Is the Inverted Yield Curve Head Fake? I Don’t Think So

Is the inverted yield curve’s recession code red a “head fake”? I have received numerous questions, all focusing on other explanations. There are three main themes: 1) the inversion is technical (rather than fundamental); 2) the inversion is an artifact of quantitative easing; and 3) the inversion is inconsistent with the strong economic numbers we are seeing. Let me tackle these questions in a Q&A.

One media story said the inversion was caused by people prepaying their mortgages and banks being awash in cash - which they invested in the 10-year Treasury bond driving down the yield. Does this make sense?

Harvey: It makes sense that in a lower-rate environment homeowners refinance their mortgages. However, the “extra cash” part of the story makes no sense. Homeowners prepay their mortgage and then refinance into a new, lower rate mortgages. It is not the case that home owners simply pay off all their mortgage debt and do not refinance. Hence, this story that banks are awash with cash because of prepayments and are deploying the cash to buy the 10-year bond is false. 

There were a number of stories this week about the inversion being purely technical as pension plans scramble to match duration in a falling yield environment. Do you agree?

Harvey: Notice the way the question is phrased. Given the falling rates, actions are taken that might further reduce the rates. Essentially, these actions are a reaction to the inversion – not the cause of the inversion. Let’s unpack. First, duration is the sensitivity of the bond price to a 1% change in the interest rate. Longer-term bonds are much more sensitive to interest rate changes than shorter-term bonds because the change in the interest rate has a much bigger impact on long-dated cash flows. This is just the concept of compounding. If you invest $100 at 1% or 2%, there is not much difference over one year ($1), however, over 30 years there is a huge difference ($46).  Suppose you are a pension fund and you have a forecast of the payments you need to make in the future to pensioners (your liabilities). You hedge these payments by having a portfolio of assets that delivers the same interest rate sensitivities as your liabilities. This portfolio of assets includes both long-term bonds and mortgages. When interest rates drop, the mortgages effectively transform from being long-term instruments to a short-term instruments because many homeowners will prepay. That is, why would a savvy homeowner continue to make payments on a 5% 30-year mortgage when they can refinance into a 3.5% mortgage? This is the duration mismatch mentioned in the media. The pension plan needs to buy long-dated bonds to retain their hedge. The buying of these bonds drives prices up and yields down. Again, importantly, this is happening after the fact. The mechanism is credible and can account for the momentum in bond yields (and the momentum in the yield curve inversion). It does not account for the inversion in the first place. Anyway, pension funds don’t have to just buy the US 10-year Treasury, plenty of other choices, including lower-coupon mortgage-backed securities, are available to them.

What about insurance companies?

Harvey: Insurance companies are in a very similar situation to pension funds. The global insurance industry is massive, around $25 trillion, and sells many products, including those with long-term savings and pension features. Again, they need to match liabilities and assets. Suppose they are obligated to pay a certain amount in 30 years. They cover it by purchasing a long-term bond with a 3% interest rate. Suppose the interest rate drops to 2%. To cover new obligations, they need to buy more of the bond. That is, to get the cash flow of the equivalent of 3%, you need 50% more of a 2%-yielding bond. Hence, as rates drop, the insurance company needs to buy more. This additional buying pressure increases prices and drives long-term yields even lower. Again, this is very similar to the pension situation. This behavior induces a momentum effect: as rates decrease, insurance companies buy more and rates decrease even more. The opposite happens when rates increase. This does not account for the yield curve inversion, but it certainly could make the inversion even larger.

Were speculators in the options market responsible for the yield curve inversion?

Harvey: I doubt it. There are many possible trades here, but let’s focus on the one mentioned in the media. Many people bet on the volatility of bonds and stocks by taking positions in call and put options. If you think volatility is going to go up, you would buy both calls and puts on the same asset. If volatility goes up (and it doesn’t matter if prices go up or down), one side of the trade will pay off. If volatility remains the same, then you lose the amount you paid for the call and the put (the option premia). Now consider the other side of the trade, where you are short the call and the put. Being short an option has a much different risk profile. If you buy the option, the maximum you can lose is the price you paid. If you sell, say a call option, your losses can be catastrophic because there is no limit to the liability. So, if you are short these options, you likely need to hedge. If interest rates unexpectedly move even lower, there is increased demand for the hedging asset, which may be a longer-term Treasury bond. However, there are other ways to hedge including buying options with different strike prices. While the mechanism is credible, like the first argument, it relies on rates already falling. Again, it might explain some of the momentum (yield curve becoming more inverted since June 30, 2019), but it does not explain the inversion.

The unemployment rate held at 3.7%. How can a recession be around the corner with such low unemployment?

Harvey: Unemployment is always lower before recessions - almost by definition. Furthermore, unemployment is a classic "lagging" indicator of a recession. In contrast, the yield curve reflects expectations about the future. Employment, profits, etc., all tell us about the past. Ideally, we want a window to the future. A better employment indicator would be CFO's hiring plans. These plans are set in advance and tell us about future employment. The Duke-CFO survey measures both employment plans as well as capital investment plans.

Is QE causing the inversion?

Harvey: Since 2017, the Fed has been reducing the size of its assets. From 2015 to 2017, the size of the Fed’s assets was constant. The Fed has now reduced its assets by about $700 billion. You would think that the selling of these bonds would reduce prices and increase yields. We have seen the opposite. There is talk the Fed may increase its assets once again – but that is just talk. It is hard to attribute quantitative easing (QE) as the cause of the inversion when the Fed’s actions are more consistent with steepening. That said, the Fed is still actively buying 10-year and 30-year bonds. You can download the data here.

My point is that the overall holdings have decreased and it is very hard to tell the story that QE is inverting the yield curve when holdings have decreased. There is one piece of data that would be very useful to have: the average maturity (or duration) of the Fed holdings. Unfortunately, these data have not been updated since 2017.

Please share this article. Let’s start a conversation.