## 1. Balance Sheet Impact: Reserves at Fed vs. RRP rate

## Source details

**Canonical URL:** [1. Balance Sheet Impact: Reserves at Fed vs. RRP rate](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14111.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14111.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14111.pdf.json)

---

### I. Introduction — nexus of financial stability, monetary policy, and plumbing
- Focus: the financial plumbing connecting bank and nonbank balance sheets, repo markets, and changes from proposed regulations and Fed intermediation.
- Core plumbing components:
  - The repo markets (bilateral repo and Tri-party repo (TPR)).
  - The bank deposit market.
- Size estimates and context:
  - U.S. Tri-party repo (TPR) market currently estimated at US$1.8 trillion from a peak of almost US$3 trillion before the Lehman crisis.
  - New York Fed work indicates the bilateral repo market over US$1 trillion; including collateral reuse the bilateral market may be around US$ 2 trillion–3 trillion, and larger than the TPR.
- Paper structure (sections referenced):
  - Section II: regulatory/financial plumbing nexus and repo rates going forward.
  - Section III: marginal rate of substitution between bilateral repo and TPR and impact on collateral velocity (re-use).
  - Section IV: increasing nexus between financial plumbing and monetary policy.
  - Section V: central banks active in QE should monitor drainage of reserves (e.g., high quality liquid assets) relative to money metrics when liftoff from ZLB.

### Box 1 — Bilateral Repo Markets: some estimates (hedge fund collateral and reuse)
- Hedge fund financing channels: (i) pledging collateral to prime brokers (PB) to borrow money, or (ii) repurchase agreements (repo) with PBs or another dealer.
- End-2007 estimate (method and figures):
  - Assets under management (AUM) of $2 trillion.
  - 27 percent share of strategies that would use repo.
  - Aggregate leverage of 4 (source: FSA hedge fund surveys, United Kingdom).
  - Implied approx US$540 billion times four = US$2.2 trillion pledged collateral could have gone to banks.
  - Due to hedging (about 60–70 percent hedged), only one-third of US$2.2 trillion could reach banks for reuse → US$750 billion pledged collateral reusable as of end-2007.
  - Threshold varies with rate cycle (closer to 60 percent near top of cycle; over 70 percent near bottom), affecting pledged collateral available for reuse.
- End-2013 estimate (method and figures):
  - AUM higher at US$2.6 trillion.
  - Aggregate leverage (including derivative use) lower at 3.5.
  - Share of HF strategies using repo around 40 percent.
  - Resulting estimate at US$900 billion (adjusted downward due to higher hedging threshold at bottom of rate cycle).
- Collateral reuse factor and market size:
  - With collateral reuse factor between 2 and 3, the size of the bilateral global repo market is at par or larger than the TPR in the U.S.
- Caveats and heterogeneity:
  - If about two-thirds of strategies are hedged, collateral from the remaining one-third may not all be reused/turned to cash by banks—it depends on bank balance sheet space and forthcoming regulations.
  - Banks differ: some curtail pledged collateral activities (example cited: UBS), others try to enter the market.

