Bond Market I: Four Models. What is driving the bond market currently, and how much higher will it take bond yields? In my new book Predicting the Markets, I discuss four of the most widely followed models used to explain the 10-year US Treasury bond yield and to predict it.

I won’t keep you in suspense: The bond yield isn’t being driven by inflationary expectations, in my opinion, but rather by the federal funds rate (FFR) and the perceptions of how high the Fed will take it during the current tightening cycle (Fig. 1 and Fig. 2).

My central premise is that the bond market perceives that the Fed isn’t behind the curve on fighting inflation but rather on top of it. This explains why the yield curve has been flattening as the Fed has been tightening. The FFR is up from near zero since late 2016 to 1.50%-1.75% currently. It is likely to be raised three times to 2.25%-2.50% by the end of this year. Two more rate hikes next year should take it up to 2.75%-3.00%. If that’s it for a while and inflation remains subdued, as I expect, then the bond yield should peak at 3.50% over the next 12-18 months. Now let’s review the bond yield models briefly and discuss which of them might be working best right now:

(1) Bond Vigilantes Model. The Bond Vigilantes Model simply compares the bond yield to the growth rate in nominal GDP on a year-over-year basis (Fig. 3). In my book, I observe, “My model shows that since 1953, the yield has fluctuated around the growth rate of nominal GDP. However, both the bond yield and nominal GDP growth tend to be volatile. While they usually are in the same ballpark, they rarely coincide. When their trajectories diverge, the model forces me to explain why this is happening. On occasions, doing so has sharpened my ability to see and understand important inflection points in the relationship.”

Here’s why I call it the “Bond Vigilantes Model”: Bond investors failed to be vigilant about inflation during the 1950s through the 1970s. So the spread between the bond yield and nominal GDP was mostly negative during this period (Fig. 4). During the 1980s through the early 2000s, the Bond Vigilantes saddled up and fought off inflationary pressures by keeping the spread above zero most of the time and widening it whenever inflation showed signs of picking up. I count five such anti-inflationary widenings during this period that were soon followed by decelerating nominal GDP growth.

Since then, inflationary pressures have been mostly dissipating, and the spread has been mostly negative. That’s because there was less reason to be vigilant, i.e., to fear reflation. The negative spread during the mid-2000s was attributed by both Alan Greenspan and Ben Bernanke to a “global savings glut,” which kept bond yields down even as the Fed raised the FFR. Since 2008, the major central banks feared deflation and purchased lots of bonds through their quantitative easing (QE) programs.

However, the Fed has been normalizing monetary policy since the end of October 2014, when QE was terminated (Fig. 5). During October 2017, the Fed started to taper its balance sheet. (See Chronology of Fed’s Quantitative Easing & Tightening.) The Fed started raising the FFR at the end of 2015 and waited until the end of 2016 to do so again. The bond yield bottomed at a record low of 1.37% on July 8 of that year. The Fed raised the FFR three times during 2017 and is expected to do so three or four times this year. The bond yield rose to 3.11% last Thursday, the highest since July 2011, reflecting the Fed’s normalization of its balance sheet and interest rates.

(2) Inflation Premium Model. In my book, I observe: “The Inflation Premium Model posits that the nominal interest rate is equal to the real interest rate plus a premium reflecting inflation expectations. That sounds plausible in theory, but it’s not particularly useful for predicting interest rates. It certainly doesn’t make very much sense for short-term interest rates—particularly not for the overnight federal funds rate, since inflationary expectations over such a short period are irrelevant unless the economy is plagued by hyperinflation.”

A few pages later, I explain that the model “assumed even greater importance in the fall of 2015. The minutes of the October Federal Open Market Committee (FOMC) meeting began with a very dense section titled ‘Equilibrium Real Interest Rates.’ It focused on the concept of r-star (r*), which is the ‘neutral’ or ‘natural’ real interest rate. This was defined in the minutes as ‘the level of the real short-term interest rate that, if obtained currently, would result in the economy operating at full employment or, in some simple models of the economy, at full employment and price stability.’ …

“Does r* have any practical significance? In the past, there was a presumption that it is reasonably stable, I can’t find a way to measure it that results in a reasonably stable metric. More importantly, Fed officials have acknowledged that it might have been lower than they expected following the Great Recession. …

“Daily data are available for the yield on the 10-year US TIPS [i.e., Treasury Inflation-Protected Securities] since 2003” (Fig. 6). “Presumably, it should be a market-based measure of the real interest rate, and the yield spread between the nominal 10-year Treasury bond and its comparable TIPS should be a measure of inflationary expectations over the next 10 years at an annual rate. In the short period of available data, the nominal 10-year Treasury bond yield has been more highly correlated with the TIPS yield than with the yield spread between the two” (Fig. 7). “This suggests that the TIPS and Treasury yields share a common driver, and that it’s not inflation but the other component of their yields—i.e., the real interest rate, which is hardly a stable fixture.”

