absorbing structural liquidity surpluses and serve as the primary reference for the short end of the yield curve.
    - ISF operations function at the margin to smooth day-to-day imbalances and keep the interest rate close to the policy rate.

### B. Monthly forecasts
- Toolkit design and horizons
  - The toolkit produces:
    - Short-term daily forecasts over a 7–30 day horizon for calibrating regular and fine-tuning operations.
    - Multi-month forecasts to gauge likely evolution of structural liquidity and reserve demand.
  - The same model suite is used across horizons; parameters for model selection, reconciliation, and interval construction are calibrated to operational needs at each horizon.
  - Forecast vintages are stored with production date and target horizon to enable ex-post evaluation and to distinguish statistical outputs from judgmental overlays.
  - Underlying R scripts and examples of rolling-origin back-testing are described in an accompanying handout.
- Role of multi-month forecasts
  - For Currency in Circulation (CiC):
    - CiC exhibits a stable upward trend and well-behaved seasonal pattern.
    - Multi-month forecast provides a baseline trajectory for the expansion of non-interest-bearing liabilities.
    - Baseline can be combined with assumptions on structural components of reserve demand (e.g., average precautionary buffer held by banks) to infer path of structural liquidity.
    - Monthly CiC projections from the toolkit produce a monthly average; the BCU’s existing day-of-month profile is then applied to derive an internally consistent daily path.
  - For Government Deposits (GovD):
    - Large and irregular swings and absence of a clear structural trend limit usefulness of purely statistical multi-month projections.
    - Medium-term GovD forecasts are used more as scenario inputs built around the fiscal program, known policy measures, and planned debt-management operations than as stand-alone projections.
    - GovD forecasts are updated when new budget information, financing plans, or cash-management strategies become available.
- Illustration
  - Figure 2 (CiC Monthly Forecast): blue line indicates monthly median CiC evolution over time; orange line indicates the IMF’s forecast. Data source: BCU and IMF calculations.

### C. Forecast Models and Evaluation
- Institutional and publication recommendations
  - Recommendation Annex 1: Publish liquidity forecasts after internal validation, including measures of forecast uncertainty, once internal performance is robust to increase market transparency and support more active liquidity management by banks.
    - Publish short-term liquidity forecasts only after seven-day FRFA has been adopted as the main operation and the new forecasting framework has been thoroughly tested and demonstrated robust across several maintenance periods.
    - Publication should include explicit measures of forecast uncertainty and an operational link to the planned weekly seven-day fixed rate full allotment (FRFA) operation at the policy rate.
    - Central publication with prediction intervals would allow banks to benchmark, assess likely surplus/deficit over maintenance periods, and incorporate forecast error probability into bidding strategies.
    - Current practice: forecasts used only internally, no systematic publication of point projections, forecast bands, or performance indicators; interim updates overwrite vintages making ex post error evaluation impossible.
    - IMF toolkit produces standard deliverables (tables, fan charts, simple coverage diagnostics, exception logs) that could form basis for regular publication once validated.
  - Recommendation Annex 2: Develop a dedicated, structured database integrating liquidity data, balance sheet information, and forecast vintages to replace multiple Excel files and enhance analytical capabilities.
    - Rationale: current forecasting process relies on numerous Excel files, tacit knowledge, manual steps, partial documentation, and no version control; operational risk is high and procedures are labor-intensive and not easily transferable.
    - Benefits: strengthen data governance, traceable changes to inputs/outputs, cleaner data pipelines, ex-post evaluation, more robust to staff turnover.
- Technical implementation and governance
  - The proposed forecasting framework is operationalized as a comprehensive R package:
    - Provides a robust data pipeline transforming inputs into standardized outputs with documented models and intermediary steps.
    - Separates statistical forecast from judgmental adjustments.
    - Allows comparison of forecast accuracy across models and over time.
    - Uses free, open-source software; BCU already has proficient R programmers and internal usage to leverage.
    - A separate technical handout contains detailed model specifications, validation protocols, and fully documented R code.
  - Institutional setup:
    - Liquidity-forecasting unit: produce and maintain forecast vintages for autonomous factors and reserve demand.
    - Monetary-operations area: use forecasts to calibrate size and timing of open market operations.
    - Back office and accounting units: provide timely, high-quality source data.
    - Coordination with the Treasury and key stakeholders follows a predictable calendar with agreed deadlines for transmitting cash-flow information and reconciling outstanding balances.
    - Internal procedures should specify sequence from data extraction to forecast production, documentation of judgmental adjustments, and escalation rules for large forecast errors or anomalies.
- Statistical approach and model selection
  - Forecasts organized around a small set of well-specified models per liquidity factor and horizon.
  - Candidate models typically include:
    - Naïve and seasonal-naïve benchmarks
    - Exponential-smoothing (ETS) variants
    - ARIMA/SARIMA models with optional regressors for calendar effects, tax dates, salary payments, and other events
  - Forecast generation and evaluation:
    - Fixed horizons (e.g., 7 and 30 calendar days operationally; several months for planning)
    - Out-of-sample evaluation using rolling-origin procedure
    - Performance assessed using bias (Mean Error), accuracy metrics (MAE, RMSE), and interval scores and coverage rates for predictive distributions
    - Model selection and combination based on multi-horizon statistics rather than in-sample fit
    - Hierarchical or reconciled forecasts used where multiple frequencies are relevant so daily paths aggregate exactly to target monthly averages
  - Factor-specific predictability and modeling guidance:
    - CiC: stable trend, persistent seasonality; parsimonious time-series models with calendar regressors and occasional intervention dummies typically sufficient for high accuracy at daily and multi-month horizons.
    - GovD: larger idiosyncratic swings; main gains from combining statistical models with structured institutional information (issuance calendars, one-off operations, Treasury cash-flow projections) rather than complex statistical specifications.

