## wpiea2025171-source-pdf

## Source details

**Canonical URL:** [wpiea2025171-source-pdf](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025171-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025171-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025171-source-pdf.pdf.json)

---

### Intermediaries and cross-border payments: roles and scale
- Financial intermediaries execute the vast majority of cross-border payments, primarily to settle trades in financial assets.
- For 77 percent of cross-border payment transactions, the transacting parties are financial intermediaries; most of the remaining 23 percent is also processed by financial intermediaries for end-recipients that are households or non-financial businesses (Cerutti et al., 2025b).
- Households generally favor safe, local-currency assets; intermediaries are crucial for channeling savings into international markets.
- Cross-border payment frictions — including both fees and settlement delays — affect intermediaries’ profits and hence their willingness to intermediate cross-border capital flows.

### Market size and quantitative importance of payment frictions
- Global payment and asset figures reported in the source:
  - "30.5  trillion  in  2023  (WTO,  2024)"
  - "total customer-related cross-border payments of USD 190 trillion in 2023 (Cerutti et al., 2025b)."
  - Cerutti et al. (2025a) estimate the global market for cross-border payments "approached one quadrillion dollars in 2024."
  - At end-2023, financial intermediaries held "88 percent of global financial assets," with "banks comprising 39 percent and non-bank financial intermediaries (NBFIs) 49 percent (FSB, 2024a)."
  - "As of 2021, 76 percent of the world population had an account at a financial institution (including mobile money providers) (World Bank, 2022)."
  - In only "five countries (Australia, Canada, New Zealand, UK, US) does more than a quarter of the population own stocks."
- Direct monetary and opportunity costs:
  - Wholesale transfers incur "on average, a 10 basis points charge" for acquiring, servicing, and redeeming foreign financial assets (Cerutti et al., 2025b).
  - A total of "20 basis points lost, on average, between purchasing an asset and receiving its returns" is a relevant investment consideration.
  - The yield difference between 10-year Treasuries issued by Hong Kong SAR and the US is "below 40 basis points most of the time (as measured during 2014-2024) and rarely exceeds 80 basis points."
  - For "92 percent of wholesale transfers, the transfer between the correspondent and the recipient bank is completed within a business day" (FSB, 2024b).
  - "794 billion dollars is in-transit at correspondent banks" at any given moment (McKinsey and SWIFT, 2018, as of 2016).
  - Continuous reinvestment of cash locked during settlement (one business day on the New York Stock Exchange) "reduces portfolio management costs by 22 percent" (J.P. Morgan, 2023).
  - "7.6 percent of countries had at most two wholesale payment corridors" (FSB, 2024b).
- Technological and geopolitical forces:
  - Initiatives to reduce frictions include central bank digital currency (BIS, 2023; Garratt et al., 2024), stablecoins, and unbacked crypto assets.
  - Geopolitical fragmentation can raise costs; some corridors see costs rise "by more than an order of magnitude" (Duffie, 2023; Eichengreen, 2022).

### Model framework and frictions
- Two endowment economies (Home and Foreign); representative household in each consumes tradable and non-tradable goods.
- Households save/borrow via domestic bonds; financiers intermediate imbalances in bond demand and enable current and capital account imbalances.
- Financiers:
  - Consolidated financial intermediaries (e.g., correspondent banks and asset managers) form a unit mass with heterogeneous risk-bearing capacity γ_i ∈ (0,1] and CDF G(γ) = γ^α for α > 1.
  - Two-stage game:
    - Stage 1: financiers decide whether to enter the market for cross-border intermediation, paying fixed cost F to establish payment rails (extensive margin).
    - Stage 2: financiers choose intermediation quantity q_i to maximize mean-variance objective max_{q_i} {Ω q_i − (γ_i / 2) Var(Ω q_i)}, balancing expected profits Ω against exchange rate risk and variable costs.
  - Variable costs / risk heterogeneity captured by parameter α (intensive margin).
- Key financial definitions:
  - Home bond gross return R := 1 + r; Foreign bond gross return R* := 1 + r*.
  - Per-unit expected net profit from intermediation: Ω ≡ R − R* E[e2] / e1.
- Payment-friction channels in the model:
  - Fixed cost F to establish a payments channel (affects entry/extensive margin).
  - Heterogeneous risk-bearing capacity (α) captures variable-cost-like effects and affects the intensive margin.
  - Reductions in variable costs are modeled as a universal increase in risk-bearing capacity (lower α increases intensive supply).

### Equilibrium characterization (selected lemmas and expressions)
- Current account (Lemma 1): CA_1 = CA(e_1, Q_T1, Q_N1); CA(·) is increasing and concave in e_1.
- Capital account (Lemma 2): KA_1 summarized as
  - KA_1 = (α / (α−1)) Ω / Var(e_2) [ R* / e_1 ]^2 ( 0.5 Ω^2 / (F · Var(e_2)) [ R* / e_1 ]^2 )^{α−1}
  - KA_1 = KA(R, R*, e_1, α, F), which is decreasing in e_1.
- Frictionless limit (Lemma 3): as F → 0 and/or α → 1, Ω → 0 and e_1 → R* / R (UIP deviations vanish).
- Parameter restriction imposed: |Ω| < 0.5 R* e_1.

### Main results: levels and UIP deviations (Proposition 1)
- Lower cross-border payment frictions (reduction in F and/or α) lead to:
  - A smaller UIP deviation |Ω|.
    - (a) e1 depreciates if CA1 > 0 (current account surplus).
    - (b) e1 appreciates if CA1 < 0 (current account deficit).
  - A larger capital account and current account position (greater magnitude of imbalances).
- Mechanisms:
  - Lower fixed costs increase the number of profitable financiers (extensive margin).
  - Lower variable costs increase the quantity each financier intermediates (intensive margin).
  - Both margins raise intermediation for any given UIP deviation, increasing capital flows and reducing equilibrium UIP deviations.

