Currency Risk in Long-Term International Investment
Long-term international investment has one inadvertent consequence – the creation of currency risk. Black (2002) defines currency risk as “the risk that changes in exchange rates will affect the profitability of any activity between the time when one is committed to it and the time when it is carried out. This affects foreign trade, foreign lending, and foreign direct investment. Commitment may arise from a contract, as in export sales or foreign currency loans, or from incurring sunk costs, as in foreign direct investment or setting up foreign distribution systems for exports.
It is possible to reduce currency risk by use of forward currency markets, at a cost, but only for relatively short time periods. Currency risk causes the local currency value of the foreign receivables or investments to fluctuate dramatically because of pure currency movements. Typically, the academic literature on currencies has misunderstood currency and suggested that currencies have no long term return, difficult to predict, and difficult to take advantage of for the reason that markets are extremely liquid.
Therefore, typical recommendations include either that companies and investors should remove this uncompensated volatility through naively hedging back into the base currency or leaving the risk uncompensated. The latter recommendation is often misinterpreted and consequently left unmanaged. The effective financial management of such cash flows or investments provides a completely different perspective as naive hedging/unhedging of currency risk implies a strong view that the base currency will appreciate/depreciate against the foreign currency.
Furthermore, the currency market has many non-profit participants and while exact currency levels cannot be predicted, the future direction of currencies can be anticipated Exposure of International Corporate Bond Returns to Exchange Rate Risk This part of the paper looks at the effects of exchange rate variability on corporate bond returns, trying to ascertain the extent to which currency risk impacts the borrowing costs of companies. The relevance of this study becomes apparent in the context of European Monetary Union.
In fact, one of the benefits of currency unification that has been touted is the reduction of financing costs to European companies, from decreasing exchange rate risk. This chapter examines the likelihood and extent of these potential gains. Exchange rate changes may be correlated with corporate bond returns for two reasons: first, because the company’s expected future cash flows may depend on exchange rates; secondly, because exchange rates may affect the required market discount rates on these cash flows.
This change in the required discount rate can arise from changes in default-free interest rates or from changes in credit risk premia. This section tries to isolate the changes in default-free interest rates in corporate bond returns and concentrate on the other sources of returns. With that purpose, studying credit returns is vital, that is the returns on corporate bonds above the returns on risk-free bonds with the same promised payments. Credit returns are particularly interesting owing to the option-like pay-off of corporate bonds.
Since they just depend on changes in the solvency of the issuing company, credit returns are very sensitive to events that affect the future prospects of the company. Even more so than equity returns. The riskiness of corporate credit returns has both idiosyncratic and systematic sources. However, only the latter may be priced and are thus of interest to us. Finding no covariation of credit spreads with such factors would be evidence that they are not priced. However, finding some covariation still leaves the question of whether the risk source is priced.
It is thus important to examine potential risk factors such as stock market returns, default-free interest rate changes and exchange rate changes, with a special interest in this last factor, since it is the one that will be affected by monetary union. The significance of exchange rate risk in explaining asset returns is an open, and insufficiently explored, question in the empirical asset pricing literature. Most papers have concentrated on stocks and default-free bonds, with mixed conclusions. Jorion (1991) finds little evidence of priced exchange rate risk in US stock returns, whereas Chow et al.
(1997) do find evidence of exchange rate changes affecting the returns of US stocks and bonds, including long-term corporate bonds. This part of the article adds to the literature by looking at a different asset class. Additionally, the sample extends coverage to include bonds denominated in several currencies, and is thus informative about the effects of currency risk in a wide variety of firms. Prior attempts at identifying the systematic factors of corporate bonds have primarily used portfolios of US dollar-denominated bonds organized on credit ratings.
Chang and Huang (1990) find evidence for a two-factor latent-variable model using interest rate variables as the instruments. Fama and French (1989) find that future excess returns on stock and bond portfolios can be forecast by default spreads, term spreads, and dividend yields. These studies examine bonds’ excess returns, that is, the return on a corporate bond less the risk-free interest rate for the holding period. This excess return is affected by two forces: movements in the riskless interest rate and changes in default premia. It is also critical to focus our analysis on credit returns in order to isolate changes in default premia.
Studies that have looked at credit returns explicitly are few in number. Sarig and Warga (1991) quantify the magnitude of credit returns for zero-coupon US corporate bonds while Litterman and Iben (1991) show how to extend the analysis to include coupon bonds. These studies do not address the dynamics of credit returns or try to identify the systematic factors that affect them. Changes in exchange rates will affect asset returns in as much as their cash flows are not completely hedged within the firm. The effectiveness of exchange rate hedging is a function of the uncertainty and duration of future cash flows.
Long-term cash flows will be harder to hedge than short-term cash flows owing to their increased uncertainty. Looking at short-horizon returns may understate exchange rate risk if firms can hedge their short-term cash flows effectively. But for future cash flows, where the long-term effects of exchange rate changes are difficult to ascertain, hedging effectiveness is doubtful. This is the motivation of Chow et al. (1997) in looking at long-horizon returns to assess whether exchange rate risk is priced. It is therefore important to examine the exchange rate dependence of credit returns at different horizons.
