Foreign Bonds With Currency Hedging

Reader Question

Hey LTR, Here is a question, would it be possible get a higher return on fixed income by buying bonds in a foreign country (assume maybe a foreign country that has developed financial markets so it's risk equivalent to buy in US or close to risk equivalent) and then hedging the currency exposure to get a higher interest rate? I have zero expertise in both areas, maybe the currency hedge cost wipes out any interest rate premium gain. I'm sure others have thought of this. Thanks

My Reply

Good question. Unfortunately, it is as you guessed -- the currency hedge wipes out the deal. Here is why this happens...

You can think of any trade as going long whatever you buy and short whatever you sell. Investing in a regular domestic 10-year bond, for example, is going (long 10-year bond + short 0-year bond). That "0-year bond" is how I like to think of cash when discussing investing. Your net gain from the trade, relative to having not made the trade, is the return from your long investment minus the return from your short investment. If you hold to maturity and the 10-year bond ends up returning 40% while 0-year bond (cash) returns 30% in the same time span, then the trade netted you 10%. It's true that your total return is 40%, but you would have got 30% even if you did not make this particular trade. So the net gain from the trade is just 10%.

Once you start thinking in these terms, it becomes easy to analyze any trade, including investing in foreign bonds with currency hedging... Let's say you're buying 10-year Australian bonds. You go (long 10-year AUS bond + short 0-year US bond). The 0-year US bond is, of course, your US dollar-denominated cash. Then you need to establish your currency hedge. All that is is buying US cash with AUS cash, meaning you go (long US 0-year bond + short AUS 0-year bond). Add up your two trades. Short 0-year US bond from the first trade cancels with long US 0-year bond from the second trade. Your net trade becomes (long 10-year AUS bond + short 0-year AUS bond).

Now hopefully the problem becomes clear. If Australia has higher interest rates than the US, it will (generally speaking) have them higher across the yield curve, from 0-year bonds to 10-year bonds. Since your trade is (long 10-year AUS bond + short 0-year AUS bond), your gain from higher interest rate on the 10-year AUS bond is offset by the similarly higher interest rate from 0-year AUS bond. Whether Australia's interest rates are higher than US rates by 2% or 3% or even 10% across the yield curve, this currency-hedged trade would still net you big fat zero relative to simply buying the US 10-year bond to begin with.

The reality is even worse. One of the frictional costs of participating in the forex market is that you will not get all of the expected interest on the currency you are long while paying more than expected interest on the currency you are short. I have never traded forex and honestly don't even know the magnitude of this spread, but it is there. That might add another 0.1% or 0.2% or whatever the spread happens to be to the cost of the currency hedge. Now what was a break-even trade actually becomes a losing one.

This is why currency-hedged foreign bond investing is not often done in practice. The only way it stands good chance of being more profitable than investing in domestic bonds of same maturities is if the domestic yield curve is quite flat while the foreign yield curve is quite steep. For example, let's suppose 0-year US bond pays 2% while US 10-year bond pays 2.5%. Your expected profit (assuming the yield curve stays frozen in time -- which, of course, it won't) is just 0.5% per year if you go (long 10-year US bond + short 0-year US bond). But suppose 0-year AUS bond pays the same 2% while 10-year AUS bond pays 5%. Your expected profit from investing in currency-hedged 10-year AUS bond is now 3% minus whatever the frictional costs happen to be (which, remember, include the broker's skimming off the interest rate).

However even that last scenario is not fool-proof. The yield curves are not frozen in time, so it's quite possible that the short-end of the Australian yield curve quickly rises close to the long-end level, making the Australian yield curve flat, similarly to the US curve. Now your short 0-year AUS bond position will be costing you, say, 4% a year instead of 2% a year, which combined with the ever-present frictional costs could make your initially promising trade a barely profitable or even a losing one.

And, of course, if you never establish the currency hedge you expose yourself to the currency risk which is not present in investing in domestic bonds. Sometimes taking such currency risk pays off big time. If you had been investing in emerging markets bonds for the past 15 years or so you would have reaped great profits, both from high starting interest rates and currency appreciation relative to the dollar. I think such investments might have even exceeded the terrific returns from emerging market equities and definitely exceeded the very substantial returns from plain old domestic US bonds over the same time period. But, of course, currency risk could just as easily move against you -- as was likely felt (at least until a few months ago) by Japanese investors of the recent years who decided to look for higher yields abroad.