### Main results: volatility and shock-type dependence (Propositions 2 and 3; Lemma 4)
- Lemma 4 identity for volatility impacts:
  - ∂Var(x)/∂(-Fric) = 2 Cov( ∂x/∂(-Fric), x ).
  - A reduction in frictions unambiguously increases volatility if sign(dx/ds) = sign( ∂/∂(-Fric) (dx/ds) ) for all shocks s.
- Real shock (shock to Q_T1) — Proposition 2:
  - Lower frictions decrease exchange rate volatility.
  - Lower frictions increase capital flow volatility.
  - Intuition: With lower frictions, adjustment to a real shock shifts from exchange-rate movements toward capital flows; in the frictionless limit, exchange-rate volatility goes to zero and shocks are absorbed by capital flows.
- Financial shock (shock to R*) — Proposition 3:
  - Lower frictions increase exchange rate volatility (unambiguous).
  - Effect on capital flow volatility is ambiguous:
    - Small decreases in frictions increase capital flow volatility.
    - Large enough reductions in frictions decrease capital flow volatility.
  - Intuition: Two opposing effects — increased responsiveness of financiers amplifies capital-account reactions (raising volatility), while larger pre-shock imbalances steepen the current account and can reduce capital-flow responses (lowering volatility).

### Comparative statics and intuition
- Equilibrium solves CA(e_1) + KA(Ω, e_1, α, F) = 0 with E[e_2] normalized to 1.
- As F and/or α fall:
  - Extensive and/or intensive supply of intermediation increase.
  - Capital account becomes more sensitive to UIP deviations; in the frictionless limit UIP deviations are eliminated.
  - Adjustment between quantity margins (net trade and financial flows) and price margins (exchange rate) depends on the slope of the current account at the initial equilibrium.

### Policy-relevant implications and recommendations
- Marginally lower cross-border payment frictions can lead to larger and more volatile capital flows.
- Policymakers should:
  - Be attentive to the source of shocks (real vs. financial), since volatility responses differ by shock type.
  - Consider the size of friction changes, because the qualitative effect on capital flow volatility can reverse for sufficiently large reductions in frictions.
  - For countries joining platforms that reduce fixed and variable costs (shared digital platforms, CBDCs, stablecoin infrastructure), expect:
    - Larger capital flows and more exchange rate volatility after joining.
    - If exchange rate volatility is a concern (e.g., due to high financial dollarization), consider:
      - Maintaining a higher level of foreign exchange reserves to smooth short-run volatility following financial shocks.
      - Reducing financial dollarization.
      - Maintaining larger fiscal buffers to help firms and financial institutions withstand shocks.
  - Use country-specific estimates of current and capital account elasticities, current levels of frictions, and expected magnitudes of decline to assess likely outcomes on capital flow volatility.

### Model assumptions, robustness, and calibration notes
- Representative household with log utility and Cobb-Douglas tradable/non-tradable consumption; results generalize to any current account increasing and concave in e_1.
- Financiers combine asset management and payment conduit roles; in reality these roles may be split and profits divided across entities.
- Frictions abstracted into fixed cost F and heterogeneity parameter α; qualitative implications similar across channels though channels operate differently.
- Cross-border payment frictions are modeled separately from trade frictions (shipping, tariffs), which are typically much larger and outside the paper’s focus.
- Calibration values used for illustrative figures (Table C.1):
  - QT1 = 2
  - QN1 = 1
  - R = 1.02
  - R* = 1.025
  - Var(e2) = 2
  - ρ = 0.35
  - p∗1 = 1
  - ∆QT1 = −0.01 QT1 = −0.02
  - ∆R∗ = 0.025
- Robustness checks: qualitative results hold when allowing intermediary profits to enter foreign household’s budget constraint; illustrative magnitudes are stylized.

### Key takeaways
- Reducing cross-border payment frictions:
  - Lowers equilibrium UIP deviation Ω and moves e_1 toward R* / R in the frictionless limit.
  - Tends to depreciate home currency when starting from a current account surplus (and the opposite for a deficit).
  - Dampens exchange-rate volatility to real shocks while amplifying capital-flow volatility to real shocks.
  - Amplifies exchange-rate volatility to financial shocks; effect on capital-flow volatility to financial shocks is ambiguous and depends on friction size and initial CA.
- Mechanism summary:
  - Extensive margin (F): lower F → more entrants → greater aggregate intermediation.
  - Intensive margin (α): lower α → greater intermediation per entrant.
  - Volatility outcomes hinge on the covariance between a variable and its sensitivity to frictions (Lemma 4).
- Practical implication: assess current frictions, expected friction reductions, intermediary entry and risk-bearing capacity, and CA/KA elasticities when designing policies to reduce cross-border payment frictions.

*Source: wpiea2025171-source-pdf - introduction of new payment rails, including some potentially based on central bank digital*

### introduction of new payment rails, including some potentially based on central bank digital

### introduction of new payment rails, including some potentially based on central bank digital

### Intermediaries and cross-border payments: roles and scale
- Financial intermediaries execute the vast majority of cross-border payments, primarily to settle trades in financial assets.
- For 77 percent of cross-border payment transactions, the transacting parties are financial intermediaries; most of the remaining 23 percent is also processed by financial intermediaries for end-recipients that are households or non-financial businesses (Cerutti et al., 2025b).
- Households generally favor safe, local-currency assets; intermediaries are crucial for channeling savings into international markets.
- Cross-border payment frictions — including both fees and settlement delays — affect intermediaries’ profits and hence their willingness to intermediate cross-border capital flows.