*Source: Annex IV. Liquidity Management (tarea2026037) — IMF technical assistance report.*

### Annex V. Current Instruments

### Annex V. Current Instruments

### Overview
- Describes the main instruments used to implement the BCU’s interest-rate-based operational framework.
- Outlines: the structure of reserve requirements and averaging mechanism; the interest-rate corridor and standing facilities; fine-tuning operations (ISF); and the role of LRM as principal sterilization and benchmark instrument in pesos.
- Notes that the inflation-targeting framework entered into force in 2005, and that since September 2020 the BCU has operated an interest-rate-based framework with the overnight interbank peso rate (TMM) aligned to the policy rate Tasa de Politica Monetaria (TPM) set by the Monetary Policy Committee (COPOM).
- COPOM composition and cadence:
  - Composed of three members of the Board of Directors of the BCU and three senior officials appointed by the Board, with input from the Ministry of Finance via the Macroeconomic Coordination Committee.
  - Meets roughly every 6-8 weeks, with about 8 scheduled meetings per year.

### Reserve Requirements — Structure and intent
- Reserve requirements absorb excess liquidity and reinforce financial stability, incentivizing longer-term peso funding and discouraging reliance on short-term FX liabilities.
- High FX reserve requirement ratios support dedollarization and reflect dollarization risk; described as "among the highest in the region."
- Reserve requirements vary by currency and original maturity, serving liquidity management and prudential objectives.
- Peso (UYU) and UI reserve requirement ratios by deposit category:
  - Demand & <30-day time: 15 %
  - 30–90 days: 3 %
  - 91–180 days: 2 %
  - 181–366 days: 1 %
  - >1 year: 0 %
- Foreign currency (FX, residents) reserve requirement ratios:
  - ≤180 days: 28 %
  - >180 days: 20 %
- FX, non-residents (net of eligible FX assets): 28 %
- Design effects:
  - Withdraws structural liquidity and creates buffers against deposit volatility.
  - Declining peso schedule encourages longer-term local-currency funding and supports open market operations in guiding short-term rates toward long-term rates.

### Reserve Requirement Remuneration
- Peso reserves (UYU): remunerated at 60% × TPM on the remunerable base (net of average vault cash for the month).
- US dollars (FX): remunerated at Fed funds midpoint − 0.50%; applies to reserve requirements for USD-denominated liabilities.
- UI (indexed pesos): no interest; receive UI indexation for the maintenance month (adjustment equals the UI change over the maintenance month).
- BCU circulars and communications set these parameters.

### Effective Reserve Ratios, Deposit Composition, and Compliance
- Effective reserve requirement ratio history and magnitudes:
  - Effective reserve requirement ratio remained steady between 12.5 percent and 13.0 percent since January 2022.
  - Prior to 2022, the effective ratio was close to 20 percent.
- Deposit composition and maturity effects:
  - Deposits with maturity below 30 days make up over 80 percent of the entire deposit base subject to MRR.
  - Share of deposits with maturity above 30 days increased from 11 percent to 18 percent after a large cut of the reserve requirement ratio applied to longer-term deposits.
- Aggregate balances:
  - As of the end of 2024, banks’ balances at the BCU combining USD and UYU totaled approximately 10–11 percent of GDP.
- Averaging and intramonth flexibility:
  - Compliance calculated on a monthly average (“período de cómputo”), allowing within-month day-to-day flexibility; cash in vault in pesos counts toward the peso reserve requirement base.
  - The averaging framework reduces money-market volatility and keeps the reserve requirement structural.
- Fulfilment behavior:
  - Using data from November 2024 onward, the vast majority of bank-month observations lie below 120 percent of the required level; for most banks fulfilment ratios are consistently below 110 percent.
  - Typical excess reserves are in the order of only 0–10 percent above the minimum requirement.
  - Intramonth paths (Nov 2024–Oct 2025) show many banks allow cumulative position deviations more than ±30 percent from the eventual requirement at various points; many run into cumulative deficits after around day 25.
  - The BCU supplies a modest liquidity buffer early in the period and gradually withdraws it later in the maintenance period.
- Non-compliance penalties:
  - Financial charge on shortfall calculated at a punitive rate (example: 1.5× the transaction price for pesos or Fed + 2 percent for FX), possible reputational or supervisory consequences.
  - BCU monitors compliance monthly and does not allow carryover of excess or deficit into subsequent periods.

