The flow of stablecoins and the spillover effects in the foreign exchange market

The rapid expansion of the stablecoin market has made it an important participant in the global financial system. The vast majority of stablecoins are pegged to the US dollar, but it is noteworthy that over 70% of the cumulative net inflows from fiat currencies to stablecoins come from non-US dollar currencies. This means that most stablecoin transactions inherently involve foreign exchange conversions relative to the US dollar, thus forming a parallel, stablecoin-based foreign exchange ecosystem. This phenomenon raises two key questions: how deeply is this crypto-based foreign exchange market connected to the traditional foreign exchange market? What kind of spillover effects will it have on the traditional foreign exchange market and global capital flows? This paper systematically provides evidence of causal spillover effects from the stablecoin market to the foreign exchange market using daily data from four US dollar-pegged stablecoins (USDT, USDC, DAI, BUSD) against 27 fiat currencies (from January 2021 to November 2025). The study finds that an exogenous increase in net inflows of stablecoins significantly widens the price deviations between stablecoins and traditional foreign exchange, leads to depreciation of local currencies, and worsens synthetic dollar financing conditions (i.e., increases the dollar premium). These findings suggest that stablecoins have become an emerging component of the global currency market, directly impacting financial stability. The Research Institute of Financial Technology at Renmin University of China (WeChat ID: ruc_fintech) has compiled the core parts of the study. Authors | Iñaki Aldasoro, Paula Beltran, Federico Grinberg

Compilation | Chen Yazhou

1. Data and Core Definitions

The data for this paper comes from 64 centralized exchanges (including Binance, Coinbase, Kraken, etc.). For each stablecoin-fiat currency pair, daily closing prices, trading volumes, and inflows from fiat currencies to stablecoins and reverse outflows are collected. To construct currency-level indicators, a volume-weighted average price is used, and the traffic from all exchanges is aggregated.

Core Definition 1: Price Deviation

A subject wants to purchase stablecoins with local currency through two routes: a direct route (buying stablecoins with local currency on a centralized exchange) and an indirect route (first converting local currency to US dollars in the spot foreign exchange market, then using US dollars to buy stablecoins). In a frictionless complete market, the costs of both routes should be equal. Price deviation is defined as the direct price divided by the indirect price multiplied by 100%. When this ratio is greater than 100%, direct purchase is more expensive, indicating an arbitrage opportunity (buying stablecoins with US dollars and then converting back to local currency for sale); and vice versa. The further this ratio deviates from 100%, the greater the arbitrage friction.

Core Definition 2: Net Inflow Rate

The net inflow rate is defined as: inflows from fiat currency to stablecoins minus reverse outflows, divided by the outstanding stock of the currency-stablecoin pair from the previous period. This indicator measures the new net funds entering the stablecoin ecosystem from a specific fiat currency (as a percentage of the stock), reflecting the net demand pressure for stablecoins. A positive value indicates net funds flowing from fiat currency to stablecoins.

Core Definition 3: Covered Interest Rate Parity Deviation

Covered interest rate parity deviation is a core indicator for measuring the cost of dollar financing in the traditional foreign exchange market. It equals (forward exchange rate minus spot exchange rate) divided by the spot exchange rate, minus the interest rate differential between the local currency and the US dollar. When this deviation is negative, it indicates that synthetic dollar financing is more expensive than directly borrowing US dollars, meaning there is a "dollar premium." The larger the negative value, the higher the cost of dollar financing. This is a classic indicator for measuring financing frictions in the traditional foreign exchange market.

2. Three Key Facts

Fact 1: There is a significant and heterogeneous price deviation between stablecoins and the traditional foreign exchange market. In the sample, major high-liquidity currencies (such as the US dollar, euro, and British pound) have smaller price deviations (averaging about 0.05%-0.3%) and are closely clustered around 100%, indicating relatively effective arbitrage. However, for economies facing high inflation, exchange rate volatility, or capital controls (such as the Argentine peso, Nigerian naira, and Turkish lira), the average price deviation can reach several percentage points (for example, the average deviation of the Argentine peso is about 1.06%, with a maximum of 14.8%), and the volatility is severe. This fact indicates that there are persistent arbitrage frictions between crypto and traditional markets, and the size of these frictions is positively correlated with macroeconomic vulnerability.