### Model framework
- Two endowment economies, each with a representative household consuming tradable and non-tradable goods.
- Households save/borrow via domestic bonds; imbalances in bond demand are intermediated by financiers who enable current and capital account imbalances.
- Financiers represent consolidated financial intermediaries that both operate and use cross-border payment rails (e.g., correspondent banks and asset managers).
- A unit mass of financiers with heterogeneous risk-bearing capacity play a two-stage game:
  - Stage 1: financiers decide whether to enter the market for cross-border intermediation, incurring a fixed cost reflecting the cost of establishing the payment rail (e.g., becoming a correspondent bank).
  - Stage 2: financiers choose the amount of cross-border intermediation to supply, facing a mean-variance trade-off balancing expected profits against exchange rate risk and a variable cost of using payment rails.
- Reductions in variable costs are modeled as a universal increase in the risk-bearing capacity among financiers, since lower variable costs increase expected profit per unit of risk.

### Main results: levels and UIP deviations
- Reduced frictions lead to:
  - Smaller deviations from uncovered interest rate parity (UIP).
  - Increased capital flows.
- Mechanisms:
  - Lower fixed costs increase the number of financiers that operate profitably (extensive margin).
  - Lower variable costs increase the quantity each financier intermediates (intensive margin).
  - Both margins raise intermediation for any given UIP deviation, increasing capital flows and reducing equilibrium UIP deviations.

### Main results: volatility and shock-type dependence
- Volatility effects depend on the type of shock (real vs. financial) and on the size of friction changes.
- For a real shock:
  - Reduced frictions increase capital flow volatility.
  - Reduced frictions decrease exchange rate volatility.
  - Intuition: In a frictionless world UIP holds so exchange rate is set by interest differentials and real shocks must adjust via capital flows; with frictions exchange rate also adjusts, so lowering frictions shifts adjustment toward capital flows and away from exchange rates.
- For a financial shock:
  - Lower frictions always increase exchange rate volatility.
  - Effect on capital flow volatility is ambiguous:
    - Small decreases in frictions increase capital flow volatility.
    - Large enough reductions in frictions decrease capital flow volatility.
  - Intuition:
    - Lower frictions increase the responsiveness (elasticity) of financiers’ extensive and intensive margin decisions to financial conditions.
    - Smaller frictions move the capital account-current account equilibrium toward larger imbalances; because the current account is concave in the exchange rate, larger pre-shock imbalances imply shocks induce larger exchange rate moves and potentially smaller capital flow responses.

### Policy-relevant implications
- Marginally lower cross-border payment frictions could lead to larger and more volatile capital flows.
- Policymakers planning for changes in cross-border payment frictions should:
  - Be attentive to the source of shocks relevant to their context (real vs. financial), since volatility responses differ by shock type.
  - Consider the size of friction changes, because the qualitative effect on capital flow volatility can reverse for sufficiently large reductions in frictions.

### Connection to literature and novelty
- Two strands of related literature:
  - Papers studying macro implications of introducing new means of payment (e.g., CBDCs, crypto assets), which frequently assume UIP holds and thus do not generate net capital flows.
  - Papers that incorporate UIP deviations but typically do not model payment frictions explicitly.
- This paper’s novelty: focuses on payment frictions faced by intermediaries as a channel affecting UIP deviations, capital flows, and exchange rate level and volatility.
- Closest related works (Gabaix and Maggiori, Basu et al., Kekre and Lenel, Dao et al.) also emphasize intermediaries’ risk-bearing capacity; this paper adds the role of payment frictions (fixed and variable costs) and their effects across shock types.

### Context and motivation
- Introduction of new payment rails (potentially CBDCs, stablecoins, or unbacked crypto assets) and geopolitical tensions that could fragment systems are likely to change cross-border payment frictions.
- Geoeconomic fragmentation could reverse US-dollar system economies of scale and raise the cost of cross-border payments.
- Emerging and developing economies may be especially vulnerable because volatile capital flows and exchange rate movements can trigger financial instability, notably in financially dollarized economies.

*Source: wpiea2025171-source-pdf - introduction of new payment rails, including some potentially based on central bank digital*

### 30.5  trillion  in  2023  (WTO,  2024),  compared  to  total  customer-related  cross-border  pay-

### wpiea2025171-source-pdf - 30.5  trillion  in  2023  (WTO,  2024),  compared  to  total  customer-related  cross-border  pay-

### Market size and the role of intermediaries
- Global cross-border payments estimates in the text:
  - "30.5  trillion  in  2023  (WTO,  2024)"
  - "total customer-related cross-border payments of USD 190 trillion in 2023 (Cerutti et al., 2025b)."
  - Cerutti et al. (2025a) estimate the global market for cross-border payments "approached one quadrillion dollars in 2024."
  - Difference between estimates largely due to exclusion of financial institution–related payments in the USD 190 trillion figure (e.g., Swift message type 202).
- Intermediary concentration and asset ownership:
  - At end-2023, financial intermediaries held "88 percent of global financial assets," with "banks comprising 39 percent and non-bank financial intermediaries (NBFIs) 49 percent (FSB, 2024a)."
- Household behavior and home bias:
  - "As of 2021, 76 percent of the world population had an account at a financial institution (including mobile money providers) (World Bank, 2022)."
  - In only "five countries (Australia, Canada, New Zealand, UK, US) does more than a quarter of the population own stocks."
  - Households that own stocks display a preference for domestically-issued assets.