### Interest-Rate Corridor and Standing Facilities
- Formal corridor:
  - Standing lending and deposit facilities define a broad 1–15 percent corridor.
  - Example standing rates as of November 10, 2025: overnight deposit window at 1 percent and overnight credit window at 15 percent.
- Practical operation:
  - The formal corridor is wide and rarely binds; standing facilities function as backstops and are largely dormant under normal conditions.
  - Interbank rates trade within a tight ±15-basis-point band around the target, limiting incentives to arbitrage toward standing-facility rates.
  - Frequent ISF fine-tuning operations create a de facto narrow trading band (a "mini-corridor") around TPM.

### ISF Fine-Tuning Operations — Cadence and effects
- ISF activity metrics (January 2020 to mid-2025):
  - Total ISF interventions: 2,380 across 1,019 distinct days.
  - Average gap between consecutive ISF days: 1.9 days (median = 1).
  - 66 percent of gaps equal to one day.
  - 90th percentile gap: 4 days (maximum of 14 days).
  - Streaks of consecutive ISF days: 246 streaks of length ≥ 2; median streak length = 4; P90 = 5.
  - ISF days are spread evenly across weekdays (approximately 20 percent each).
  - Monthly activity averages 38 interventions (in 16 ISF days).
- Operational characteristics:
  - High-frequency, irregular ISF operations produce a near-daily market presence despite uneven spacing and intensity.
  - The mini-corridor generated by daily ISF operations keeps the overnight rate within a few basis points of the policy rate—far tighter than the formal 1–15 percent standing facilities corridor.
  - The effective operating band around TPM is delivered through ISF operations while standing facility windows remain as last-resort backstops.

*Annex V. Current Instruments — tarea2026037.*

### 7.3 bp (median 5 bp; P90 20 bp), narrowing over time from 16.3 bp in 2020 to 2.9–3.3 bp in 2024–2025.

### tarea2026037 - 7.3 bp (median 5 bp; P90 20 bp), narrowing over time from 16.3 bp in 2020 to 2.9–3.3 bp in 2024–2025.

### Mini-corridor and short-end market dynamics
- Key statistics and pattern:
  - Observed TMM–TPM spread: 7.3 bp (median 5 bp; P90 20 bp), narrowing from 16.3 bp in 2020 to 2.9–3.3 bp in 2024–2025.
  - Formal corridor delineated by standing facilities spans approximately 1,400 bp.
- Operational outcome:
  - The TMM tracks the TPM inside a narrow band while the formal bounds remain largely inactive.
  - This creates a highly predictable close that limits incentives for active bilateral liquidity management and price discovery at the very short end.
  - Market participants increasingly condition intraday positioning on expected central bank presence rather than interbank trading.
- Survey evidence:
  - Peso liquidity is “usually guaranteed” when the BCU is active; money-market operations commonly seen at (or very near) the policy rate and tenors of 2–3 days.
  - Volumes are not systematically pre-signaled, complicating day-to-day positioning and consistent with an irregular intervention cadence.

### ISF operations, reserve averaging, and banks’ intramonth behavior
- ISF frequency and incentives:
  - High frequency of ISF operations, even when banks’ reserve balances fluctuate around the required reserve average, suggests ISF spreads may not sufficiently incentivize use of reserve requirement averaging.
  - Current ISF pricing structure appears to encourage frequent reliance on central bank operations for day-to-day liquidity management rather than smoothing via reserve averaging.
- Empirical distribution and interpretation:
  - Most observations concentrated in quadrants 1 and 3 and along the axes, indicating co-movement or closeness to monthly averages.
  - Nearly 24 percent of data points are in quadrants 2 and 4, indicating cases where banks neither engage in interbank transactions to manage liquidity nor make more effective use of reserve averaging.
  - The notable share of observations in Q2 and Q4 suggests the current ISF pricing mechanism may require revision to better support intended policy objectives.
- Timing of fine-tuning operations:
  - No simple one-to-one mapping between individual reserve “gaps” and the direction of OMOs.
  - Liquidity provision to banks formally above same-day requirement is rare and concentrated early in the maintenance period.
  - Liquidity provision to banks appearing below same-day requirement is more frequent and distributed throughout the month.
  - Liquidity absorption dominates, especially in the last week of the period and particularly at banks that, on a same-day basis, sit below the contemporaneous requirement.
  - Many banks front-load reserve holdings to secure compliance early, then run down cumulative surpluses; late-month same-day deficits do not necessarily imply expected non-compliance with the monthly average.

