Holding cash is a key risk management tool, and corporations are holding more cash than ever. Where should the corporate treasury park that cash? In which currency should these balances be held–US dollars, Euros, Yen, Swiss Francs, Australian Dollars or what? The choice of the currency denomination of cash investments is the flip side of the problem of selecting the currency denomination of debt. Modern capital markets confront corporate treasury with a broad array of opportunities for borrowing and investing in various currencies.
Assuming the expected returns in all currencies are fairly priced for risk (i.e., Uncovered Interest Rate Parity holds), the answer will depend on the multinational structure of the company’s business and its exposure to fluctuations in the different currencies. Some companies will be better off stashing cash in dollar denominated securities because they anticipate future net cash outflows denominated in dollars, while others will be better off stashing cash in Euro denominated securities and others in Yen, and so on. Many companies will have an optimal mix of cash stashed in a variety of currencies. We’ll call the company’s optimal mix under the assumption of Uncovered Interest Rate Parity the company’s Benchmark currency portfolio.
Other factors will matter, too, such as international tax rules, concerns about capital controls and so on.
But what about that big assumption we made up front? What if expected returns in all currencies are not fairly priced for risk, so that Uncovered Interest Rate Parity does not hold? What if investments in certain currencies are generating big profits, while investments in other currencies are generating losses?
It is well documented that a speculative portfolio built by purchasing high interest rate currencies and selling low interest rates currencies—the carry trade portfolio—has been very profitable over many years.The figure below, taken from a recently published paper by Burnside, Eichenbaum, Kleschelski and Rebelo (here is the free working paper version), shows the cumulative return to an investment in the carry trade portfolio between 1976 and 2009 as compared against the cumulative return to an investment in US stocks and the return to an investment in US Treasury Bills.
The carry trade portfolio earns almost the same return as the stock portfolio and has a significantly lower standard deviation, and therefore a much larger Sharpe ratio. Another way to view the same point is to construct a levered carry trade portfolio so that it has the same risk as the market. In that case, the return to the levered carry trade portfolio is much higher than a portfolio of US stocks as seen in the figure below, taken from this paper by Lustig and Verdelhan. The carry trade portfolio is marked as “levered HML”.
1 dollar invested in the US stock market in 1971 yielded 40 dollars at the end of 2005. In contrast, the same 1 dollar invested in the levered carry trade yielded 135 dollars.
These results are very suggestive that Uncovered Interest Rate Parity does not hold.
In that case, many corporate treasurers are tempted to deviate from the company’s Benchmark currency portfolio and stash a larger portion into the profitable currencies and a smaller portion in the unprofitable currencies. Their company’s cash portfolio becomes a combination of the Benchmark currency portfolio plus a speculative portfolio. The profitability of the carry trade is obviously going to be enticing to any corporate treasury. See our post on Nintendo’s cash-currency strategy.
But should it be? Is past performance a reliable guide to future returns? Are there some big risks that this overlooks? What about the costs paid by the company when the carry trade makes a big loss?
More on this in forthcoming posts.