### II. The financial plumbing — example from U.S. Federal Reserve and implications
- QE effects on bank balance sheets:
  - QE purchases of HQLA from nonbanks converted nonbank assets into deposit liabilities at banks and corresponding bank asset of reserves at the Fed.
  - QE converted good collateral in the market into excess balances at banks.
  - Banks facing leverage ratio constraints due to QE seek balance sheet “space” for higher-return financial intermediation/non-depository activities.
- Fed RRP trial and rate wedge:
  - Fed started trial “testing” a repo rate floor via the RRP (last September, as stated).
  - Interest on excess reserves (IOER) is currently paid at a rate of 25bps.
  - Fed’s reverse repo program currently pays interest at a rate of up to 5 bps.
  - Even with 25 bps IOER pulling repo rates up, Fed “put” a 3 bps to 5 bps floor (via RRP).
  - Gagnon/Sack proposal: eliminating the wedge between RRP floor and IOER and making RRP full allotment would allow Fed to set price (P) and let market determine quantity (Q) of reserves on Fed’s balance sheet.
- Accounting mechanics example:
  - For every US$100 million of reverse repos: RRP liability on Fed’s balance sheet increases by US$100 million and bank excess reserves liability decreases by US$100 million; Fed’s total balance sheet unchanged.
- Operational restrictions and “capped rehypothecation”:
  - RRP operational structure restricts reuse of collateral outside the Triparty system.
  - Collateral from the Fed can be reused within the GSD of DTCC in the GCF Triparty system, but reuse only within Triparty repo liabilities; reuse cannot be used for posting at central clearinghouses, bilateral derivatives markets, or bilateral repo market.
  - Collateral remaining as asset on the Fed’s balance sheet and within Triparty has limited velocity; securities in market possession have velocity; those at the central bank do not.
- Interaction between Triparty RRP and bilateral repo:
  - The bilateral repo market (core of bank/nonbank nexus) is outside the Triparty framework.
  - Nonbank RRPs with the Fed create bank balance sheet space allowing banks to be more active in bilateral repo, enhancing link between Triparty and bilateral markets.
  - Demand from bilateral repo could lead banks (if they have balance sheet space after adjusting for HQLA/leverage ratio/LCR) to make markets for clients not eligible for RRP, undertaking collateral transformation including substitution of their balance sheet collateral with RRP collateral.
- Example of potential rate wedge and intermediation frictions:
  - A bank could lend to the Fed under RRP (assume at 25 bps) or lend to a hedge fund at 30 bps — bilateral repo rate above the 25 bps RRP.
  - Alternatively, repo between pension fund and bank may be 20 bps or even 10 bps after factoring the FDIC levy that increases with a larger balance sheet.
  - Resulting wedge around the 25 bps RRP could be from 30 bps to 10 bps.
  - Wedge removable only if hedge funds could deal with pension funds directly; inability highlights role of financial intermediation and frictions.