### Payment frictions and their quantitative importance
- Definition: a payment friction is any market imperfection preventing the costless and instant exchange of money in return for a good or service.
- Direct monetary costs:
  - Wholesale transfers for acquiring, servicing, and redeeming foreign financial assets incur "on average, a 10 basis points charge" (Cerutti et al., 2025b).
  - A total of "20 basis points lost, on average, between purchasing an asset and receiving its returns" is a relevant investment consideration.
  - For comparison: the difference between yields on 10-year Treasuries issued by Hong Kong SAR and the US is "below 40 basis points most of the time (as measured during 2014-2024) and rarely exceeds 80 basis points."
- Opportunity costs from settlement delays:
  - FSB (2024b): for "92 percent of wholesale transfers, the transfer between the correspondent and the recipient bank is completed within a business day" (this excludes originator-to-correspondent leg).
  - McKinsey and SWIFT (2018) (as of 2016): at any given moment "794 billion dollars is in-transit at correspondent banks."
  - J.P. Morgan (2023): continuous reinvestment of cash locked in during settlement (one business day on the New York Stock Exchange) "reduces portfolio management costs by 22 percent."
- Heterogeneity and corridor constraints:
  - FSB (2024b) reports "7.6 percent of countries had at most two wholesale payment corridors," implying limited competition and higher fees in some corridors.
- Technological and geopolitical forces:
  - Initiatives aim to reduce frictions via central bank digital currency (BIS, 2023; Garratt et al., 2024), stablecoins, or unbacked crypto assets.
  - Geopolitical tensions risk exacerbating frictions; "fragmentation of liquidity across multiple networks can raise costs (Duffie, 2023; Eichengreen, 2022)," with some corridors seeing costs rise "by more than an order of magnitude."

### Model: structure and frictions
- Economy and agents:
  - Two-country open-economy model with Home and Foreign, two periods t= 1,2, and two goods: tradable good T and non-tradable good N.
  - Single representative household per country; financiers are a unit mass based in Foreign who intermediate capital flows.
- Preferences and timing:
  - Consumption bundle: C_t = (C_Tt)^ρ (C_Nt)^{1−ρ} with ρ∈(0,1).
  - Period-one utility: u_1 = ln(C_1) + β E[ln(C_2)] with β∈(0,1].
  - Tradable good endowments can be traded internationally at zero cost; non-tradables cannot be stored.
- Financial instruments and financiers:
  - Home bond B (denominated in tradable good) with gross return R := 1 + r; Foreign bond B* with gross return R* := 1 + r*.
  - Financiers decide to pay a fixed entry cost F and choose an intermediation quantity q_i; they start with no capital and match purchases and sales across countries.
  - Financiers maximize a mean-variance objective: max_{q_i} {Ω q_i − (γ_i / 2) Var(Ω q_i)}, where Ω denotes expected net profits per unit and γ_i ∈ (0,1] is heterogeneous risk-bearing capacity with CDF G(γ) = γ^α for α > 1.
- Payment frictions in the model:
  - Fixed cost F to establish a payments channel (extensive margin).
  - Heterogeneous risk-bearing capacity (γ_i > 0) or equivalently variable costs captured by parameter α (intensive margin).
  - A reduction in F to F′ < F increases entry (extensive margin). A reduction in α to α′ < α increases intensive supply of intermediation.

### Equilibrium characterization and key lemmas
- Current account (Lemma 1):
  - Home's period-one current account: CA_1 = Q_T1 − (1 / e_1)^{1−ρ} C_T*1 (Q_N1 / Q_N*1) with C_T*1 = (1 / (1+β*)) [ Q_T*1 + (1 / (1 + r*)) E[Q_T*2] ].
  - CA_1 summarized as CA(e_1, Q_T1, Q_N1); CA(·) is increasing and concave in e_1.
- Capital account (Lemma 2):
  - Home's period-one capital account:
    - KA_1 = (α / (α−1)) Ω / Var(e_2) [ R* / e_1 ]^2 ( 0.5 Ω^2 / (F · Var(e_2)) [ R* / e_1 ]^2 )^{α−1}
  - Summarized as KA_1 = KA(R, R*, e_1, α, F), which is decreasing in e_1.
  - Intuition: KA sign follows Ω (the UIP deviation); capital account responds to profits available to financiers and to frictions F and α.
- Frictionless limit (Lemma 3):
  - In the limit as frictions are eliminated (F → 0 and/or α → 1), Ω → 0 and e_1 → R* / R.
  - Without frictions, UIP deviations are eliminated and the exchange rate adjusts to the ratio of interest rates.
- Parameter restriction:
  - The model imposes |Ω| < 0.5 R* e_1 (profits per unit of intermediation are less than half the expected gross yield on Foreign bonds).

### Comparative statics and intuition for results
- Equilibrium exchange rate e_1 solves CA(e_1) + KA(Ω, e_1, α, F) = 0 with E[e_2] normalized to 1.
- As frictions fall (lower F and/or lower α):
  - Extensive and/or intensive margins of intermediation increase.
  - Capital account becomes more sensitive to UIP deviations; in the frictionless limit UIP deviations are eliminated.
  - Higher returns net of transaction and opportunity costs can entice intermediaries to hold more foreign assets, but equilibrium effects may also alter exchange rate risk and hence the final portfolio allocation.
- The model provides a framework to quantify how changes in cross-border payment frictions affect exchange rates, capital flows, and their volatility.

*Italic: Source — wpiea2025171-source-pdf (canonical URL provided in metadata).*

### 5.1    The impact of changes in frictions on levels

### 5.1    The impact of changes in frictions on levels

### Proposition 1 — main findings
- Proposition 1 (Impact on levels). Lower cross-border payment frictions—i.e., a reduction in F and/or α—results in:
  - 1. A smaller UIP deviation |Ω|, which implies:
    - (a) A depreciation of the exchange rate e1 if the economy has a current account surplus CA1 >0, or
    - (b) An appreciation of the exchange rate if the economy has a current account deficit CA1 <0.
  - 2. A larger capital account and current account position (whether surplus or deficit).