Fact 2: Net inflows of stablecoins are highly correlated with price deviations, local currency depreciation, and the widening of the dollar premium. Through linear projection regression (tracking the impact from 0 to 15 days after the shock), it is found that an increase in net inflows of stablecoins immediately and significantly raises the price deviation (increasing on the day of the shock); subsequently leading to depreciation of local currencies in the traditional spot market (peaking 1-3 days later); while causing a decrease in the short-term (3 months) covered interest rate parity deviation (i.e., increasing the dollar premium), with no significant impact on the 12-month covered interest rate parity. This indicates that the balance sheet constraints of arbitrageurs are tighter over shorter time horizons. These correlations constitute preliminary evidence of the impact of stablecoins on traditional markets, but cannot yet be interpreted as causal.

Fact 3: The above correlations cannot be directly interpreted as causal relationships. There are obvious reverse causality and omitted variable issues. For example, the expected depreciation of the local currency may simultaneously drive capital flight into stablecoins (increasing net inflows) and lead to the depreciation itself. Changes in global risk sentiment (such as an increase in the VIX index), local regulatory shocks, and other omitted factors may also simultaneously affect the demand for stablecoins and traditional exchange rates. Therefore, a more rigorous causal identification strategy is needed.

3. Constrained Arbitrage Model

To establish a causal identification framework, this paper constructs a theoretical model. The core settings are as follows:

1. Domestic residents demand two types of assets simultaneously: stablecoins (for on-chain access to US dollars) and synthetic dollars obtained through foreign exchange swaps (for hedging and trade financing). The two are imperfect substitutes, with demand changing linearly with stablecoin premiums and dollar premiums. When the stablecoin premium rises, demand for stablecoins decreases; when the dollar premium widens (making the covered interest rate parity more negative), demand for synthetic dollars decreases.

2. Stablecoin issuers provide fully elastic supply at the US dollar peg, meaning the dollar price of stablecoins is always $1.

3. US investors supply dollars to the swap market, with supply increasing as the dollar premium widens (the higher the premium, the more supply).

4. A representative intermediary is active in both the stablecoin market and the swap market, with its cost function containing three components: swap-specific costs (proportional to the square of the swap position), stablecoin-specific costs (proportional to the square of the stablecoin position), and cross-market costs (proportional to the square of the sum of the two positions). The cross-market cost parameter is key, as it indicates the additional marginal cost incurred by the intermediary from holding total local currency exposure. When this parameter is greater than zero, expanding the stablecoin position increases the marginal cost of swap provision, thus creating spillover from the stablecoin market to the covered interest rate parity.

The model's equilibrium solution shows:

• A shock to stablecoin demand will increase the stablecoin price deviation (self-effect).

• If the cross-market cost parameter is greater than zero, it will also simultaneously increase the covered interest rate parity deviation (i.e., causing the dollar premium to rise, cross-market spillover).

• Through substitution effects (when stablecoins become more expensive or the dollar premium rises, residents turn to spot purchases of US dollars), local currencies depreciate.

When the cross-market cost parameter is zero, the two markets are completely decoupled, and shocks to stablecoins do not affect covered interest rate parity or the dollar premium.

In a multi-country extension, the model introduces "cross-book participants"—traders active in multiple fiat currency-stablecoin trading books, who face a total conversion budget constraint (the sum of conversion amounts across all currencies is fixed). When the urgency to convert in country A rises, participants withdraw from country B, leading to a decrease in net inflows in country B, thus affecting prices and exchange rates in country B. This cross-country connection provides a theoretical basis for instrumental variables in causal identification.

4. Causal Identification Strategy and Main Results

4.1 Identification Challenges

Ordinary least squares regression suffers from severe bias because local stablecoin demand shocks are highly correlated with other factors affecting local exchange rates (such as expected depreciation, changes in local interest rates, and pressures on the banking system). A source of exogenous variation is needed.

4.2 Fine Instrument Variable Construction

Step 1: Remove global common factors (such as global risk appetite, Bitcoin price trends, changes in the US dollar index, etc.) from the net inflow rate using a factor model to obtain heterogeneous components. This step is equivalent to subtracting all currency common fluctuations from the net inflow of each currency-stablecoin pair.

Step 2: For a target currency and stablecoin, the instrument variable is the size-weighted sum of heterogeneous shocks from all other currencies, with weights based on each currency's market share in that stablecoin market (e.g., trading volume share). Since the target currency's own shocks are excluded, this instrument variable is uncorrelated with local factors of the target currency (such as local monetary policy and changes in local capital controls); at the same time, positive shocks from other countries will attract conversion activities away from the target currency through the budget constraints of cross-book participants, thus reducing net inflows of the target currency, satisfying the relevance condition.

