The inversion of the yield curve has raised concerns in the media over the possibility of a recession in the next 12 to 18 months. Considering the inversion has been a reliable indicator during the past 40 years, it could be reckless not to pay attention to it. However, the U.S. bond market has also significantly transformed in the past decade, which raises the question – Does a yield curve inversion still have the same predictive power it had in the past? In short, I think that this time is different for several reasons, but the yield curve remains an important indicator that we should still pay attention to. Let’s start with what has changed.
We usually look at the spread between the 10-year and three months treasury bond yields, and over recent weeks the spread has turned negative. However, why is the logic behind its predictive power of future recession remains unclear. The most commonly used story is that investors expect short-term interest rates to fall in the coming year or two as the Federal Reserve reacts to an economic slowdown, and thus investors stockpile on long-term bonds; this buying spree pushes down long-term bond yields and leading to an inverted yield curve. This explanation only takes into account the changes in short term interest rates.
However, we should note that not only short-term interest rate affects long-term bond yields, the term premium — which is the compensation investors demand for holding long-term bonds in case of a sudden burst in inflation – also has an important role in affecting long-term bond prices.
We assume the term premium remains flat or doesn’t play a significant role in inverting the yield curve. One of the reasons why this time may be different is that the recent developments in the term premium.
Source: FRED and Federal Reserve Bank of New York
In the chart above are the spread of 10-year to 3-months bond yields and the term premium of a 10-year treasury bond. It clearly shows that this time the term premium has plummeted to a record low level – the lowest it has ever been. Back in 2007, during the previous yield curve inversion, the term premium has also trended downward but remained above zero and fell to its lowest level only as the yield curve inverted. This time, however, the term premium has been in negative territory for a better part of 3 years. And a negative term premium only pushes further down long-term yields. The negative term premium may suggest that people have concerns over deflation rather than inflation. Former Federal Reserve Chairman Ben Bernanke has discussed why the term premium is so low, but this issue still requires research. In any case, the negative term premium suggests that the yield curve inversion isn’t only about investors expecting lower interest rates, but also they don’t demand companion for inflation risk.
The second reason why this time could be different refers to the Fed’s balance sheet. In the past, the Fed’s balance sheet was much smaller relative to the size of the economy, and its composition was different. However, the Fed’s balance sheet’s size is roughly 20% of the U.S. GDP and a sizable proportion of its balance sheet comprises of long term bonds and mortgage-backed securities.
Through the QE programs the Fed aimed to lower long-term bond yields (and thus encourage investment). According to some estimates (see here, opens pdf) for every 10% of GDP of bond purchases (as part of the QE programs), the 10-year bond yield fell by anywhere from 15 to 240 bp. Since late 2017, the Fed has been cutting down its balance sheet in a normalization process that should end in the coming months. However, this process has been done by mostly letting short- and medium-term bonds to expire without repurchasing new ones. As a result, the changes in the size and maturity of the Fed’s balance sheet may have also contributed to the fall in long-term term premiums and, indirectly, the flattening of the yield curve.
Finally, the composition of bondholders has expanded to include a larger share of international bondholders than in the past. Not only Japan and China are the largest foreign holders of U.S. Treasuries with over a trillion-dollar worth of bonds for each country; all countries are estimated to hold a total of nearly $6.5 trillion (as of April 2019) or 30% of the national public debt. Back in 2007, this share was only 22%. How could the rise in foreign holdings affect the yield curve? Foreign holders of U.S. Treasuries have been contributing to the drop in bond yields (for more see here and here) including the term premium as countries – mainly China and Japan – have been exporting their savings to the U.S. and are likely to sell off their holdings only if they need to defend their currency.
All these reasons could make it harder for the inverted yield curve to be a reliable predictor of recessions as in the past. But it doesn’t mean we should discount the recent inversion all together; after all, the issues listed above have culminated for years and yet only recently the yield curve has inverted. This inversion may have had two direct culprits: The December rate hike, and the concerns over the trade war. The former, even ex-ante, appeared as a policy mistake; since then, the Fed has reversed course by lowering the dot plot from two hikes in 2019 back in the December 2018 meeting to none in the March meeting. Moreover, the market expects between two and three rate cuts by the end of the year. And in the upcoming meeting the Fed may even indicate a rate cut or two in the dot plot. As for the trade war, at this point, it is the U.S. against the world, as no country is safe from a trade dispute, although so far, Trump’s primary rival remains China. The markets expect these disruptions to have a harsh effect on investments and trade flow that could eventually push to slower economic growth even in the U.S. Albeit, at this point it seems still a stretch as to how this trade war, at its current form, lead to a recession. Perhaps, a deterioration of the trade war with other neighboring countries and Europe along with the already global slowdown could be enough to instigate more economic weakness that will eventually affect U.S. consumers and firms.
The bottom line is that yes, this time might be different. However, no, it does not mean we should discount the changes in the yield curve. It’s still an essential indicator, albeit the factors I have listed above might make it less reliable than in the past. Also, if the trade war rhetoric was to subside (But I doubt it will); or if the Fed were to start lowering interest rates to buy “recession insurance” (I still think the Fed will move very slowly on this issue – much slower than the market currently expects), then these developments could eliminate the recent yield curve inversion until, at least, the next Presidential tweet.