### Intuition and mechanism
- When fixed and/or variable costs of intermediation fall, the quantity of intermediation offered by financiers increases for any given UIP deviation |Ω|. Thus the equilibrium UIP deviation must decrease.
- For a country with a CA surplus (KA in deficit), Ω = R − R* − 1/e1 < 0. With interest rates exogenously fixed, a reduction in the UIP deviation |Ω| must be accompanied by an increase in e1 (a depreciation).
- Analogous logic with flipped signs applies for CA deficits (appreciation when frictions fall).
- A reduction in frictions increases the capital account response to any given |Ω| (capital account curve shifts), while the current account is unaffected by the reduction in F and/or α. The equilibrium where current account surplus and capital account deficit balance shifts outwards to larger imbalances and a depreciated exchange rate.
- The extent of adjustment on the quantity margin (net trade and financial flows, x-axis) versus the price margin (exchange rate, y-axis) depends on the slope of the current account in the initial equilibrium (see equation 12).

### Visual depiction
- Figure 2 illustrates an economy initially with a current account surplus: reduced frictions shift the capital account from the blue line to the green line; where the capital account intersects the y-axis (no UIP deviation) reduced frictions have no impact; equilibrium shifts outward to larger imbalances and a depreciated exchange rate.

---

### 5.2    The impact of changes in frictions on volatility

### How impacts on volatility are determined (Lemma 4)
- Lemma 4. Let x be the quantity of interest (i.e., capital flows or exchange rates). Then:
  - (i) The impact of a reduction in cross-border payment frictions on the volatility of x is given by
    - ∂Var(x)/∂(-Fric) = 2 Cov( ∂x/∂(-Fric), x ).
  - (ii) A reduction in cross-border payment frictions unambiguously increases volatility if, in reaction to a shocks, it holds that
    - sign(dx/ds) = sign( ∂/∂(-Fric) (dx/ds) ),
    - i.e., x and the responsiveness of x when frictions are reduced (∂x/∂(-Fric)) always move in the same direction. If they always move in opposite directions, volatility is reduced; otherwise the impact is ambiguous.

### Intuition for Lemma 4
- An endogenous variable is more volatile with respect to shocks if its level responds more to them. A reduction in frictions increases volatility if it amplifies that responsiveness.

### Volatility from a real shock (Q_T1) — Proposition 2
- Proposition 2 (Impact on volatility following a real shock). Following a shock to Q_T1, lower cross-border payment frictions, i.e., a reduction in F and/or α,
  - 1. Decrease exchange rate volatility, and
  - 2. Increase capital flow volatility.
- Intuition:
  - A shock to Q_T1 shifts the current account left (contractionary) or right (expansionary). Lower frictions increase aggregate intermediation responsiveness to a given exchange rate change, flattening the capital account curve. A given current account shock then dissipates more through capital account volatility and less through exchange rate volatility.
  - In the limit of unconstrained financiers (UIP holds), the capital account is flat and exchange rate volatility is zero; shocks are entirely absorbed by capital flow volatility.

### Volatility from a financial shock (R*) — Proposition 3
- Proposition 3 (Impact on volatility following a financial shock). Following a shock to R*, lower cross-border payment frictions, i.e., a reduction in F and/or α,
  - 1. Increase exchange rate volatility, and
  - 2. Have an ambiguous effect on capital flow volatility.
- Intuition — two effects:
  - Effect 1: Lower frictions increase the responsiveness of the capital account, amplifying its reaction to R* shocks and raising both exchange rate and capital flow volatility.
  - Effect 2: Lower frictions increase imbalances (Proposition 1), which steepens the current account curve. This steepening raises exchange rate volatility but reduces capital flow volatility in response to the same shock.
  - Net result: exchange rate volatility unambiguously increases; capital flow volatility is ambiguous — the first effect dominates for small reductions in frictions (increasing capital flow volatility), while the second dominates for larger reductions (decreasing capital flow volatility).

### Visual depiction
- Figure 3 illustrates capital flow and exchange rate volatility responses to a real shock as frictions are reduced.
- Figure 4 illustrates capital flow and exchange rate volatility responses to a financial shock as frictions are reduced.
- (Figures are illustrative; vertical axis values are not quantitative.)

---

### 6    Discussion — policy relevance and model assumptions

### Modeling assumptions and robustness
- Representative household: Current account expression (equation 12) derived from a stylized representative household with logarithmic period utility consuming a Cobb-Douglas bundle of tradable and non-tradable goods. Results generalize to any current account that is increasing and concave in the exchange rate.
- Financiers: Model combines asset management and payment conduit roles into a single agent for tractability; in reality these roles can be separate and the operating profit |Ω| would be divided across entities depending on financial sector market structure.
- Microfoundations for frictions: Frictions are abstracted into two groups — fixed costs (F) and variable costs (α). Reductions in both have qualitatively similar implications, with differences in channels (fixed costs directly affect the extensive margin; variable costs directly affect both extensive and intensive margins).
- Cross-border payments vs trade in goods: The model centers on payments for trade in assets; changes in payment frictions may affect trade in goods and services but those frictions (e.g., shipping costs, tariffs) are typically orders of magnitude larger than wholesale payment frictions and thus are not the paper’s focus.

### Example interpretation and policy implications
- Financial platforms (shared digital platforms enabling direct transactions and settlement across intermediaries in different countries) reduce fixed costs (cost to join platform) and variable costs (cost and speed efficiency vs correspondent banking).
- For a policymaker in a country historically vulnerable to financial shocks joining such a platform:
  - Expect larger capital flows and more exchange rate volatility after joining.
  - If concerned about exchange rate volatility (e.g., due to high financial dollarization), consider:
    - Maintaining a higher level of foreign exchange reserves to smooth short-run volatility following financial shocks.
    - Reducing financial dollarization.
    - Maintaining larger fiscal buffers to help firms and financial institutions withstand shocks.
- Direction of change in capital flow volatility depends on:
  - Current level of cross-border payment frictions.
  - Expected magnitude of their decline after joining the platform.
  - Elasticities of current and capital accounts with respect to exchange rates.
  - These inputs can be estimated for a given country and combined with the framework to guide expectations on capital flow volatility.