### Standing facilities and intraday arrangements
- Access and eligibility:
  - Access limited to institutions that perform financial intermediation (IIF) with current accounts at the BCU—principally licensed banks and other supervised financial intermediaries (e.g., casas financieras and cooperativas de intermediación financiera).
  - Other entities may hold BCU accounts (AFAPs, insurers, payment institutions, public bodies, exchanges, CCPs, etc.) but are not eligible for the standing credit window.
- Facility design and terms:
  - Two distinct standing lending windows provide secured overnight credit at a penalty rate; the standing deposit window allows unlimited overnight placements at the posted floor rate.
  - Facilidad Permanente de Liquidez (FPL): requested during operating hours as an overnight repo against eligible public securities.
  - Facilidad de Crédito a tasa de penalización: opens after market close to resolve end-of-day shortfalls or regularize unpaid intraday credit.
  - Shared terms for lending facilities: pricing at a published penalty rate; collateral in eligible public securities; usage capped at 30 percent of the institution’s average minimum reserve requirement from the second prior month; access on up to three business days per month; principal and interest due the next business day; BCU may sell or retain pledged collateral if unpaid.
  - Intraday liquidity facility: same-day, non-remunerated credit against pre-registered collateral; unpaid balances at end of day automatically regularize as an overnight repo at the standing penal rate.
- Operational implications:
  - Wide-corridor/active-operations configuration keeps the overnight rate anchored but increases day-to-day operational demands on the central bank.
  - Because there is no remunerated floor close to the policy rate, the BCU must actively absorb residual surpluses and cover temporary shortfalls via OMOs and ISF near the policy rate.
  - Survey feedback: institutions plan around OMOs/ISF and treat standing windows as backstops; some note a corridor closer to market conditions would make the system more self-correcting.

### Letras de Regulación Monetaria (LRMs): role, issuance, and market structure
- Instrument design and role:
  - LRMs are peso-denominated, zero-coupon securities issued by the BCU as its main structural sterilization instrument.
  - Typical maturities: 30, 90, 180 and 360 days; legal framework allows up to 24 months (maturities beyond one year not used in recent periods).
  - Placed at a discount and redeemed at par; treated by market participants as free of credit risk.
  - LRM yields closely linked to the monetary policy rate and expectations of its future path; LRM cut-offs and secondary-market curve form the main reference for short-term risk-free peso rates.
- Stock and composition:
  - Stock rose from about UYU 333 billion at end-2023 (~10 percent of GDP) to roughly UYU 425 billion in November 2025 (~12 percent of GDP), larger than the monetary base.
  - The 30-, 90-, 180-, and 360-day LRMs account for 7 percent, 27 percent, 23 percent, and 43 percent of the total outstanding stock, respectively.
- Market scale and concentration (July 2024–June 2025):
  - Total securities operations: about UYU 3.3 trillion, equivalent to roughly 99 percent of GDP.
  - Primary issuance: LRMs account for about 45 percent; private-sector certificates of deposit (CDs) represent over 50 percent; Treasury notes and other securities together account for less than 5 percent.
  - Secondary trading: LRMs represent close to 90 percent of all secondary trading in domestic securities; international issues 7 percent; Treasury notes 3.9 percent; other securities 0.6 percent.
  - Private CDs secondary operations amount to only about 0.1 percent of total private-sector volumes (around UYU 1.7 billion out of roughly UYU 1.3 trillion), implying a secondary-to-primary ratio close to zero.
- Auction mechanics and market practices:
  - Auctions: uniform-price (Dutch) mechanism; participants submit sealed bids specifying amount and yield; BCU accepts bids up to the cut-off volume and all successful bidders awarded at the single cut-off yield.
  - BCU discretion: may allot less than the announced amount if rates offered are deemed too high and can allocate more than the tender amount when needed.
  - Two-tier structure: competitive tranche by designated primary dealers (six primary dealers participate in the competitive tranche) and a non-competitive tranche roughly 20 percent of the competitive amount open to broader set of banks and institutional investors at the weighted-average auction yield.
  - Auction calendar and settlement: regular weekly calendar—typically with 1-month bills early in the week, 3-month bills mid-week, and alternating 6- and 12-month bills later—with T+0 settlement for short tenors and T+1 for 12-month bills.
  - Post-auction transparency: results (amounts offered and awarded, stop-out, bid-to-cover) released immediately.
- Price discovery features and market depth:
  - Auction outcomes keep short-end LRM yields tightly aligned with the policy rate, anchoring a benchmark curve used across the system; market depth thins beyond 6–12 months.
  - 30–180-day cut-offs cluster near the TPM; rollovers are orderly; issuance size/frequency broadly adequate.
  - Longer-tenor pricing often anticipates policy changes; longer tenors can drift ahead of turning points.
  - Sensitivity to bid-to-cover: 30–90d curves flat and low across bins (anchored short end), while 180–720d curves slope down (more market-driven price discovery at longer tenors).
  - Market participants use LRM yields and the policy rate as primary references for internal funds-transfer pricing and pricing wholesale deposits and loans in pesos.
- Market participants’ concerns and suggestions:
  - Some participants view BCU’s discretionary adjustments as lacking transparency.
  - Respondents ask the BCU to pre-announce cut-off conventions (including tie-breaks) and apply them consistently; variability observed between “cutting by rate” and “cutting by volume” and changes in cut-off conventions during 2024–25, especially in certain 90-day auctions.
  - Some propose moving from a Dutch-style to an English-style format to reinforce price discovery at the cut-off.
- Auction participation timing pattern:
  - Median adjudicated volume and bid-to-cover ratios at competitive auctions for 30d and 90d show a gradual decline from W1 to W5, indicating reduced demand later in the month.
  - 180d tenor: similar downward trend with a sharper drop in bid-to-cover ratios.
  - 360d tenor: more variability, peak in W2 then decline; dispersion (P25–P75) narrow for short tenors and wider for 360d.

