I had long wondered why the yields on US T-bonds and the yields on Japanese and European sovereign bonds would not converge. The conventional theory is that when interest rates are relatively high in country A, then investors in country B should buy the bonds of country A, lowering the yield in country A and raising the value of country A’s currency. Over the last 12 months, this did in fact happen with respect to the US and the Eurozone. Our dollar rose relative to the euro, our T-bond yields fell, and so our T-bond prices rose. But the Eurozone sovereign bond yields did not rise despite the euro dropping in value. In fact, the Eurozone bond yields dropped a lot. The yield on the 10-year German bund fell by 112 basis points in 12 months as of 9/25/19. The yield of the 10-year US T-bond fell 136 basis points. So there was a little convergence, but not much.
Why did European investors accept negative yields on their sovereign bonds while the US 10-year T-bond had a positive yield of 2% or higher. I had read that the Europeans had to hedge against currency losses, which did not make much sense to me, because the US dollar was strong, and if they spent their euros to buy US-dollar-denominated T-bonds, the European and Japanese purchases could make the dollar rise even further.
In an article entitled “How to Make Money from Money-Losing Bonds” by Katherine Greifeld, which appeared in the Sept. 16 issue of Bloomberg Business Week, I discovered why the bond yields have not converged. It’s because of currency futures and short-term interest rate differentials. Perhaps more sophisticated investors than I assumed that most of us knew why the differential yields existed, but I’m not stupid, and I saw no Seeking Alpha articles that even asked the question, much less addressed it. And the implications of the question and the answer are considerable, because the answer predicts, almost assuredly, that long-term Treasury bond yields will continue to decline, even if the EU bond and Japanese bond yields do not decline further. That is, the bond yields are going to converge.
Greifeld indicated that US investors were buying negative-yielding Japanese bonds and simultaneously shorting the US dollar against the yen. So in essence, they buy yen now, and sell those yen in the future to redeem the US dollar at a price agreed on 12 months in advance. When they sell their bonds, after 12 months say, they will get fewer yen than they paid, but they use those yen to buy back the dollar at a much lower price than what they sold it for 12 months prior, and so they get about 2.5% more dollars than they sold. They make more than they would have, had they just bought 10-year US T-bonds, and just retained those bonds.
The issue for me is; why are the currency futures priced like that? Why does the bet that the US dollar will decline relative to the euro and the yen persist?
On 9/23/19, the closing cash price for euros on the CME was one euro equals $1.09924. At the same time the futures price that would settle in September 2020 was one euro to $1.1285. So if a US investor sold dollars on 9/23 to buy euros, she could buy 909.72 euros for 1,000 dollars. Then she can promptly sell those euros in the futures market with a contract settlement on the third Friday of September 2020. At one euro to $1.1285, the 909.72 euros would fetch $1,026.62, which is an absolutely safe return of 2.66% on the investment, quite a bit higher than the 10-year US Treasury yield, and free of the risk that the 10-year Treasury bond could decline in price because of inflation.
The same principles hold for buying Japanese yen. However, there is a problem in either buying the Japanese yen or the euro and holding them for a year. The interest rate on bank reserves in the Eurozone is negative 0.6%. So holding euros in cash will cause them to decline 0.6%, and our 909.72 euros will only be worth 904.26 euros after one year, which would only buy $1,020.46. So the return would be 2.05%, but that is still higher than the yield on a 10-year T-bond. If you bought a 10-year French bond or a 10-year Japanese bond, which yields about -.20%, then the foreign currency only drops about 0.2% for the year, and the 909.72 euros would be worth 907.9 euros, which could buy $1,024.56, a return of 2.46% on negative-yielding French bonds.
Of course, you could make higher returns on your euros by buying riskier bonds, but we’re talking about institutions, like pension funds, which can’t afford to take on default risk. They want safe returns proffered by safe sovereign bonds.
So dollar futures are in persistent contango against certain currencies, in particular the euro, the pound, the Swiss franc and the yen. Why? In the US, the IOER (interest rate on excess reserves) is 1.8%. In the Eurozone it is negative 0.6%. The difference is 2.4%. If you hold dollars, then your $1,000 will be worth $1,018 after one year. If you hold 1,000 euros, you will have 994 euros after one year. That is why the contango exists and persists. It is because of the difference in short-term interest rates, the cash rates.
The cash rate in Mexico is 8%, 6.2% higher than the US rate. The 12-month forward futures price for the peso is 5.2% lower than the current price. It would appear that the difference in short-term interest rates between countries is the single most important determinant by far of the currency future prices.
I calculated above that the one-year futures price on the dollar relative to the euro was 2.66% higher than the current price. That is greater than the difference between the two cash rates (2.4%). The percentages are not equal because clearly the futures market has to price in expectations of short-term interest rate changes by both parties. Actually, given the expectation that the US will lower its Fed funds rate (and IOER) by a greater amount than the Eurozone will lower their cash rate, one would think that the dollar is underpriced by the futures market. If short-term interest rates converge, which is probable, then the contango of the currency futures will diminish.