The Risk Exposures of Corporate Bond Returns This part of the paper discusses the factors that affect returns to corporate bonds. Previous research on corporate bonds has focused on explaining excess returns over short-term default-free interest rates. Unfortunately, this approach mixes the two effects of term-structure movements and default-spread movements. To examine changes in default spreads independently of movements in the term structure, we form credit returns, by taking the return to a corporate bond and subtracting the return to a riskless bond with the same promised payments.
It is believed that this definition is more representative of actual movements in default spreads for corporations (Dermine and Hillion 1999). Corporate bonds are contingent claims on the value of the firms that issue them. The bond only receives the promised payments (coupon and principal) if the firm does not default. A simple model can be use to illustrate the determinants of corporate bond returns. This article assumes that default happens when the value of the assets of the firm, V, falls below the value of its liabilities, K, with both processes assumed to follow diffusions.
Default can then be modelled as the first time X = log (V/K) passes through 0. The dynamics of X can in general depend on any state variables that affect the assets or the liabilities of the firm. Further assumptions that the write-down in case of default is a constant fraction, W, of the price of a default-riskless bond with the same promised payments, the time t price of a zero-coupon corporate bond with maturity T has been shown by SaaRequejo and Santa-Clara (1997) to be: C(t,T) = P(t,T) (1 – WQ(t, T)), (1)
where C(t, T) denotes the corporate bond price, P(t, T) the default-riskless price and Q(t, T) the forward risk adjusted probability that X win fall below 0 between dates t and T, that is, the probability of default before T. The (continuously compounded) return to corporate zero-coupon bonds between dates t and s is given by Therefore: r c (t, s, T) = r p (t, s, T) + r x (t, s, T), (3) where the return to the corporate bond is decomposed into the return to a default- riskless zero-coupon bond with the same maturity and the credit return.
As illustrated, the credit return is due to changes in the forward risk adjusted probability of default. This decomposition holds approximately with coupon bonds. The return to a coupon corporate bond can be decomposed into the return to a default-riskless zero-coupon bond with the same coupon rate and principal and the credit return. The credit return depends on the evolution of the state variable X, which determines the probability of default. The dynamics of X depend on any systematic factors and idiosyncratic shocks that affect the assets and liabilities of the firm, which are left unspecified by the model.
Furthermore, the form of the dependence of credit returns on the risk factors is a priori unknown, since the forward risk adjusted probability of default depends on X generally in a nonlinear way, and the solvency ratio itself may vary nonlinearly with the factors. The purpose of this section is to investigate and try to determine what systematic factors affect credit returns on corporate bonds. The form of the dependence to linearity, which can of course be justified as a first-order approximation, has been restricted.
The variables which are use to explain credit spreads are stock market returns, changes in interest rates and changes in exchange rates. These variables have been studied before in other asset pricing studies. The return on the stock market is a standard explanatory variable in asset pricing models. This variable should have explanatory power for the variation in the value of the assets and the liabilities of the firm, and hence for changes in the probability of default and credit spreads. Corporate bond returns are driven by changes in risk-free interest rates and changes in default premia.
Campbell (1987) and others have shown that expected excess returns on stock and bond portfolios are related to changes in interest rates. Chang and Huang (1990) use interest rates as instruments for time-varying risk premia of corporate bond returns while Fama and French (1989) find evidence that the term premium (riskless long-term rate less the short-term rate) is related to excess returns of corporate bonds. The effect of interest rates on credit returns is not obvious because term-structure movements are removed with the matching bond.
Additionally, Saa-Requejo and Santa-Clara (1997) show that the forward risk adjusted probability of default depends on the volatilities of default-riskless interest rates and that these volatilities may depend on the interest rates themselves. Therefore changes in the probability of default should be related to changes in the yield curve. There is residual interest in interest rate changes and stock market returns. This paper mainly interests in examining the exposure of corporate bond returns to exchange rate risk.
Basically, it endevours to ‘filter out’ the effects of the former variables in order to assess the true impact of exchange rate changes. Exchange rate changes will affect returns if the asset’s cash flows are not completely hedged. Given the possibility of default, perfect currency hedging of corporate bonds is not possible. Whether the residual exchange risk can be diversified away or not is an open question. Chow et al. (1997) argue that the lack of empirical evidence for priced exchange rate risk may arise from holding periods being too short.
They claim that if exchange rate risk is priced, it should show up in long-term returns. The rationalization for the existence of long-horizon exchange rate exposure is that short-term cash flows can be effectively hedged whereas the effect of long-term exchange rate exposure on future cash flows is uncertain, making hedging less effective. If this conjecture is true, then it is expect that the significance of exchange rate exposure will increase with the holding period length. Source of the Data
The data for this study are provided by Datastream, which provides extensive coverage of financial and economic variables for a large cross-section of countries. The interest is in the more developed markets, where liquidity is higher and pricing is more reliable. For this reason, it specifically concentrates on bonds denominated in the currencies of France, Germany, Switzerland, and Great Britain, as well as the United States and Canada. These currencies provide the bulk of the bonds tracked and produce a sufficient sample size for each currency.