### Box 2 — General Collateral (GC) Rate
- Definition and measurement issues:
  - GC rate used as a benchmark or proxy for collateral within the Triparty repo framework; GCF rate provided by DTCC used for simplicity.
  - Ambiguity in GC concept: unclear whether GC refers to the offer rate the money fund/bank sees, the bid rate a highly rated borrower sees from the money fund, or the bid rate an unrated borrower sees from the rated borrower.
- Market mechanics and counterparty differences:
  - Surplus money can be deposited at a bank to receive IOER (less bank’s cut) or placed in a money fund to receive the RRP rate (adjusted for money fund’s overall return).
  - Highly rated borrowers can access money from money funds at IOER/RRP plus a very tiny risk adjustment of 1–2 basis points.
  - Less-rated borrowers who own HQLA borrow from banks at IOER minus expense (which exceeds that for highly rated borrowers).
  - Unrated borrowers without HQLA borrow at IOER (plus even higher expense) plus collateral transformation fee, or at RRP rate plus collateral transformation with higher fee relative to borrowers with HQLA.
- Role of Triparty versus bilateral repo markets:
  - GC rate (via Tri-party repo) was aligned to the Fed Funds rate prior to 2008 crisis.
  - GC rate does not provide information on the bilateral repo market, which may be larger than the Triparty repo market.
  - Collateral desks in bilateral markets sort large incoming blocks of collateral for higher-value pieces that can be rehypothecated bilaterally; remainder flows to Triparty repo.
- International perspective:
  - ECB did not resort to a floor when repo rates turned negative after cutting deposit rate to zero in July 2012; repo rates of good collateral remained below zero for much of that period.
  - EONIA moved into positive territory as excess liquidity declined with LTRO repayments.
  - U.K. does not provide a floor to repo rates (RONIA); their bank rate is at 50 bps.
  - Fed provides a floor via RRP (contrast with ECB and U.K.).
- RRP, collateral velocity, and financial plumbing:
  - Expanding RRP (counterparty limits, higher rate, more eligible counterparties) shifts bilateral repo demand inward, reducing bilateral repo volumes and the marginal rate of substitution between Tri-party and bilateral repo markets.
  - RRP takes money funds off bank balance sheets, freeing up bank balance sheet space (and HQLA) for non-depository activities such as collateral services and derivatives intermediation.
  - If RRP remains overnight, described as “accounting drainage” rather than drainage of reserves; this mutes collateral velocity and reduces overall financial lubrication.
  - Distinction between ownership and possession: Fed ownership of securities (duration removal) differs from market need for collateral services, which transfer with possession. By replacing reserves with RRP, the Fed transfers possession back to the market with restricted velocity.
- Regulatory impact: SLR and balance sheet considerations:
  - Reserve deposits are included in the SLR calculation; banks may prefer the Fed do RRP with their customers rather than take deposits themselves to place at IOER.
  - With sizable use of RRP, the SLR becomes less binding on U.S. banks (relative to RWA) as the Fed takes low-risk, low-return business off banks’ balance sheets.
  - Note: SLR is not effective until Jan 1, 2018.
- Numerical illustration (as described in source):
  - Notional Contract 1,000,000$
  - RoE hurdle rate for a typical Bank 8%-10%
  - Leverage Ratio 5%
  - Borrow in Fed Funds Market 7 basis points
  - IOER Rate 25 basis points
  - Annual Arbitrage Income 1,800$
  - Equity Capital Tied to $1m Trade 50,000$
  - RoE on Trade (W / CET1) 3.60%
  - Reserves less than RoE hurdle rate; Return may exceed hurdle rate due to repo netting
- Scenario: RRP rate as the policy rate (Box 3 summary):
  - As RRP rate rises with stronger output and inflation pressures, the Fed’s cost of doing reverse repos increases.
  - Anticipated changes as RRP rises:
    - RRP rates moving from ZLB toward 25 bps to 50 bps to 75 bps and higher.
    - More eligible counterparties (e.g., asset managers, CCPs).
    - Potentially higher bids for quantities of collateral by eligible counterparties.
  - Financial intermediation consequences:
    - RRP provides principal protection and some return to nonbanks like money funds and GSEs, but does not enable collateral reuse.
    - Deposits moving away from banks create balance sheet space for non-depository activities; IOER may become the floor for those not eligible to bid for RRP and may increase along with RRP.
  - Comparison: Reserve Bank of Australia’s CLF noted as an alternative mechanism to address HQLA scarcity.
- Asset sales versus liability rearrangement:
  - Selling Fed assets outright allows assets to move directly to final holders and increases collateral velocity (which may conflict with a controlled liftoff from ZLB).
  - Rearranging liabilities (reserves versus RRP) requires balance sheet accommodation at each step and can constrain collateral reuse and mobility.
- Risks and policy implications:
  - Uncertainty about links between Triparty and bilateral markets, hedge fund demand for money, and pension/insurer demand for collateral can create wedges among bilateral repo rate, GC rate via the Triparty repo, and rate on Fed’s RRP operations.
  - A successful lift off from ZLB should avoid leaving structural wedges; exit strategy should consider that a sizable, quick reduction in reserve balances could create wedges between these rates.
  - Central banks active in QE should monitor drainage of reserves (e.g., high quality liquid assets) relative to money metrics when liftoff from ZLB.

*Source: IMF working paper section titled "1. Balance Sheet Impact: Reserves at Fed vs. RRP rate" (excerpts).*