---

*Source: excerpt from the IMF working paper chapter “5.1    The impact of changes in frictions on levels” (wpiea2025171-source-pdf).*

### References

### References

### Monetary policy, open economies, and exchange rates
- Adrian, T., C. J. Erceg, M. Kolasa, J. Linde, and P. Zabczyk(2022a): “Managing Monetary  Tradeoffs  in  Vulnerable  Open  Economies,”  CEPR  Discussion  Papers  16972, CEPR.
- Adrian, T., F. Grinberg, T. Mancini-Griffoli, R. M. Townsend, and N. Zhang (2022b):  “A Multi-Currency Exchange and Contracting Platform,” IMF Working Papers 2022/217, International Monetary Fund.
- Bacchetta,  P.,  M.  Davenport,  and  E.  van  Wincoop(2022):   “Can  Sticky  Portfolios  Explain  International  Capital  Flows  and  Asset  Prices?”Journal  of  International  Economics, 136, 103583, NBER International Seminar on Macroeconomics 2021.
- Basu,  S.  S.,  E.  Boz,  G.  Gopinath,  F.  Roch,  and  F.  D.  Unsal(2023):   “Integrated Monetary and Financial Policies for Small Open Economies,” IMF Working Papers 2023/161, International Monetary Fund.
- Benigno,  P.,  L.  M.  Schilling,  and  H.  Uhlig(2022):  “Cryptocurrencies,  Currency Competition,  and  the  Impossible  Trinity,”Journal  of  International  Economics,  136, 103601.
- Cova, P., A. Notarpietro, P. Pagano, and M. Pisani(2022):  “Monetary Policy in the Open Economy with Digital Currencies,”Bank of Italy Temi di Discussione (Working Paper) No, 1366.
- Fanelli, S. and L. Straub(2021):  “A Theory of Foreign Exchange Interventions,”The Review of Economic Studies, 88, 2857–2885.
- Ferrari  Minesso,  M.,  A.  Mehl,  and  L.  Stracca(2022):   “Central  Bank  Digital  Currency in an Open Economy,”Journal of Monetary Economics, 127, 54–68.
- George,  A.,  T.  Xie,  and  J.  D.  Alba(2021):  “Central  Bank  Digital  Currency  with Adjustable Interest Rate in Small Open Economies,”Available at SSRN 3605918.
- Ikeda, D.(2022):  “Digital Money as a Medium of Exchange and Monetary Policy in Open Economies,” IMES Discussion Paper Series 22-E-10, Institute for Monetary and Economic Studies, Bank of Japan.
- Engel, C. and S. P. Y. Wu(2024):  “Exchange Rate Models are Better than You Think, and  Why  They  Didn’t  Work  in  the  Old  Days,”  NBER  WP  32808,  National  Bureau  of Economic Research.

### Covered Interest Parity, UIP deviations, and dollar role
- Accominotti, O., J. Cen, D. Chambers, and V. Degorce(2025):  “Covered Interest Parity:  The Long Run Evidence,” Working paper.
- Albagli,  E.,  L.  Ceballos,  S.  Claro,  and  D.  Romero(2024):   “UIP  Deviations: Insights from Event Studies,”Journal of International Economics, 148, 103877.
- Avdjiev, S., W. Du, C. Koch, and H. S. Shin(2019):  “The Dollar, Bank Leverage, and Deviations from Covered Interest Parity,”American  Economic  Review:  Insights, 1, 193–208.
- Du,  W.,  A.  Tepper,  and  A.  Verdelhan(2018):  “Deviations  from  Covered  Interest Rate Parity,”The Journal of Finance, 73, 915–957.
- Bacchetta, P., J. S. Davis, and E. van Wincoop(2023):  “Dollar Shortages, CIP Deviations, and the Safe Haven Role of the Dollar,” NBER Working Papers 31937, National Bureau of Economic Research, Inc.
- Dao, M. C., P.-O. Gourinchas, and O. Itskhoki(2025):  “Breaking Parity:  Equilibrium Exchange Rates and Currency Premia,” Working paper.

### Cross-border payments, correspondent banking, and CBDCs
- BIS(2023):  “Lessons learnt on CBDCs,” Report submitted to the G20 Finance Ministers and Central Bank Governors, Bank for International Settlements, Basel, Switzerland.
- Cerutti, E., M. Firat, and M. Hengge(2025a): “Global Cross-Border Payments: A$1 Quadrillion Evolving Market?”  IMF Working Papers 2025/120, International Monetary Fund.
- Cerutti, E., M. Firat, and H. Perez-Saiz(2025b):  “Estimating the Impact of Digital Money on Cross-Border Flows:  Scenario Analysis Covering the Intensive Margin,” IMF Fintech Notes 2025/002, International Monetary Fund.
- CPMI(2020): “Enhancing Cross-Border Payments: Building Blocks of a Global Roadmap,” Stage 2 report to the G20 – technical background report, Bank for International Settlements, Basel, Switzerland.
- CPMI(2024):  “Linking Fast Payment Systems Across Borders:  Governance and Oversight – Final Report,” Tech. rep., Committee on Payments and Market Infrastructures, Bank for International Settlements.
- Garratt, R., P. Koo Wilkens, and H. S. Shin(2024):  “Next Generation Correspondent Banking,” BIS Bulletins 87, Bank for International Settlements.
- Duffie, D.(2023):  “Fragmentation Risks to the Dollar-Dominated International Financial Order,”Keynote Speech at Asian Bureau of Finance and Economic Research.
- Eichengreen, B.(2022):  “Sanctions, SWIFT, and China’s Cross-Border Interbank Payments System,”CSIS Briefs.
- IMF(2020):  “Digital Money Across Borders:  Macro-Financial Implications,” IMF Policy Paper No. 2020/050.
- Cerutti, E., M. Firat, and M. Hengge(2025a): “Global Cross-Border Payments: A$1 Quadrillion Evolving Market?”  IMF Working Papers 2025/120, International Monetary Fund.
- Cerutti, E., M. Firat, and H. Perez-Saiz(2025b):  “Estimating the Impact of Digital Money on Cross-Border Flows:  Scenario Analysis Covering the Intensive Margin,” IMF Fintech Notes 2025/002, International Monetary Fund.