### Implications and policy-relevant issues highlighted
- The narrow effective mini-corridor at the very short end reduces incentives for bilateral interbank trading and active liquidity management, increasing reliance on central bank operations.
- ISF pricing may not sufficiently encourage reserve-averaging behavior; revision of the ISF pricing mechanism is suggested to better support intended reserve framework objectives.
- Operational demands on the BCU are elevated under a wide formal corridor with active OMOs/ISF near the policy rate; a corridor closer to market conditions could make the system more self-correcting.
- Auction transparency and pre-announced cut-off conventions are important to avoid unexpected shifts in market rates; some market participants request pre-announced rules and consistency, and consider a move to an English-style auction to reinforce price discovery.
- LRMs concentrate issuance and trading, defining the short-end benchmark curve, but market depth thins beyond 6–12 months, constraining price discovery and transmission further out the curve.

*Source: tarea2026037 - 7.3 bp (median 5 bp; P90 20 bp), narrowing over time from 16.3 bp in 2020 to 2.9–3.3 bp in 2024–2025.*

### Annex VI. Money Market Overview

### Annex VI. Money Market Overview

### Central bank operations and rate management
- The BCU conducts daily fine-tuning operations to keep the overnight interbank rate close to the TPM.
- Fine-tuning instruments are remunerated within a narrow corridor around the policy rate (±20 basis points), which effectively pins the overnight rate close to the TPM.
- The central bank routinely absorbs or injects liquidity to accommodate short-term shocks to banks’ reserve positions.

### Market structure and transaction patterns
- Short-term operations are concentrated in the peso “call garantizado” segment.
- Most volumes in that segment are transacted at maturities between two days and one week.
- Across these short maturities, the BCU accounts for an estimated 50–70 percent of turnover, highest in the 3–6 day bucket; the remainder reflects bilateral trades between banks.

### Bank funding, peso CDs, and interaction with BCU operations
- Peso CDs are overwhelmingly issued at very short maturities, mostly overnight or within a few days, and are sold to pension funds.
- These nominal peso CDs are well suited to managing immediate funding needs but do not materially lengthen the maturity profile of peso liabilities or provide a structural buffer for meeting reserve requirements.
- Bank-level data indicate that in many months roughly 30 percent of instances in which banks place surplus liquidity with the BCU through fine-tuning operations coincide with new peso CD issuance by the same institutions on the same day.
  - On those days banks earn a spread over the policy related rate to raise very short term pesos from investors while simultaneously depositing cash at the central bank, rather than allowing surplus and deficit positions to be matched through the interbank market.
  - This pattern suggests nominal peso CDs are used primarily as a tactical wholesale funding instrument and that the BCU intermediates a significant share of short term peso liquidity coming from institutional investors.
- A major long-term funding source for banks lies outside this nominal-CD pattern: inflation indexed UI CDs placed mainly with pension funds, often with maturities around five years and in some cases close to ten years, which play a central role in providing stable term funding and supporting the indexed yield curve.

### Implications for market development and liquidity management
- The BCU’s strong footprint supports a tight alignment between the target rate and market conditions but survey evidence suggests it reduces incentives for banks to manage liquidity actively or trade reserves among themselves.
- Excess liquidity is usually placed in fine-tuning deposits rather than redistributed via the interbank market, which may limit the development of deeper unsecured and secured money markets over time.
- The coexistence of private wholesale funding and central bank intermediation in the very short segment implies limited interbank matching of surplus and deficit positions, reinforcing tactical use of short nominal CDs.