The above currencies correspond to the major Eurobond markets. Datastream tracks bonds that are alive as of the current day. It is not a historical database with prices for matured or defaulted issues, so the issue of sample selection bias may be important. The bias, if any, is mitigated by the fact that both bonds which have defaulted and bonds which have matured and paid off are not included in the present sample. These offsetting forces make the source of bias hard to sign and presumably insignificant. Corporate bond data
Corporate bond data are obtained from Datastream’s bond database. The database includes a price history for the individual issues and their associated characteristics such as coupon rate, derivative features, and time to maturity. The study starts with all bonds in the database and eliminate bonds which are issued by the government, public authorities, supranationals, or any other public entity. The remaining bonds have been issued by private companies subject to credit risk. A complication that arises from using corporate bonds is the presence of embedded options.
Embedded options change the return decomposition from the model shown in the previous section. Therefore, the study eliminates callable bonds, bonds with warrants, convertible bonds, putable bonds, and bonds with sinking fund provisions. This restriction avoids evaluating the embedded options, which would be difficult given their variety and complexity. It is well known that corporate bonds do not trade very often and market prices are observed infrequently. In the absence of market prices, quoted bond prices come from either dealers’ bid-ask sheets or from matrix pricing of similar bonds.
Datastream claims to use a sale price or a dealer bid-ask price for its quoted bond prices, although differentiation between market price and dealer bid-ask price is not provided. Conclusion This article looks at the risk exposures of credit returns in the international corporate bond market as a particular way of studying long-term international investment. Corporate bond returns are affected by riskless interest rate and default factors. Credit return is defined as the return on a corporate bond less the return on an equivalent default-free bond.
The credit return is interpreted as compensation for bearing default risk. One potential factor affecting credit returns is exchange rates, both in their potential effects on companies’ cash flows and in the market required rate of return. This factor gains importance in light of the European Monetary Union process, since monetary union would eliminate exchange rate risk. If a risk premium does exist for exchange rate risk, than a benefit of monetary union would be the potential elimination of the associated risk premium in companies’ borrowing costs.
The examination as to whether credit returns are related to foreign exchange changes after controlling for other sources of systematic risk shows an interesting result. The relationship between credit returns and the different risk factors is explored at different holding-period horizons. Initially, the systematic factors that affect credit returns have been examined. The factors chosen are related to interest rates, stock market returns, and exchange rates. A linear pricing relation between credit returns and the above factors have been assumed.
Regressions are performed to determine which factors are significantly related to credit returns. The resulting factor loadings are inputs to an unconditional multivariate asset-pricing model. The results show a significant negative relationship between credit returns and the term spread. The slope of the term structure is found to be significant at the longest horizon while the excess stock market factor is weakly significant across currencies and horizons. The results show that credit returns are affected by interest rates and to a lesser extent by excess stock returns.
Therefore, credit returns of corporate bonds are not related to exchange rates at short to medium horizons. This result is consistent across currencies. The results for the longest horizon, one year, show evidence of a significant relationship between exchange rates and credit returns. This is checked to see whether this result is robust to the definition of the exchange rate factor. After analyzing two alternative exchange rate factors, similar results have been found to the trade-weighted index factor.
The conclusion drawn is that there is a negligible relationship between credit returns and exchange rates for short or medium horizons. Only at the longest horizon studied is there some evidence of exchange rates affecting credit returns. Although whether this risk is priced remains an open question, given their small exposure, the impact of currency union on the borrowing costs of companies will most likely be small. Works Cited Black, John. “Currency Risk”. A Dictionary of Economics. Oxford University Press, 2002. Campbell, John Y.
“‘Stock Returns and the Term Structure'”, Journal of Financial Economics, 18(1987): 373-99. Chang, Eric C. and Roger D. Huang. “‘Time Varying Return and Risk in the Corporate Bond Market'”, Journal of Financial and Quantitative Analysis, 25(1990): 322-40. Chow, Edward H. , Wayne Y. Lee and Michael E. Solt. “‘The Exchange-Rate Risk Exposure of Asset Returns'”, Journal of Business, 70(1997): 105-23 Dermine, Jean and Hillion, Pierre (eds. ). European Capital Markets with a Single Currency. Oxford University Press: Oxford, 1999. Fama, Eugene F. and K. R. French.
“‘Business Conditions and Expected Returns on Stocks and Bonds'”, Journal of Financial Economics, 25(1989): 23-49. Jorion, Philippe. “The Pricing of Exchange Rate Risk in the Stock Market'”, Journal of Financial and Quantitative Analysis, 26(1991): 363-76. Litterman, R. and T. Iben. “‘Corporate Bond Valuation and the Term Structure of Credit Spreads'”, Journal of Portfolio Management, (1991): 52-64 Saa-Requejo, Jesus and P. Santa-Clara. ‘Bond Pricing with Default Risk’, working paper, UCLA, 1997. Sarig, Oded and Arthur Warga. “‘Some Empirical Estimates of the Risk Structure of Interest Rates'”, Journal of Finance, 44(1989): 1351-9.