### Global financial cycle, capital flows, and financial stability
- Acalin, J.(2023): “The Global Financial Cycle Meets Global Imbalances,” Working paper.
- Gabaix, X. and M. Maggiori(2015):  “International Liquidity and Exchange Rate Dynamics,”The Quarterly Journal of Economics, 130, 1369–1420.
- Davis, J. S. and E. van Wincoop(2018):  “Globalization and the Increasing Correlation Between Capital Inflows and Outflows,”Journal of Monetary Economics, 100, 83–100.
- Davis, J. S. and E. van Wincoop(2024):  “A Theory of Capital Flow Retrenchment,”Journal  of  International  Economics, 150, 103952.
- Goldberg, L. and N. Cetorelli(2011):  “Global Banks and International Shock Transmission:  Evidence from the Crisis,”IMF Economic Review, 59, 41–76.
- Gelos, G. and R. Sahay, eds. (2023):Shocks and Capital Flows:  Policy Responses in a Volatile World, Washington, DC: International Monetary Fund.
- FSB(2024a):  “2024  Global  Monitoring  Report  on  Non-Bank  Financial  Intermediation,” Tech. rep., Financial Stability Board.
- FSB(2024b):  “Annual  Progress  Report  on  Meeting  the  Targets  for  Cross-border  Payments: 2024 Report on Key Performance Indicators,” Tech. rep., Financial Stability Board.
- Cooper,  I.,  P.  Sercu,  and  R.  Vanpée(2013):   “The  Equity  Home  Bias  Puzzle:   A Survey,”Foundations and Trends in Finance, 7, 289–416.
- Ardalan, K.(2019): “Equity Home Bias: A Review Essay,”Journal of Economic Surveys, 33, 949–967.
- Dornbusch, R.(1976): “Expectations and Exchange Rate Dynamics,”Journal of Political Economy, 84, 1161–1176.
- Gabaix, X. and M. Maggiori(2015):  “International Liquidity and Exchange Rate Dynamics,”The Quarterly Journal of Economics, 130, 1369–1420.
- Duffie, D.(2023):  “Fragmentation Risks to the Dollar-Dominated International Financial Order,”Keynote Speech at Asian Bureau of Finance and Economic Research.

### Theoretical, methodological, and miscellaneous contributions
- Chen, K., M. Kolasa, J. Lindé, H. Wang, P. Zabczyk, and M. J. Zhou(2023): “An Estimated DSGE Model for Integrated Policy Analysis,” IMF Working Papers 2023/135, International Monetary Fund.
- Fanelli, S. and L. Straub(2021):  “A Theory of Foreign Exchange Interventions,”The Review of Economic Studies, 88, 2857–2885.
- Cooper,  I.,  P.  Sercu,  and  R.  Vanpée(2013):   “The  Equity  Home  Bias  Puzzle:   A Survey,”Foundations and Trends in Finance, 7, 289–416.
- Bacchetta,  P.,  M.  Davenport,  and  E.  van  Wincoop(2022):   “Can  Sticky  Portfolios  Explain  International  Capital  Flows  and  Asset  Prices?”Journal  of  International  Economics, 136, 103583, NBER International Seminar on Macroeconomics 2021.
- Gabaix, X. and M. Maggiori(2015):  “International Liquidity and Exchange Rate Dynamics,”The Quarterly Journal of Economics, 130, 1369–1420.
- Duffie, D.(2023):  “Fragmentation Risks to the Dollar-Dominated International Financial Order,”Keynote Speech at Asian Bureau of Finance and Economic Research.

*References — wpiea2025171-source-pdf*

### Chapter 3, International Monetary Fund.

### Chapter 3

### Model setup and key definitions
- Households consume tradable (CTt) and non-tradable (CNt) goods; prices PNt, PTt and real exchange rate p t ≡ PNt/PTt appear in budget constraints (A.1)–(A.5).
- Lifetime income in units of the tradable good:
  - Ȳ ≡ p1 QN1 + QT1 + 1/(1 + r) (p2 QN2 + QT2) (equation A.5).
- Euler equation (intertemporal first order condition):
  - 1/CT1 = β(1 + r) E[1/CT2] (equation A.10).
- Intra-temporal allocation between tradables and non-tradables yields:
  - CN1 = (1 − ρ)/ρ · CT1 / p1 (equation A.12).
- Home period-one current account:
  - CA1 = QT1 − p1 (ρ/(1 − ρ)) QN1 (equation A.20).
- Real exchange rate relation (equation A.21):
  - p1 = p∗1 e1 (1−ρ)/1.
- Home current account expressed in terms of e1 (equation A.24):
  - CA1 = QT1 − (1/e1) (1−ρ)/1 CT∗1 (QN1 / QN∗1), with CT∗1 given by (A.25).

### Financiers, intermediation profits, and capital account
- Per-unit expected net profit from intermediation (Ω) (equation A.26):
  - Ω ≡ R − R∗ E[e2] / e1.
- Financier optimal intermediation (second-stage) (q∗i):
  - q∗i = Ω / (γi [R∗ / e1]^2 Var(e2)).
- Entry condition and indifferent financier risk-bearing ability (ˆγi):
  - ˆγi = 1/2 · Ω^2 / (F · [R∗/e1]^2 Var(e2)).
- Aggregated capital account (KA) (after aggregating financiers):
  - KA = α/(α − 1) · Ω / Var(e2) · [R∗/e1]^2 · (0.5 · Ω^2 / (F · Var(e2)·[R∗/e1]^2))^(α−1) (expression from A.2 derivation).

