The Reason of Treasuries Won’t Go Away

There have been some reports in the media about how the 10-year Treasury, if hedged into either euros or yen, now has a yield of roughly zero. The implication is that if European and Japanese bond buyers have been a major driver of demand for U.S. bonds, then they would presumably be sensitive to the hedged yield, not the nominal yield.

If that yield advantage has gone away, could demand for U.S. Treasuries also go away? There is probably something to this line of thinking, but the conclusion is much narrower than people are making it out to be.

First, what do we mean by a hedged yield? If you are a European bond buyer and you want to keep your currency exposure in euros, then any non-euro bonds you buy have to be hedged back into your home currency. There isn’t just one way to accomplish this, but a common way would be to use a basis swap.

In simple terms, you pay U.S. LIBOR (currently 0.81%), you receive EURIBOR (which is LIBOR but in euros; that rate is currently -0.30%), plus one side or the other has to pay to make the currency element happen, which is around 0.50% right now paid by the euro side of the trade. If you add all that together, you get 1.61% per year in hedging costs, which is basically equal to the Treasury yield. So when I use the term “hedged yield”, I mean the Treasury yield less this hedging cost. Since the hedge cost is 1.61% and the Treasury yield is 1.58%, the hedged yield is -0.03%.

Let’s assume that our European investor faces just two choices: buy American Treasuries and hedging, or German Bunds. The investor might just compare this hedged yield to the Bund yield and pick whichever is higher. On that very simplistic basis, the investor is basically indifferent. The hedged yield of -0.03% is almost exactly the same as the 10-year Bund yield at -0.05%.

That wasn’t true at the beginning of the year, when the hedged yield was 1.20% and German Bunds were yielding just 0.62%. That kind of yield advantage, even after hedging away the currency risk, is bound to create demand for U.S. bonds.

So now that the hedged yield and the Bund yield have converged, does that mean this source of demand will go away? And if so, doesn’t that mean rates in the U.S. need to rise? I think this is taking this analysis too far, for a few reasons outlined below.

You can see that the net hedged yield (yellow) is sometimes way above or way below the German Bund yield (red). This persists for years at a time and reverses for no obvious reason. It is also not predictive of future yield changes. A regression of the current spread between hedged yield and Bund yield vs. future Treasury yield changes showed no relationship at all.

This might seem counter-intuitive. Why would there be such a weak relationship between these two yields that are such obvious substitutes?

In part that is because this isn’t an arbitrage…

While they are substitutes, the hedge isn’t without risk. Over the life of this kind of trade, the hedger is exposed to changes in LIBOR. If LIBOR in the U.S. were to rise or if EURIBOR were to fall, the cost of holding the trade would go up. Lo and behold, we’ve been seeing LIBOR rising recently without any corresponding change in longer-term rate expectations.

This has to do with some friction around changes in money market rules and is likely to be a one-off, if not a fleeting issue. It could be that buyers are betting that U.S. LIBOR will fall back once this money market issue has passed.

Regardless, this is why there is not now, nor will there ever be an iron-clad relationship between these two yields. But there is probably a bigger reason why there seems to be a random relationship between the rates.

Mostly because hedged buyers probably aren’t the marginal buyers

Most large global investors have a healthy need for U.S. dollars. Financial institutions, such as insurers, are likely to do plenty of business in the U.S., therefore aren’t necessarily hedging anyway. These institutions might also have an increasing preference for USD exposure, which again means no hedging.

We can see this if we look historically at the U.S. vs. German chart above. The correlation between the two yields without hedges is 0.91. With the hedge, it is just 0.88. In other words, the hedge cost doesn’t explain anything about the variation.

The hedging example everyone is using also makes the assumption that the buyers’ cost of funds are based on EURIBOR. That is also not necessarily true. Insurers are a good example. Their cost of funds is based on premiums and liabilities, not EURIBOR. Or, take sovereign wealth funds and foreign currency reserve holdings. Neither of these really faces a traditional cost of funding.

It is possible that hedged investors weren’t the marginal buyers in the past, but are now. It is possible that the unusual circumstances of negative rates have caused buyers to get more creative. If so, the fact that the hedge cost has risen to be equal to the yield on the 10-year will be something of a governor on fresh demand. But it definitely can’t cause a material back-up in Treasury yields. If the bonds are just a smidge too expensive and yields back up a bit, the demand comes back.

Both comments and pings are currently closed.

Comments are closed.