### 1. Balance Sheet Impact: Reserves at Fed vs. RRP rate ..........................................................12

### 1. Balance Sheet Impact: Reserves at Fed vs. RRP rate

### I. Introduction — nexus of financial stability, monetary policy, and plumbing
- Increasing research links financial stability considerations to monetary policy decisions (e.g., “Market Tantrums and Monetary Policy,” Feroli et al.; speech by Fed’s Jeremy Stein).
- Paper focus: the financial plumbing connecting bank and nonbank balance sheets, repo markets, and changes from proposed regulations and Fed intermediation.
- Core plumbing components:
  - The repo markets (bilateral repo and Tri-party repo (TPR)).
  - The bank deposit market.
- Size estimates and context:
  - U.S. Tri-party repo (TPR) market currently estimated at US$1.8 trillion from a peak of almost US$3 trillion before the Lehman crisis.
  - New York Fed work indicates the bilateral repo market over US$1 trillion; including collateral reuse the bilateral market may be around US$ 2 trillion–3 trillion, and larger than the TPR.
- Structure of the paper (sections referenced):
  - Section II: regulatory/financial plumbing nexus and repo rates going forward.
  - Section III: marginal rate of substitution between bilateral repo and TPR and impact on collateral velocity (re-use).
  - Section IV: increasing nexus between financial plumbing and monetary policy.
  - Section V: central banks active in QE should monitor drainage of reserves (e.g., high quality liquid assets) relative to money metrics when liftoff from ZLB.

### Box 1 — Bilateral Repo Markets: some estimates (hedge fund collateral and reuse)
- Hedge fund (HF) financing channels: (i) pledging collateral to prime brokers (PB) to borrow money, or (ii) repurchase agreements (repo) with PBs or another dealer.
- End-2007 estimate method and numbers:
  - Assets under management (AUM) of $2 trillion.
  - 27 percent share of strategies that would use repo.
  - Aggregate leverage of 4 (source: FSA hedge fund surveys, United Kingdom).
  - Implied approx US$540 billion times four = US$2.2 trillion pledged collateral could have gone to banks.
  - Due to hedging (about 60–70 percent hedged), only one-third of US$2.2 trillion could reach banks for reuse → US$750 billion pledged collateral reusable as of end-2007.
  - Note: threshold varies with rate cycle (closer to 60 percent near top of cycle; over 70 percent near bottom), affecting pledged collateral available for reuse.
- End-2013 estimate method and numbers:
  - AUM higher at US$2.6 trillion.
  - Aggregate leverage (including derivative use) lower at 3.5.
  - Share of HF strategies using repo around 40 percent.
  - Resulting estimate at US$900 billion (adjusted downward due to higher hedging threshold at bottom of rate cycle).
- Collateral reuse factor and market size:
  - With collateral reuse factor between 2 and 3 (largely due to inter-dealer collateral moves), the size of the bilateral global repo market is at par or larger than the TPR in the U.S.
- Caveats and heterogeneity:
  - If about two-thirds of strategies are hedged, collateral from the remaining one-third may not all be reused/turned to cash by banks—it depends on bank balance sheet space and forthcoming regulations.
  - Banks differ: some curtail pledged collateral activities (example cited: UBS), others try to enter the market.