4.3 Two-Stage Least Squares Estimation Results

The F-statistic of the first-stage regression is far above conventional thresholds (over 500), indicating that the instrument variable is strong.

Second-stage estimation results (shock response, for every 1% increase in stablecoin net inflow):

• Price Deviation: increases by about 40 basis points, with the effect lasting about 10 days after the shock.

• Spot Exchange Rate (local currency): depreciates by about 5 basis points, with the effect diminishing after a few days.

• 3-Month Covered Interest Rate Parity Deviation: decreases by about 5 to 10 basis points (i.e., dollar premium increases), with significant effects concentrated in the short term.

• 12-Month Covered Interest Rate Parity Deviation: no significant change.

Key ratio: the absolute value of the covered interest rate parity coefficient divided by the price deviation coefficient is about 0.44, indicating that about 44% of the stablecoin price pressure is transmitted to the traditional dollar financing market. This intermediate value between 0 (complete separation) and 1 (complete integration) indicates a significant but not complete connection between the stablecoin market and traditional markets.

5. Counterfactual Analysis and Dynamic Amplification

5.1 Cross-Market Costs as Key Drivers

Through model calibration (using literature and the above instrumental variable estimates), counterfactual analysis shows:

• Doubling the cross-market cost parameter nearly doubles the spillover of covered interest rate parity (from -6 basis points to -12 basis points), with exchange rate depreciation increasing from 6 basis points to 9 basis points.

• If the cross-market cost parameter approaches zero (complete separation), the spillover of covered interest rate parity disappears, and exchange rate depreciation is halved.

• Stablecoin-specific costs mainly affect the price deviation itself, with limited impact on the spillover of covered interest rate parity.

Policy Implications: In economies with open capital accounts, intermediaries can hedge local currency exposure through international operations, effectively reducing cross-market costs; whereas in economies with strict capital controls, regulations amplify cross-market costs by preventing the international diversification of intermediaries' balance sheets, thereby enhancing spillover effects.

5.2 Dynamic Extensions and State Dependence

The paper further constructs a dynamic model, introducing:

• The persistence of stablecoin demand shocks (set with an autoregressive coefficient of 0.8, meaning the half-life of the shock is about 3 days)

• The impact of intermediary wealth on risk-bearing capacity (costs divided by current wealth; the less wealth, the higher the cost per unit position)

Simulation results show:

• Shocks lead to a reduction in intermediary wealth of about 5%, lowering risk-bearing capacity and amplifying subsequent price responses.

• When intermediary initial wealth is below 50% of the steady-state level, the effects of the same shock double.

• Cumulative impulse responses indicate that static analyses may underestimate the true spillover costs by 5-6 times.

5.3 Redemption Frictions and Amplification of Runs

When introducing a redemption discount mechanism (during large redemptions, due to fire-sale and settlement frictions, each unit of stablecoin can only be redeemed for less than $1), liquidity shocks and friction shocks occurring simultaneously (such as the Terra/Luna collapse or the Silicon Valley Bank incident triggering USDC depegging) can cause spillover effects to reach five times the baseline. This explains why stablecoin stress events can have extraordinary impacts on traditional markets.

6. Conclusion and Policy Implications

This paper systematically demonstrates for the first time the existence of causal spillover relationships between the stablecoin market and the traditional foreign exchange market. The main conclusions are as follows:

1. An exogenous increase of 1% in net inflows of stablecoins will expand the stablecoin-fiat price deviation by about 40 basis points, lead to a depreciation of the local currency by about 5 basis points, and increase the short-term dollar premium by 5-10 basis points.

2. About 44% of the stablecoin price pressure is transmitted to the traditional dollar financing market, indicating a significant but not complete connection between the two markets.

3. Cross-market frictions are the main determinants of the size of spillovers, and capital controls amplify these frictions.

4. Spillover effects are state-dependent: when intermediary capital is exhausted, the same shock will have a greater impact; when redemption frictions and liquidity shocks overlap, they can amplify several times.

Policy Recommendations:

1. Prudent requirements for stablecoin intermediaries: capital buffers, reserve liquidity requirements, and limits on concentrated currency exposures can reduce spillover channels from the source.

2. Incorporate into macroprudential regulation: policymakers (especially in emerging markets) should include monitoring of the stablecoin market within the framework of foreign exchange and capital flow management, with changes in spillover ratios serving as early warning indicators of increased interconnectedness. As the stablecoin market continues to expand and mature, similar liquidity shocks may produce larger price responses in the future, and spillover risks may increase with market development. This finding has profound implications for understanding capital flows and financial stability in the digital age.