Reverse Yankee bonds
Published as part of the The international role of the euro, June 2025.
US firms face different borrowing costs depending on whether they issue bonds in US dollars or in euro. Bonds issued by US firms in a foreign currency are known as “Reverse Yankees” in market parlance. Historically, the share of these bonds denominated in euro relative to those denominated in US dollars has been tightly correlated with differences in medium to long-term expectations about US and euro area policy rates, as measured by ten-year overnight index swap (OIS) rates (Chart A, panel a). The OIS differential widened in 2024, making the issuance of bonds in euro more attractive relative to US dollars. In turn, euro-denominated Reverse Yankee bonds as a share of total issuance increased.
In addition to risk-free rate differentials, borrowing costs depend on the relative credit spread for bond issuers in the two currencies and on the cost of hedging foreign exchange (FX) risk. The premium paid over risk-free rates by US companies may differ between euro area and US markets owing to differences in the investor base and in the perceived risks between the two markets. Moreover, when firms issue liabilities in a foreign currency, they are exposed to exchange rate risk. For firms with assets and operations in the corresponding foreign currency, this risk may be more limited. Other firms may hedge their positions using the FX market or the basis swap market. Accordingly, this box follows the methodology developed in recent research and builds two measures of borrowing costs: one with FX risk hedging (yellow line in Chart A, panel b) and the other without (blue line).[1],[2]
Chart A
Evolution of US corporate bond issuance and cost of issuing bonds in euro for US firms
a) Share of euro-denominated bonds issued by US firms and expected policy rate differentials between the United States and the euro area | b) Unhedged and hedged euro-US dollar borrowing cost differentials |
|---|---|
(percentages; basis points) | (basis points) |
 |  |
Sources: Dealogic, Moody’s Analytics, Bloomberg and ECB staff calculations.
Notes: In panel a), the “Reverse Yankee issuance share” is calculated as the share of euro-denominated bonds issued by investment-grade US firms divided by their total issuance volume in all currencies; "Euro-US dollar risk-free rates” refers to the respective ten-year overnight index swap rates, which measure market-implied expectations of future policy paths. In panel b), the computation of residual credit spread differentials follows Liao (2020) and Caramichael et al. (2021) and uses data on investment-grade and senior unsecured US dollar and euro-denominated bonds (with outstanding amounts larger than USD 50 million) issued by US companies. To assess the impact of currency denomination on the pricing of credit risk, a separate regression for every cross-section is estimated as: , where is the option-adjusted corporate bond spread, is the currency effect, are the firm fixed effects, are maturity bucket fixed effects and are rating fixed effects. The residual credit spread is represented by the estimated coefficient . “Unhedged borrowing cost differential” is the sum of the residual credit spread and the OIS rate differential. “Hedged borrowing cost differential” is the sum of the residual credit spread and the covered interest parity (CIP) deviation (or cross-currency basis). Ten-year risk-free rates and the forward premium are used for computing interest rate differentials and CIP deviations to match the average maturity of the bonds in the sample. The data reported in the chart is monthly, using the first available observation each month; all underlying variables are daily except for credit spreads, which are weekly. The total number of observations is 892,280 and the total number of bonds is 6,212. Some residual credit spread data for 2020 are missing and therefore interpolated.
Borrowing cost differentials without FX risk hedging are measured as the yield differential between bonds denominated in euro and in US dollars. This is determined by (i) the difference between risk-free rates in the United States and the euro area, and (ii) the difference in corporate bond spreads paid by US firms in US dollars and euro, respectively, over the respective risk-free rates. To derive the latter, spreads of individual US firms that issue corporate bonds in both currencies are regressed on currency, firm, maturity bucket and rating fixed effects for each monthly cross-section. The estimated time-varying currency coefficient is the residual credit spread, which indicates how much a US firm would save or overpay, relative to risk-free rates, when issuing in euro instead of US dollars. This spread is added to the risk-free rate differential between the euro and the US dollar to derive the relative unhedged borrowing cost. The unhedged borrowing cost differential has historically been negative, thus making borrowing in euro more attractive, all else being equal (blue line in Chart A, panel b). The combination of declining euro risk-free rates and broadly stable residual credit spreads translated into lower costs for borrowing in euro without hedging against FX risk in 2024.
If a company chooses instead to hedge against FX risk, other components need to be added. The borrowing cost differential between the US dollar and the euro is then the residual credit spread plus any deviation from covered interest parity (CIP), also known as the cross-currency basis, which represents the extra cost to be borne by the issuing firm when it hedges euro liabilities into US dollars.[3] It can be thought of as the unhedged borrowing cost differential plus the forward premium. Assuming that the FX risk is fully hedged until the bond’s maturity, this hedged relative borrowing cost indicator (yellow line in Chart A, panel b) is slightly positive, i.e. unfavourable for the euro, suggesting that, on average, US firms do not have a financial incentive to borrow US dollars synthetically through the euro market.[4] In fact, the hedged borrowing cost differential has recently been more favourable in absolute terms for US dollar issuance because of compressed corporate bond spreads in the United States and, to a lesser extent, because of CIP deviations.