(3) Yield Curve Model. The Yield Curve Model posits that bond yields are determined by expectations for short-term interest rates over the maturity of the bond. These expectations are embedded in the “term structure of interest rates,” as reflected in the shape of the yield curve. As I explain in the book, “The slope of the yield curve reflects the ‘term structure’ of interest rates. Think of the 10-year yield as reflecting the current one-year bill rate and expectations for that rate over the next nine years. … An ascending yield curve indicates that investors expect short-term interest rates to rise over time. … A flat yield curve suggests that investors expect short-term rates to remain stable for the future. … An ‘inverted’ yield curve has a downward slope, suggesting that investors are scrambling to lock in long-term yields before they fall.”

I conclude as follows (italicization added here for emphasis): “In this Yield Curve Model, inflation matters a great deal to markets because it matters to the central bank. Investors have learned to anticipate how the Fed’s inflationary expectations might drive short-term interest rates, and to determine yields on bonds accordingly. So the measure of inflationary expectations deduced from the yield spread between the Treasury bond and the TIPS might very well reflect not only the expectations of borrowers and lenders but also their assessment of the expectations and the likely response of Fed officials! The data are very supportive of these relationships among inflation, the Fed policy cycle, and the bond yield.”

Since the election of President Donald Trump on November 8, 2016 through last Friday, the FFR is up 125bps, the two-year Treasury note yield is up 168bps, and the 10-year Treasury bond yield is up 118bps. The spread between the two-year yield and the FFR is up 45bps over this period, while the spread between the 10-year yield and the FFR is down 7bps and the spread between the 10-year and two-year yields is down 50bps (Fig. 8).

My interpretation of these data: The bond market is anticipating that the Fed will raise the FFR, as widely expected and reflected in the two-year yield, but that inflation will remain subdued, as reflected in the narrowing of the spread between the 10-year and two-year yields.

(4) Flow of Funds Model. During the 1970s, Henry Kaufman, the chief economist of Salomon Brothers at the time, provided detailed analyses of the supply and demand for bonds. Although he had a great bearish call on bonds, I doubt it resulted from his flow-of-funds approach. More likely, he recognized that inflationary pressures were building. He was dubbed “Dr. Doom” by the press for his downbeat forecasts, accurate though they were for a while.

During the 1980s, there were plenty of doomsayers who focused on the mounting federal government budget (Fig. 9 and Fig. 10). Many of them predicted higher inflation and bond yields. I predicted that the secular forces of disinflation would prevail, leading to lower deficits notwithstanding the federal budget problem.

The current outlook for the federal budget deficit has deteriorated significantly as a result of the Tax Cuts and Jobs Act passed late last year and a profligate congressional budget spending agreement between Democrats and Republicans early this year. Once again, we will have a test of whether the bond market is driven by supply/demand fundamentals. I currently expect that the increasing supply of US Treasury bonds will find enough demand at yields closer to 3.00% than to 4.00% as long as inflation remains subdued, as I expect.

Bond Market II: Global Savings Glut. The global savings glut theory postulated by Greenspan and Bernanke in the mid-2000s remains controversial. They claimed it explains why bond yields and mortgage rates remained low even though the Fed was raising the FFR. The theory’s critics say that the Fed raised the FFR too little, too late, and too predictably (in increments of 25bps per meeting from June 30, 2004 to June 29, 2006), thus setting the stage for the credit excesses that led to the mortgage meltdown of 2007 and 2008.

In any event, the US bond market has certainly become more globalized. This might partially explain why the 10-year US Treasury bond yield has been below the growth rate of US nominal GDP and should remain so. The comparable German and Japanese bond yields remain well below 1.00% (Fig. 11).

Both the ECB and BOJ recently were unpleasantly surprised to see their core CPI inflation rates fall during April to 0.7% from 1.0% in March in the Eurozone and to 0.2% from 0.3% in Japan (Fig. 12).

Dr. Ed Yardeni is the President of Yardeni Research, Inc., a provider of independent global investment strategy research.