It follows that, if the US dollar futures are no longer in contango, then US bonds will be much more attractive to foreign investors, and so the yields on the US T-bonds and the European and Japanese sovereign bonds will converge.
The negative sovereign bond yields in the Eurozone and in Japan stem from coercion (as Hussman calls it) by their central banks. The central banks force negative short-term interest rates on bank reserves. The banks cannot get rid of these negative-yielding reserves in the aggregate. They have been imposed on banks in the aggregate by the central bank buying bonds, and the money to buy these bonds ends up in bank reserves. The intent is for banks to lend those reserves out, but that doesn’t get rid of the reserves in aggregate. It just changes ownership of the reserves. When one bank lends money to a person, business or government, the money is subtracted from the bank’s reserves, but it is going to be added back to the reserves of all banks when the borrowed money is spent and deposited in bank accounts. Forcing negative short-term interest rates on the banks forces everyone to seek investments with higher returns, and those assets then get bid up in price such that the expected, risk-adjusted, total return from those assets becomes equal to the short-term interest rates; that is, it becomes negative.
Those negative returns are prone to be exported. As Lyn Alden Schwartzer has pointed out on Seeking Alpha, the Eurozone and Japan have declining workforces. The only prospect for economic growth in those regions is for the growth rate in the productivity of that workforce to exceed the rate of decline in the numbers of the workforce. Economic growth also requires that their trading partners have economic growth so that exports to their trading partners grow. Both regions are net exporters and have large current account surpluses. These current account surpluses and QE mean that the domestic banks can finance the large fiscal deficits of their own governments to the point where the bonds have negative yields.
The US does not have a declining workforce, and our economic growth prospects as a result are relatively good. The US has a huge current account deficit. That means we have little savings, except for our corporations, which prefer to buy back their own stock. The US cannot finance its own debt without providing a positive yield on that debt, which brings me back to the original question: why don’t the Eurozone and Japan buy our debt to the extent that the yield on US T-bonds is zero? And I have answered that question. Because of the way the foreign exchange and futures markets operate, the differential in US short-term rates and the rates in the Eurozone and Japan approximately equals the differential in rates between their long-term sovereign bonds.
We know (well, we’re pretty sure) that the US is going to lower its short-term interest rates, whether Trump is president or not. How much doesn’t matter. We will lower them faster than Japan or the Eurozone will. Therefore, those rates will converge. Therefore, the long-term sovereign bond yields will converge. Therefore, US T-bond prices will continue to rise.
Have I left anything out? Well, I’ve ignored inflation risk I guess, but I don’t think it’s much of a risk. If inflation arises, it will probably occur both here and in the Eurozone. The issue is; what will be the response of the central banks? My argument is that US T-bonds will only drop in price (and yields increase) if the short-term interest rate differential increases because the US raises short-term rates more than the Eurozone does. You can sell your T-bonds in advance of the Fed raising short-term rates because the Fed tends to be market driven. T-bill rates have tended to move a month or more before the Fed changed rates.
I have long accepted that the Federal Reserve had little influence over T-bond rates. I saw that during the times of QE1, QE2 and QE3, yields moved up while the Fed was buying the bonds, and they moved down when the Fed stopped buying. My surmise was that the market expected inflation because of the QE and disinflation when QE ended. But now I’ve done a flip-flop. I’m arguing that the Fed can control or influence long-term rates through their control of short-term rates in the context of the global economic system. During QE1-QE3, the Fed did not change short-term interest rates. When the Fed did raise the Fed funds rate after 2016 and the Eurozone lowered its cash rate, T-bond prices tended to fall despite the perceived threat of deflation.
I don’t know if the Fed is aware of its ability to control long-term rates in this way. It probably is. They’re smart people. I think they don’t publicize some things that they are well aware of. They knew we were in a liquidity trap (“pushing on a string”), but never said so outright. Whether they know what they’re doing or not, need not stop us from making money from their actions. I’ve made a lot of money (for me) from T-bonds since mid-November 2018. I anticipate making a lot more.
Let me know where my logic is flawed, or what I have overlooked. I suspect that since October of last year T-bonds rose more than could be expected simply from a narrowing of the gap in short-term rates. Changes in investor expectations of future interest rate moves affect T-bond prices. Changes in the perceived risk of investing in assets that are an alternative to T-bonds will affect the price of T-bonds, and so will changes in the currency exchange rates that are independent of the differential in short-term rates, like inflation accelerating faster in the US than in Europe.
On Wednesday, September 25, T-bond yields rose a lot, much more than those in Europe. Of course, we have to wait for the European bond market to open to see if it follows the US bond market. As I write, the European bond market has not yet opened. The dollar rose a lot despite the tanking of the T-bonds, which should not occur. Gold fell, which is to be expected if the dollar is strong. The 3-month T-bill rate fell, which should help T-bonds, but it didn’t.
Perhaps T-bonds were simply overbought. The press tied the decline in T-bonds and the rally in stocks to the release of transcript of the Ukrainian phone call. Who knows? I like my thesis for the longer term and bought a lot of TMF today (the 25th). It will be interesting to see what European bonds do on the 26th as well as US T-bonds.
Disclosure: I am/we are long TMF, PTY, REM. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.