*Source: Annex VI. Money Market Overview, tarea2026037.*

### Annex X. Comparative Table of Central Bank Liquidity

### Annex X. Comparative Table of Central Bank Liquidity

### Forecasting & Operational Disclosure (per central bank)
- Banco Central del Uruguay (Uruguay)
  - Publishes Forecasts?: No public bulletin (internal daily planning)
  - Forecast Horizon: Short-term (daily/few days)
  - Forecast Frequency: Daily (internal)
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Limited formal mechanisms (Treasury cash still matters)
  - Main Operations & Framework: Sterilization mostly via LRM bills and daily OMO toolset; LRM is the workhorse for structural absorption
  - Notable Practices / Innovations: BCU “Instrumentos emitidos”; “Instrumentos del BCU y GC”
- Banco Central de la República Argentina (Argentina)
  - Publishes Forecasts?: No public liquidity forecast
  - Forecast Horizon: Immediate (day-to-day)
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Informal/operational coordination
  - Main Operations & Framework: LELIQ auctions; repos/FX swaps as needed
  - Notable Practices / Innovations: Framework changes have been frequent
- Banco Central do Brasil (Brazil)
  - Publishes Forecasts?: No public liquidity forecast (transparency via reports/calendars)
  - Forecast Horizon: Up to weeks/months (internal)
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Strong info-sharing with Treasury
  - Main Operations & Framework: Repos, reverse-repos; standing facilities; Selic system
  - Notable Practices / Innovations: Clear corridor implementation; robust comms/reporting
- Banco de México (Mexico)
  - Publishes Forecasts?: Partial — policy/operations info; no explicit daily forecast bulletin confirmed
  - Forecast Horizon: Short-term (daily operations)
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Legal info-sharing framework
  - Main Operations & Framework: Repos, deposits, FX swaps; standing facilities
  - Notable Practices / Innovations: Transparent announcements & releases
- Banco Central de Chile (Chile)
  - Publishes Forecasts?: No standalone public liquidity forecast
  - Forecast Horizon: Maintenance-period focus
  - Forecast Frequency: —
  - Forecast Errors Published?: —
  - Coordination with Fiscal Authority: Coordinated calendars with Treasury
  - Main Operations & Framework: Fixed-rate/full-allotment-style MROs; BCCh bills; repos
  - Notable Practices / Innovations: Detailed legal/operational circulars publicly posted
- Banco de la República (Colombia)
  - Publishes Forecasts?: No dedicated forecast bulletin
  - Forecast Horizon: 1–2 weeks (main ops)
  - Forecast Frequency: Weekly main; daily monitoring
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Treasury flows incorporated
  - Main Operations & Framework: Full-allotment repos; CB bills
  - Notable Practices / Innovations: Clear operations rulebook
- Česká národní banka (Czech Republic)
  - Publishes Forecasts?: Yes — public daily liquidity page
  - Forecast Horizon: Daily (forecast & outturn)
  - Forecast Frequency: Daily
  - Forecast Errors Published?: Yes — discloses forecast error
  - Coordination with Fiscal Authority: MoF flows included
  - Main Operations & Framework: Fixed-rate, full-allotment (American auction) absorption
  - Notable Practices / Innovations: Unusual transparency: daily forecast + error
- Danmarks Nationalbank (Denmark)
  - Publishes Forecasts?: Yes — publishes cash-flow effects/ops info
  - Forecast Horizon: Short-term structural (weeks)
  - Forecast Frequency: Weekly ops
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Strong — gov’t deposits central to liquidity
  - Main Operations & Framework: Fixed-rate, full-allotment weekly operations; unlimited current accounts
  - Notable Practices / Innovations: Classic peg implementation with transparent ops
- European Central Bank (Eurozone)
  - Publishes Forecasts?: Yes — weekly liquidity analysis
  - Forecast Horizon: Maintenance period (~6 weeks)
  - Forecast Frequency: Weekly
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: NCBs share gov’t cash-flow data
  - Main Operations & Framework: Fixed-rate, full-allotment; reserve averaging; fine-tuning
  - Notable Practices / Innovations: Publishes daily liquidity conditions & weekly analysis
- Federal Reserve (USA)
  - Publishes Forecasts?: No public liquidity forecast (ample-reserves regime)
  - Forecast Horizon: Up to ~1y (internal scenarios)
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Yes — info-sharing on Treasury TGA
  - Main Operations & Framework: Ample-reserves floor system; minimal active OMOs; SRF backstop
  - Notable Practices / Innovations: Implementation notes after each FOMC set tools
- Bank of England (UK)
  - Publishes Forecasts?: No public liquidity forecast
  - Forecast Horizon: ~10-day horizon (internal)
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Coordinated with UK DMO
  - Main Operations & Framework: Sterling Monetary Framework; weekly FRFA repo
  - Notable Practices / Innovations: Red Book parameters & facilities
- Bank of Canada (Canada)
  - Publishes Forecasts?: Yes — posts daily settlement balance target
  - Forecast Horizon: Next day
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Strong integration with gov’t cash
  - Main Operations & Framework: Corridor (or floor when needed); morning OMOs
  - Notable Practices / Innovations: Transparent daily targets
- Monetary Authority of Singapore (Singapore)
  - Publishes Forecasts?: No public “forecast” doc (FX-targeting; publishes OMO schedules/practices)
  - Forecast Horizon: Daily assessment; semiannual FX stance
  - Forecast Frequency: Daily
  - Forecast Errors Published?: —
  - Coordination with Fiscal Authority: Close MoF links; gov’t balances key
  - Main Operations & Framework: Daily OMOs, repos/swaps, MAS Bills; standing facility
  - Notable Practices / Innovations: Clear documentation of daily 9:45 sessions
- Bank of Japan (Japan)
  - Publishes Forecasts?: Partial — publishes daily operations summaries (not forward-looking)
  - Forecast Horizon: Next-day reserves; monthly averages
  - Forecast Frequency: Daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Tight MoF coordination
  - Main Operations & Framework: Fixed-rate, full-allotment funds-supplying; JGB purchase ops
  - Notable Practices / Innovations: Standing facilities cap volatility under ample reserves
- Reserve Bank of Australia (Australia)
  - Publishes Forecasts?: No public liquidity forecast (framework & OMO details published)
  - Forecast Horizon: Weekly ops cycle
  - Forecast Frequency: Weekly
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Strong (Cash Coordination Committee)
  - Main Operations & Framework: 2025 ample-reserves framework; FRFA repos; unlimited ES balances
  - Notable Practices / Innovations: Clear path from corridor → floor
- Reserve Bank of New Zealand (New Zealand)
  - Publishes Forecasts?: No public daily forecast