### II. The financial plumbing — example from U.S. Federal Reserve and implications
- QE effects on bank balance sheets:
  - QE purchases of HQLA from nonbanks converted nonbank assets into deposit liabilities at banks and corresponding bank asset of reserves at the Fed.
  - QE converted good collateral in the market into excess balances at banks.
  - Banks facing leverage ratio constraints due to QE seek balance sheet “space” for higher-return financial intermediation/non-depository activities.
- Fed RRP trial and rate wedge:
  - Fed started trial “testing” a repo rate floor via the RRP (last September, as stated).
  - Interest on excess reserves (IOER) is currently paid at a rate of 25bps.
  - Fed’s reverse repo program currently pays interest at a rate of up to 5 bps.
  - Even with 25 bps IOER pulling repo rates up, Fed “put” a 3 bps to 5 bps floor (via RRP).
  - Discussion of eliminating the wedge between RRP floor and IOER and making RRP full allotment (Gagnon/Sack proposal): would allow Fed to set price (P) and let market determine quantity (Q) of reserves on Fed’s balance sheet.
- Accounting mechanics (example):
  - For every US$100 million of reverse repos: RRP liability on Fed’s balance sheet increases by US$100 million and bank excess reserves liability decreases by US$100 million; Fed’s total balance sheet unchanged.
- Operational restrictions and “capped rehypothecation”:
  - RRP operational structure restricts reuse of collateral outside the Triparty system.
  - Collateral from the Fed can be reused within the GSD of DTCC in the GCF Triparty system, but reuse only within Triparty repo liabilities; reuse cannot be used for posting at central clearinghouses, bilateral derivatives markets, or bilateral repo market.
  - Thus, collateral remaining as asset on the Fed’s balance sheet and within Triparty has limited velocity; securities in market possession have velocity; those at the central bank do not.
  - Within present Triparty structure none of the collateral can be used to post at CCPs, in bilateral derivatives markets, or in the bilateral repo market, or delivered against short positions.
- Interaction between Triparty RRP and bilateral repo:
  - The bilateral repo market (core of bank/nonbank nexus) is outside the Triparty framework.
  - Fed Governor Tarullo mentioned bilateral repo market size at US$1 trillion (may be higher with collateral velocity).
  - Nonbank RRPs with the Fed create bank balance sheet space allowing banks to be more active in bilateral repo, enhancing link between Triparty and bilateral markets.
  - Demand from bilateral repo could lead banks (if they have balance sheet space after adjusting for HQLA/leverage ratio/LCR) to make markets for clients (e.g., pension funds/insurers) not eligible for RRP, undertaking collateral transformation including substitution of their balance sheet collateral with RRP collateral.
- Example of potential rate wedge and intermediation frictions:
  - A bank with surplus money could lend to the Fed under RRP (assume at 25 bps) or lend to a hedge fund at 30 bps — bilateral repo rate above the 25 bps RRP.
  - Alternatively, bank surplus HQLA earning 25 bps could be in demand by a pension fund to post at a CCP; the repo rate between pension fund/bank may not exceed 20 bps (perhaps even 10 bps, when factoring the FDIC levy to the bank that increases in line with a larger balance sheet), as the bank takes its “cut.”
  - Resulting wedge around the 25 bps RRP could be from 30 bps to 10 bps.
  - This wedge could only be removed if hedge funds could deal with pension funds directly (disintermediating banks); they cannot, highlighting the role of financial intermediation and frictions.

*Source: IMF working paper section titled "1. Balance Sheet Impact: Reserves at Fed vs. RRP rate" (excerpts).*

### Box 2. General Collateral (GC) Rate

### Box 2. General Collateral (GC) Rate

### Definition and measurement issues
- The GC rate is used as a benchmark or proxy for collateral within the Triparty repo framework; the GCF (general collateral finance) rate provided by DTCC is used for simplicity.
- Ambiguity in the GC concept:
  - Unclear whether GC refers to the offer rate the money fund/bank sees, the bid rate a highly rated borrower sees from the money fund, or the bid rate an unrated borrower sees from the rated borrower.
  - This fuzziness complicates setting a clear policy target for a collateral rate.

### Market mechanics and counterparty differences
- Deposit choices and returns:
  - Surplus money can be deposited at a bank to receive IOER (less bank’s cut) or placed in a money fund to receive the RRP rate (adjusted for money fund’s overall return).
- Funding HQLA:
  - Highly rated borrowers can access money from money funds at IOER/RRP plus a very tiny risk adjustment of 1–2 basis points.
  - Less-rated borrowers who own HQLA borrow from banks at IOER minus expense (which exceeds that for highly rated borrowers).
  - Unrated borrowers without HQLA borrow at IOER (plus even higher expense) plus collateral transformation fee, or at RRP rate plus collateral transformation with higher fee relative to borrowers with HQLA.

### Role of Triparty versus bilateral repo markets
- GC rate (via Tri-party repo) was aligned to the Fed Funds rate prior to 2008 crisis.
- GC rate does not provide information on the bilateral repo market, which may be larger than the Triparty repo market.
- Collateral desks in bilateral markets sort large incoming blocks of collateral for higher-value pieces that can be rehypothecated bilaterally; remainder flows to Triparty repo.

