One of the biggest financial market surprises in 2017 was the sustained low bond yields as the Fed raised short-term rates and the significant flattening of the yield curve. The Fed has raised rates four times since December 2016, lifting the Fed funds rate target to 1.25 percent to 1.5 percent while 10-year Treasury bond yields are virtually unchanged, at 2.5 percent.
Factors that have suppressed bond yields include declining inflation and inflationary expectations; lower real rates reflecting diminished forecasts of real economic growth; global central banks’ excessive monetary easing—artificially low policy rates and massive bond purchases and holdings; and foreign purchases of U.S. bonds, attracted by their higher yields and liquidity.
Core inflation has remained below 2 percent for most of the time since the Fed officially established 2 percent as its longer-run inflation target. Importantly, the decline in inflation in 2017—contrary to expectations that the low unemployment rate and building economic momentum would drive up inflation—materially lowered inflationary expectations.
Perhaps the largest concern about the flat yield curve is whether it is predicting slower growth, or worse. In past cycles, while the yield spread has been positively correlated with economic growth—so that a significant flattener has been statistically correlated with softer growth—the correlation has been fairly loose. However, whenever the curve inverted, with short rates higher than bond yields, for whatever reason, a recession eventually ensued (with widely varying lags). Presumably, among other things, an inverted curve signals monetary tightness and interrupts the financial intermediation process, which portends a slump in aggregate demand.
But now interest rates and the current relatively flat yield curve and other measures of monetary policy and financial conditions are giving mixed signals on the economy. Although the curve is relatively flat, other indicators suggest monetary policy and credit conditions are easy: the real Fed funds rate is slightly negative, M2 money has been growing at an ample 6 percent rate, bond yields are very low and several indexes of financial conditions register as easy. Moreover, in the real economy, there are no apparent imbalances that signal trouble.
Economic growth is clearly gaining momentum. Consumer spending growth is solid, despite relatively modest gains in real disposable personal income, and is being driven by elevated confidence and household net worth rising to record levels. Business production is increasing to meet the stronger product demand and business investment spending is advancing, with strength in durable goods shipments. Profits and cash flows continue to improve, and business confidence is strikingly high. Exports are growing moderately, benefiting from synchronized global growth and strengthening trade volumes. Tax reform legislation is expected to provide an additional lift to economic growth.
The narrow yield curve does not seem to be inhibiting financial intermediation, nor at this point do we believe it is a reliable predictor that aggregate demand will slump.
Empirical research supports the view that the relatively flat yield curve is not pointing toward an economic slump or recession. Updating earlier empirical studies conducted by Fed researchers and academics, based on the current spread between the 10-year Treasury bond yield and the 3-mont Treasury yield, with the Fed funds rate still modestly below inflation, the probability of recession is estimated to be 6%. (”The Flat Curve: Not Predicting Slowdown”, January 4, 2018”).
This is not a surprise: during previous economic expansions following periods of monetary ease, raising the Fed funds rate toward a neutral policy does not harm growth. Under current circumstances, as long as troublesome imbalances do not arise in the real economy, the Fed can continue to raise rates along the path it has forecast as appropriate—three hikes in 2018—or even moderately faster without materially raising the probability of recession.
Besides below-target inflation, several other factors have contributed to lower inflationary expectations. First, the failure of wage gains to pick up—even with the unemployment rate at 4.1 percent, decidedly below standard estimates of its natural rate—has reinforced low inflationary expectations. (The Fed and most financial market participants have the perception that wage increases are a necessary precursor to higher inflation, even though widespread research finds that wages and inflation are contemporaneous.)
Second, anecdotal evidence of lower distribution costs and product prices—the so-called “Amazon effect”—has added to the notion that inflation will stay low. Third, technological innovations continue to reduce production costs and increase the quality of new products. Finally, the Fed’s public statements that inflation is too low and that inflation may not increase quickly to 2 percent have contributed to low inflationary expectations.
Wage gains and inflation are expected to accelerate in 2018, and the December jobs numbers already showed evidence of some gains. The supply of semi-skilled labor will be absorbed, which will add to the already-tight market for skilled workers and drive up their average wages, and the acceleration in nominal GDP in 2017 will be extended into 2018, which will provide businesses more flexibility to raise product prices and grant higher wages.
Along with lower inflationary expectation, low bond yields reflect low real interest rates, which have receded as expectations of potential growth have fallen. Estimates of potential growth have declined materially to 1.8 percent (Fed and Congressional Budget Office, CBO) from 2.5 percent before the financial crisis. These forecasts reflect weak gains in productivity and disappointing capital spending growth. In response, the median FOMC member now estimates that the appropriate longer-run Fed funds rate is 2.8 percent, which, adjusted for the Fed’s 2 percent inflation target, implies a 0.8 percent real Fed funds rate. That’s down from 4.0 percent nominal and 2 percent real as recently as March 2014. Actual real rates are expected to rise in 2018, reflecting stronger growth and higher expected after-tax returns on capital, driven in part by tax reform.