  - Forecast Horizon: Maintenance period & medium-term
  - Forecast Frequency: —
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: With NZ Debt Management
  - Main Operations & Framework: Full-allotment floor; unlimited Settlement Cash balances
  - Notable Practices / Innovations: Detailed handbook & implementation notes
- Sveriges Riksbank (Sweden)
  - Publishes Forecasts?: No public liquidity forecast
  - Forecast Horizon: Weekly cycles
  - Forecast Frequency: Weekly
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: With Debt Office
  - Main Operations & Framework: Certificates (7-day); ±10bp corridor; no reserve requirements
  - Notable Practices / Innovations: Publishes policy-rate path (separate from liquidity)
- Central Bank of Nigeria (Nigeria)
  - Publishes Forecasts?: No public bulletin (internal)
  - Forecast Horizon: Short-term (daily/few days)
  - Forecast Frequency: Daily (internal updates)
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Yes — TSA flows and other autonomous factors incorporated
  - Main Operations & Framework: Regular OMOs (CBN bills) to absorb; reverse repos to inject; standing facilities with wide corridor; high CRR used actively
  - Notable Practices / Innovations: Notable use of CRR as quasi-OMO; sterilization to offset FX interventions
- Central Bank of Kenya (Kenya)
  - Publishes Forecasts?: No public bulletin (internal)
  - Forecast Horizon: Short-term (daily monitoring; weekly planning)
  - Forecast Frequency: Daily (internal); weekly review
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Yes — Treasury Single Account flows shared
  - Main Operations & Framework: OMOs via 7-day Repos/Reverse Repos; Term Auction Deposits; standing facilities; no reserve averaging
  - Notable Practices / Innovations: 2023 corridor reform; DhowCSD collateralized interbank platform; frequent fine-tuning
- Bank of Ghana (Ghana)
  - Publishes Forecasts?: No public bulletin (internal)
  - Forecast Horizon: Short-term (weekly–monthly horizon)
  - Forecast Frequency: Daily monitoring; weekly main ops
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Informal/operational coordination with MoF
  - Main Operations & Framework: Two-tier corridor: normal ±100bp; late window ±500bp; weekly BoG bills (14d/56d); repos/term deposits; unremunerated RR
  - Notable Practices / Innovations: Late-window penalty range incentivizes early squaring; active sterilization amid high inflation
- Central Bank of Egypt (Egypt)
  - Publishes Forecasts?: No explicit public forecast (announces volumes)
  - Forecast Horizon: Short-term (1-week main operation; longer tenors as needed)
  - Forecast Frequency: Weekly (main OMO operation; Tuesday)
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Strong — TSA flows & issuance calendars coordinated
  - Main Operations & Framework: Corridor system (≈100bp); weekly fixed-rate, full-allotment deposit auctions; longer-tenor corridor-linked deposits; reverse repos if in deficit; standing facilities; RR adjustments
  - Notable Practices / Innovations: Large-scale absorption with full-allotment tenders; variable-rate deposit tenders for longer maturities
- Bank Al-Maghrib (Morocco)
  - Publishes Forecasts?: No public bulletin (weekly conditions reported)
  - Forecast Horizon: Short-term (weekly main ops; fine-tuning <7 days)
  - Forecast Frequency: Weekly main auction (Tuesday) + as needed
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: High — TSA & issuance calendars integrated
  - Main Operations & Framework: Weekly fixed-rate repos at policy rate; occasional fine-tuning; longer-term repos/FX swaps; standing facilities; unremunerated RR
  - Notable Practices / Innovations: Weekly publication of system liquidity need; sustained tight alignment of o/n to policy
- Bank of Russia (Russia)
  - Publishes Forecasts?: Partial — daily liquidity data; structural outlook notes
  - Forecast Horizon: Short-term (1-week ahead) & structural (months)
  - Forecast Frequency: Daily monitoring; weekly auction cycle
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Very close — fiscal rule FX ops & MinFin flows integrated
  - Main Operations & Framework: In surplus: weekly deposit auctions (FRFA at key rate); OBR issuance for medium-term absorption; fine-tuning deposits/repos/FX swaps; standing facilities; reserve averaging
  - Notable Practices / Innovations: Regular issuance of central-bank bonds (OBR) for sterilization; detailed daily liquidity factors
- Central Bank of the Republic of Türkiye (Turkey)
  - Publishes Forecasts?: No public bulletin (internal indicators)
  - Forecast Horizon: Short-term (day-to-day; some structural)
  - Forecast Frequency: Daily (frequent adjustments)
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Partial — operational info-sharing; coordination on FX-protected deposits
  - Main Operations & Framework: Mix of 1-week repos (when needed), RR changes, targeted bond sales; FX swaps; standing facilities
  - Notable Practices / Innovations: Recent shift back toward orthodoxy; emphasis on simplification and transparency in operations; evolving corridor use
- Bank Negara Malaysia (Malaysia)
  - Publishes Forecasts?: Yes — intraday liquidity updates (FAST platform)
  - Forecast Horizon: Immediate (intraday) & short-term (few days)
  - Forecast Frequency: Multiple times daily
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Strong — Cash Coordination Committee; daily MoF flow sharing
  - Main Operations & Framework: Daily tenders (absorb/inject) across maturities; BNMN (own bills) for term sterilization; FX swaps; Islamic operations in parallel; RR (SRR) flexible
  - Notable Practices / Innovations: High-frequency transparency; dual conventional-Islamic framework; automated tenders
- Bank Indonesia (Indonesia)
  - Publishes Forecasts?: Yes — daily “Liquidity Projection” table
  - Forecast Horizon: Short-term (daily & near-term)
  - Forecast Frequency: Daily (morning + mid-day updates)
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Formal — SKB with MoF; TSA at BI; primary dealer coordination
  - Main Operations & Framework: Corridor framework; RR repos to inject; BI securities/term deposits to absorb; FX swaps; Islamic OMOs; RR with partial averaging
  - Notable Practices / Innovations: DNDF & SRBI instruments; transparency via daily projection (including Islamic segment)
- Bank of Israel (Israel)
  - Publishes Forecasts?: No public bulletin (relies on data; publishes ops/data)
  - Forecast Horizon: Short-term (current day/week); structural view months ahead
  - Forecast Frequency: Daily (internal); monthly Makam issuance
  - Forecast Errors Published?: No
  - Coordination with Fiscal Authority: Moderate — consults MoF on T-bill vs Makam volumes
  - Main Operations & Framework: Monthly Makam bills to absorb liquidity; repos to supply; standing facilities; unremunerated RR
  - Notable Practices / Innovations: Makam-based structural sterilization; slight structural deficit to anchor o/n