### Main comparative-static results (Propositions and Lemmas)
- Effect of reduced cross-border payment frictions on UIP deviation and exchange rate (Proposition 1 summary, A.3):
  - A reduction in cross-border payment frictions leads to a smaller Ω in equilibrium.
  - Reduced frictions imply a depreciation of the domestic currency e1 when starting from a CA surplus case (and analogous logic holds for CA deficits).
  - Reduced frictions increase the magnitude (“imbalances”) of CA and KA: CA surplus increases and KA deficit increases, consistent with CA1 + KA1 = 0.
- Variance decomposition identity (Lemma 4, A.4):
  - ∂Var(x)/∂(−Fric) = 2 Cov(∂x/∂(−Fric), x) (equations A.27–A.29).
  - Sign of ∂Var(x)/∂(−Fric) depends on the covariance between x and its sensitivity to frictions.
- Real shocks: exchange rate and capital flow volatility (Proposition 2, A.5)
  - Exchange rate response to a real shock s (implicit function result):
    - de1/ds = −(∂CA/∂s + ∂KA/∂s) / (∂CA/∂e1 + ∂KA/∂e1) (A.5.1 derivation; simplified in text).
  - Reduced cross-border payment frictions decrease exchange rate volatility in response to a real shock:
    - ∂/∂(−Fric) (de1/ds) has opposite sign to de1/ds, implying dampening of de1/ds when frictions decline (A.5.1).
  - Reduced cross-border payment frictions increase capital flow volatility in response to a real shock:
    - dCA/ds and ∂/∂(−Fric)(dCA/ds) have the same sign, implying amplification of CA responsiveness when frictions decline (A.5.2).
- Financial shocks: exchange rate and capital flow volatility (Proposition 3, A.6)
  - Reduced cross-border payment frictions increase exchange rate volatility in response to a financial shock (shock to R∗):
    - de1/dR∗ and ∂/∂(−Fric)(de1/dR∗) have the same sign; reduced frictions amplify exchange-rate sensitivity to R∗ shocks (A.6.1).
    - Key intermediate condition used (Lemma 5) requires equality of mixed partials for KA, which the model verifies for both F and α channels.
  - Effect on capital flow volatility in response to a financial shock is ambiguous:
    - ∂/∂(−Fric)(dCA/dR∗) can be either positive or negative depending on equilibrium CA and parameter values; reduced frictions can amplify capital-flow volatility when CA is near 0 and dampen it when CA is large (A.6.2).

### Parameter conditions and sufficient restrictions (Appendix B)
- Condition for capital account to react in line with UIP deviation (B.1):
  - sign(∂KA/∂x) = sign(∂Ω/∂x) for x = R, R∗, e1 holds if (after algebra) 0.5 · R∗/e1 > |Ω|.
  - Interpretation: UIP deviation Ω must be less than half of expected gross yield on foreign-currency bond holdings for the sign alignment to hold.
- Reduced frictions amplify capital-account responsiveness (B.2):
  - For the F channel (fixed cost), sign(∂KA/∂x) = sign(−∂KA/∂x∂F) holds generally because KA is multiplicative in h(F) := (1/F)^(α−1).
  - For the α channel (risk heterogeneity), comparable sign conditions hold under parameter restrictions that are not materially stronger than the 0.5 · R∗/e1 > |Ω| condition; a sufficient illustrative condition is that less than approximately 37% of potential financiers enter (β < 1/e), which implies −log(β) > 1 and simplifies the inequality.

### Robustness and calibration notes (Appendix C)
- Results are qualitatively robust to allowing non-negligible intermediary profits to enter the Foreign household’s budget constraint (Figure C.1 discussion).
- Calibration values used for illustrative figures (Table C.1):
  - QT1 = 2
  - QN1 = 1
  - R = 1.02
  - R∗ = 1.025
  - Var(e2) = 2
  - ρ = 0.35
  - p∗1 = 1
  - ∆QT1 = −0.01 QT1 = −0.02
  - ∆R∗ = 0.025
- Authors note these parameter choices are illustrative given model stylization and magnitudes should be interpreted as qualitative.

### Key takeaways for analysis and policy design
- Reducing cross-border payment frictions:
  - Lowers the equilibrium UIP deviation Ω and alters equilibrium exchange rate (tends to depreciate home currency in CA surplus examples).
  - Dampens exchange-rate volatility to real shocks but amplifies capital-flow (current-account) volatility to real shocks.
  - Amplifies exchange-rate volatility to financial shocks (foreign interest-rate shocks), while effect on capital-flow volatility to financial shocks is ambiguous and depends on initial CA and parameter values.
- Mechanism highlights:
  - Intermediaries’ risk-bearing capacity, entry costs (F), and heterogeneity (α) shape how KA responds to shocks and frictions.
  - The covariance between a variable and its sensitivity to frictions determines whether volatility of that variable rises or falls as frictions decline (Lemma 4 identity).
- Practical implications:
  - Policies aimed at lowering cross-border payment frictions can improve some dimensions of stability (smaller UIP deviations; lower exchange-rate sensitivity to real shocks) but may increase capital-flow volatility in key scenarios and amplify exchange-rate sensitivity to financial shocks.
  - Calibration and monitoring of intermediary entry, risk-bearing capacity, and the magnitude of UIP deviations (relative to R∗/e1) are important to assess net macro-financial effects of reducing frictions.

*Chapter 3, International Monetary Fund.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025171-source-pdf.pdf_