Global central banks have contributed significantly to low bond yields through their imposition of artificially low policy rates and massive asset purchases. Moreover, the central banks have signaled their intentions that they will be very slow to change these policies, and they are aiming toward “new normals,” with low rates and outsized balance sheets.
Maintaining unprecedented low policy rates, even after economies have been growing faster than potential, have kept short duration rates either negative or very low. The European Central Bank (ECB), Bank of Japan (BoJ), Swiss National Bank, and Swedish National Bank are imposing negative policy rates, while the Fed and Bank of England (BoE) are maintaining policy rates below inflation. The BoJ and ECB have signaled that raising their policy rates above zero is a long way off. The Fed has significantly lowered its forecasts of the appropriate path and longer-run Fed funds rate. Fixed income markets respond to such forward guidance.
Central banks’ aggressive purchases of bonds have reduced the float of outstanding sovereign bonds and lowered their yields. The Fed’s QE programs and policy of reinvesting maturing assets have resulted in holdings of $2.5 trillion in Treasury securities and $1.8 trillion of mortgage backed securities (MBS). This makes the Fed the world’s largest holder of each, with a 16 percent share of publicly-held Treasuries and 12 percent of total mortgage debt.
A critical issue is how sovereign bond yields respond when the central banks unwind their holdings. Facing uncharted policy territories of enlarged balance sheets and sustained negative real interest rates, and fearful that their “exit policies” may upset financial markets, the Fed and ECB will tip-toe toward unwinding their balance sheets. The Fed has begun its very gradual and passive balance sheet unwind, while the ECB is lagging behind, and has not given any guidance on when it may begin to reduce its holdings.
The Fed is achieving its gradual unwind by reinvesting over 90 percent of all maturing assets and allowing a very small portion of the principal of its holdings to roll off. While it has not established an official target for the ultimate size of its balance sheet, Chairman-elect Powell and several other Fed members have stated that the balance sheet will not be reduced below $2.5 trillion. Reflecting the current $1.6 trillion in currency, this suggests that the Fed’s “new normal” is to always maintain at least $1 trillion in excess reserves, which would limit the amount of the Fed’s unwind. Not surprisingly, this dovish stance—along with the messages being conveyed by the ECB and BoJ—is being reflected in low bond yields.
Ever since the 2008-2009 financial crisis, all of the Fed’s policies—low policy rates, quantitative easing and Operation Twist, and forward guidance—have all been pursued with the objective of keeping bond yields low. The Fed has succeeded, but through 2016, economic performance was disappointing. It is now ironic that, as the Fed has begun to normalize policy, it is expressing concern that the curve is too flat.
Circumstances in 2018-2019 may merit increasing rates faster than the Fed currently forecasts: 1) the tax reform legislation is expected to lift economic growth and expected after-tax returns on capital, which should raise real rates; 2) inflation may rise faster than the Fed forecasts (a gradual drift up to 2 percent ); and 3) the Fed’s maintenance of ample excess reserves in the banking system facilitates easy monetary and financial conditions.
Going forward, the probability of recession unfolding in the ensuing 12 months depends on a host of factors in the real economy, monetary policy, and external factors. Future yield spreads (and the real Fed funds rate) are endogenously determined by many moving parts in the real economy and financial markets. For now, all cyclical indicators suggest sustained economic expansion.
Probit regression models are binary response models in which the dependent variable takes on values of “0” or “1”—in our case, no recession or recession. Our analysis is based on earlier research that used probit analysis to estimate the probability of recession based on the yield curve and other variables .
The equation estimated using monthly observations from January 1961-October 2017 is:
Recession Likelihood (t) = -0.96 -0.57 Yield Spread (t-12) + 0.1 Real Fed Funds Rate (t-12)
This equation shows the strong value of the yield spread in predicting the probability of recession 12 months out. Specifically, the yield spread is inversely correlated with the probability of future recession, and statistically significant (see Table 1). A one percentage point increase in the yield spread lowers the Z-score of the probability of recession by 0.57. The real Fed funds rate is also statistically significant and positively correlated with the probability of recession with a fairly low estimated coefficient: a one percentage point increase in the real fed funds rate raises the Z-score of the probability of recession by 0.1.
Mickey Levy is the chief economist for the Americas and Asia of Berenberg Capital Markets, LLC, and member, Shadow Open Market Committee. The views expressed in this column are the author’s own and do not reflect those of Berenberg Capital Markets, LLC.
Interested in real economic insights? Want to stay ahead of the competition? Each weekday morning, E21 delivers a short email that includes E21 exclusive commentaries and the latest market news and updates from Washington. Sign up for the E21 Morning Ebrief.
 James Stock and Mark Watson, “New Indexes of Coincident and Leading Indicators” in Olivier Blanchard and Stanley Fischer, eds., NBER Macroeconomics Annual 1989, Vol.4, and Arturo Estrella and Fredric S. Mishkin, “The Yield Curve as a Predictor of U.S. Recessions”, Current Issues in Economics and Finance, Vol. 2, June 1996.