### Cross-cutting themes and operational patterns
- Forecast publication and transparency
  - Some central banks publish daily public liquidity forecasts or projections (e.g., Česká národní banka, Bank Indonesia, Bank Negara Malaysia, Bank of Canada posts daily settlement balance target).
  - Many central banks maintain internal daily planning without a public bulletin.
  - Where published, some disclose forecast errors (Česká národní banka).
- Coordination with fiscal authorities
  - Varies from strong formal integration (e.g., Banco Central do Brasil; Bank of Canada; Bank Indonesia; Central Bank of Nigeria — TSA flows incorporated) to informal or limited mechanisms (e.g., Banco Central del Uruguay; Banco Central de la República Argentina).
  - TSA flows and issuance calendars are frequently cited as key coordination channels.
- Main operational frameworks
  - Common tools: repos/reverse-repos, standing facilities, central bank bills, FX swaps, deposit/term auctions, reserve requirements (RR/CRR), and full-allotment operations.
  - Systems vary between corridor frameworks, fixed-rate full-allotment, and ample-reserves floor systems.
  - Several central banks use reserve averaging, fine-tuning operations, or central-bank bond issuance (OBR) for sterilization.
- Notable operational innovations
  - Use of high-frequency or intraday updates (Bank Negara Malaysia).
  - Dual conventional-Islamic operations (Bank Negara Malaysia; Bank Indonesia includes Islamic segment).
  - Use of CRR/ RR as quasi-OMO (Central Bank of Nigeria).
  - Corridor reforms and collateralized interbank platforms (Central Bank of Kenya; 2023 corridor reform; DhowCSD).
  - Variable late-window penalty structures to incentivize early squaring (Bank of Ghana: normal ±100bp; late window ±500bp).
- Maintenance/forecast horizons
  - Horizons reported range from intraday and next-day to maintenance period (~6 weeks) and internal scenarios up to ~1y (Federal Reserve).
  - Frequency of operational auctions ranges from daily to weekly depending on the central bank’s framework and market structure.

*Annex X. Comparative Table of Central Bank Liquidity — content as provided in the source document.*

---


_Source: https://www.imf.org/-/media/files/publications/tar/2026/english/tarea2026037.pdf_