### International perspective and examples
- ECB:
  - Did not resort to a floor when repo rates turned negative after cutting deposit rate to zero in July 2012.
  - Repo rates of good collateral (German Bunds, French Oats) remained below zero for much of that period.
  - EONIA moved into positive territory as excess liquidity declined with LTRO repayments.
- U.K.:
  - Does not provide a floor to repo rates (RONIA); their bank rate is at 50 bps.
- Fed:
  - Provides a floor via RRP (contrast with ECB and U.K.).

### RRP, collateral velocity, and financial plumbing
- Effects of expanding RRP (counterparty limits, higher rate, more eligible counterparties):
  - Shifts bilateral repo demand inward, reducing bilateral repo volumes and the marginal rate of substitution between Tri-party and bilateral repo markets.
  - RRP takes money funds off bank balance sheets, freeing up bank balance sheet space (and HQLA) for non-depository activities such as collateral services and derivatives intermediation.
- If RRP remains overnight, this is described as “accounting drainage” rather than drainage of reserves; this muting of collateral velocity reduces overall financial lubrication.
- Distinction between ownership and possession:
  - Fed ownership of securities (duration removal) differs from market need for collateral services, which transfer with possession.
  - By replacing reserves with RRP, the Fed transfers possession back to the market with restricted velocity.

### Regulatory impact: SLR and balance sheet considerations
- Reserve deposits are included in the SLR calculation; banks may prefer the Fed do RRP with their customers rather than take deposits themselves to place at IOER.
- With sizable use of RRP, the SLR becomes less binding on U.S. banks (relative to RWA) as the Fed takes low-risk, low-return business off banks’ balance sheets.
- Note: SLR is not effective until Jan 1, 2018.
- Table illustration (as described):
  - Notional Contract 1,000,000$
  - RoE hurdle rate for a typical Bank 8%-10%
  - Leverage Ratio 5%
  - Borrow in Fed Funds Market 7 basis points
  - IOER Rate 25 basis points
  - Annual Arbitrage Income 1,800$
  - Equity Capital Tied to $1m Trade 50,000$
  - RoE on Trade (W / CET1) 3.60%
  - Reserves less than RoE hurdle rate; Return may exceed hurdle rate due to repo netting

### Scenario: RRP rate as the policy rate (Box 3 summary)
- As RRP rate rises with stronger output and inflation pressures, the Fed’s cost of doing reverse repos increases.
- Anticipated changes in the new plumbing as RRP rises:
  - RRP rates moving from ZLB toward 25 bps to 50 bps to 75 bps and higher.
  - More eligible counterparties (e.g., asset managers, CCPs).
  - Potentially higher bids for quantities of collateral by eligible counterparties.
- Financial intermediation consequences:
  - RRP provides principal protection and some return to nonbanks like money funds and GSEs, but does not enable collateral reuse.
  - Deposits moving away from banks create balance sheet space for non-depository activities; IOER may become the floor for those not eligible to bid for RRP and may increase along with RRP.
- Comparison to other central bank tools:
  - Reserve Bank of Australia’s CLF is noted as an alternative mechanism to address HQLA scarcity (described for context).

### Asset sales versus liability rearrangement
- Selling Fed assets outright allows assets to move directly to final holders and increases collateral velocity (which may conflict with a controlled liftoff from ZLB).
- Rearranging liabilities (reserves versus RRP) requires balance sheet accommodation at each step and can constrain collateral reuse and mobility.

### Risks and policy implications
- Uncertainty about links between Triparty and bilateral markets, hedge fund demand for money, and pension/insurer demand for collateral can create wedges among:
  - Bilateral repo rate
  - GC rate via the Triparty repo
  - Rate on Fed’s RRP operations
- A successful lift off from ZLB should avoid leaving structural wedges; exit strategy should be mindful that a sizable, quick reduction in reserve balances could create wedges between these rates.
- Central bank roles going forward may include drainage of reserves (e.g., high quality liquid assets) alongside money metrics.

*Source: IMF Working Paper content in Box 2. General Collateral (GC) Rate*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14111.pdf_