Local projections suggest that declining hedged and unhedged borrowing costs in euro both provide an incentive for US firms to issue more Reverse Yankee bonds. On average, a 10 basis point reduction in the unhedged borrowing cost differential is associated with an increase of approximately 0.5 percentage points in the relative share of euro-denominated corporate bond issuance by US firms (Chart B, panel a). The effect is largest in the first two months after a shock. Using exchange rate-hedged borrowing cost differentials, a similar pattern emerges – lower hedged borrowing costs in euro also appear to affect the choice of issuance currency (Chart B, panel b). While the sensitivity of issuance to a 10 basis point reduction in hedged and unhedged borrowing costs is similar, the variability of unhedged borrowing costs is relatively stronger and therefore a more powerful driver.[5] In fact, univariate regressions on data since 2017 suggest that variations in unhedged borrowing costs explain up to 40% of the variations in the share of euro-denominated issuance by US firms over a six-month horizon, whereas hedged borrowing costs explain less than 20%. By components, changes in interest rate differentials are the main drivers, explaining up to 30% of the variation, with changes in residual credit spreads accounting for 17% and CIP deviations for a mere 4%.
Chart B
More attractive borrowing costs in euro are associated with stronger issuance in euro by US firms
a) Response of euro issuance share to lower unhedged borrowing costs in euro vs US dollars | b) Response of euro issuance share to lower hedged borrowing costs in euro vs US dollars |
|---|---|
(percentage points) | (percentage points) |
 |  |
Sources: Dealogic, Moody’s Analytics, Bloomberg and ECB staff calculations.
Notes: The charts show the cumulated response of the share of investment-grade bond issuance in euro versus US dollars by US firms over 12 months after a 10 basis point change in either hedged or unhedged borrowing costs, which makes borrowing in euro relatively cheaper. Local projections control for a time trend, a measure of mergers and acquisitions involving European companies compiled by Bloomberg and aggregated monthly, the 30-day average of the CBOE VIX volatility index, a COVID-19 period dummy from March to June 2020, five lags of the dependent variable and one lag of the shocked variable. Unhedged borrowing cost differentials are computed as the sum of the residual credit spread and the risk-free rates differential. Hedged borrowing cost differentials are computed as the sum of the residual credit spread and CIP deviations. Ten-year maturities for risk-free rates and the forward premium are used for computations to match the average maturity of the bonds in the sample. Monthly data are from January 2017.
The latest observations are for the end of 2024.
Overall, the findings presented in this box highlight the crucial role that changes in expected borrowing costs play as determinants of the euro’s appeal as an international funding currency. The analysis shows that expectations for future policy rate paths in the euro area and the United States, the relative pricing of corporate credit risk in the two markets and frictions in cross-currency markets all influence US firms’ decisions to issue in euro, albeit to different extents. This box shows that, in recent years, changes in risk-free rate differentials have been the main driver reflecting shifts in expectations for relative policy rates. Meanwhile, the estimated relative credit spreads and CIP deviations, which also impact borrowing costs, have remained more stable.
The analysis follows Liao, G.Y., “Credit migration and covered interest rate parity”, Journal of Financial Economics, Vol. 138, 2020, pp. 504-525 and Caramichael, J., Gopinath, G. and Liao, G.Y., “U.S. Dollar Currency Premium in Corporate Bonds”, IMF Working Paper, No 185, 2021. For further details of the estimation, see the notes to Chart A and the accompanying text.
The unhedged and hedged bond yield differentials between euro and US dollar-denominated bonds are formally defined as follows:
where is the risky bond yield, is the residualised credit spread, is the risk-free rate, and are the forward and spot exchange rates, and is the CIP deviation.From an investor demand perspective, CIP fluctuations can also drive euro area investors (especially those more exposed to exchange rate rollover risk) away from US dollar-denominated bonds towards bonds denominated in euro. See Kubitza, C., Sigaux, J-D. and Vandeweyer, Q., “The Implications of CIP Deviations for International Capital Flows”, Chicago Booth Research Paper, No 24-18, 2024.
US firms may still decide to issue in euro and hedge, for example, to cater to a broader international investor base or by using more efficient hedging strategies.
In particular, the unhedged borrowing cost differential decreased by around 50 basis points in 2024, while the euro-US dollar residual credit spread widened only marginally (6 basis points). The hedged differential, while still more favourable for US dollar issuance, decreased by only 10 basis points.