The Shift to Neutral: Understanding How Far, and How Fast the Fed is Willing to Go
We believe a perilous road awaits investors as the FOMC attempts to rapidly increase interest rates without pushing the economy into a recession. While the Fed has been successful in the past, the combination of uncomfortably high inflation, a new policy framework, and the war in Ukraine finds the Committee far behind the curve. Investors must prepare for the possibility that the Fed will be forced to sacrifice a growing economy to restore price stability. Our goal is to provide an understanding of the theoretical foundations and policy choices that led the Fed to this point as interpreting the Fed’s policy path is a critical input in determining the appropriate level of interest rates.1 At present, the risk is an undaunted Fed that can no longer backstop asset prices as it has since the Global Financial Crisis. We expect volatility to persist and positive returns to be scarce in benchmark assets until inflation falls towards 2 percent.
Before analyzing where the Fed is headed and what it means for interest rates, an understanding of the framework that guides monetary policy is necessary. The Fed is tasked with a dual mandate: maximum employment and stable prices. To achieve these outcomes, the Fed influences financial conditions through its primary tool—short-term interest rates. In particular, the Fed targets the nominal federal funds rate, an overnight interbank borrowing rate, to affect real outcomes like maximum employment. Therefore, the stance of monetary policy should be interpreted in terms of the real, inflation-adjusted interest rate because economic decisions are undertaken based on real rather than nominal quantities over the long-term. In other words, groups such as consumers and firms care about how much stuff they can buy today versus the future and the rate at which they substitute between the two.
So how does the Fed determine the correct interest rate? According to economic theory, the Fed is targeting the short-run neutral rate, r* (“r star”), consistent with the dual mandate. First, abstract away from business cycle fluctuations e.g., a recession.2 The state where the dual mandate prevails is characterized by two additional variables: the non-accelerating inflation rate of unemployment (NAIRU), u* (“u star”), and the central bank’s inflation target, π* (“pi star”). The relationship between these two and the expectations-augmented Phillips Curve, entails a trade-off between lower/higher unemployment and higher/lower inflation, controlling for inflation expectations (Ball and Mazumder, 2015)3. Putting it all together, potential real GDP is then the maximum rate of growth for the economy at full employment with stable prices.4 Growth is related to interest rates through the savings and investment curve, where lower/higher interest rates lead to higher/lower demand for investment, a component of GDP. Tying it all together, the neutral rate lies at the intersection of the savings and investment curve and potential GDP, y* (Williams, 2003)5. An interest rate below/above neutral results in growth above/below potential leading to higher/lower inflation relative to the central bank’s target as demand exceeds the productive capacity of the economy.6
In practice, however, the Fed often finds itself managing interest rates over fluctuations in the business cycle rather than in an economy in its equilibrium state. In periods when the unemployment rate is above/below NAIRU the Fed seeks to set interest rates below/above the neutral rate to increase/decrease growth. The conduct of monetary policy is further complicated by the fact that many of the variables underlying this theoretical model, u* and r*, are unobservable and unstable over time. The Fed estimates several models for these variables as well as inferring them from changes in growth, inflation, and unemployment. In a severe recession, forecasting errors around the precise level of u* and r* are inconsequential given the economy’s distance from its full potential. It’s only when the economy approaches these unobservable variables that “navigating by the stars”, or determining rate paths, becomes a difficult endeavor (Powell, 2018)7.
We have established a simple model for how the Fed thinks about the neutral rate as well as the uncertainty associated with estimating it, but we have given little thought to what determines the level of neutral. Economic theory suggests that the neutral rate is a function of economic growth and time preferences i.e., how much to consume today rather than save (Laubach and Williams, 2003)8. The observed decline in both interest rates and real GDP over the past two decades provides suggestive evidence that economic growth and time preferences are closely related. However, longer sample periods show a more tenuous relationship between the two (Hamilton, et. al, 2015)9. Moreover, during the 1970’s, arguably the closest analog to the current macroeconomic backdrop in the United States, interest rates were high at a time when real GDP growth was relatively low.
As the above indicates, the data leaves substantial room for shifts in savings and investment preferences in explaining the neutral rate. Demographic trends are a widely cited factor to account for the observed decline in interest rates since the 1990s. Demographic shifts worked through two channels, first increasing longevity resulted in higher savings demand by workers to fund longer retirements, which reduced real interest rates (Carvalho, et. al, 2016)10. Second, aging populations and declining birth rates increased the share of the population saving for retirement rather than consuming, a trend that may reverse as dependency ratios rise (Goodhart and Pradhan, 2017)11. At the same time, emerging market countries were running large current account surpluses and building precautionary savings in response to the financial crises of the 1990s, further adding to the ‘global saving glut’ (Bernanke, 2005)12. Exacerbating the effects of increased demand for savings was a shortage in the supply of safe assets, particularly following the financial crisis (Caballero, et. al, 2017)13. Other research also finds a role for higher inequality, lower public investment, and higher risk premia in explaining shifts in demand for savings and lower real interest rates (Lukasz and Smith, 2017)14. While the jury is still out on the precise cause of the decline in the neutral rate, the consensus view is that it is low relative to history and therefore provides a meaningful headwind to higher interest rates in the U.S.
Bringing it all back to the present, the economic situation confronting the FOMC today is very different from the last time the Committee began hiking interest rates in 2015. U.S. economic growth is remarkably strong with the median estimate for real GDP in 2022 in the Fed’s Summary of Economic Projections (SEP) at 2.8 percent, following growth of 5.5 percent in 2021. The U.S. economy has nearly caught up to its pre-COVID-19 growth trend (Figure 4). The unemployment rate is at 3.6 percent, below the SEP’s long-run median, with headline and core PCE inflation at 6.4 and 5.4 percent, respectively. Compare that with 2015 where unemployment was above the Fed’s own long-run estimate according to the SEP at the time, additionally subsequent labor market developments indicate that the FOMC was overestimating the equilibrium level of the unemployment rate (Crump et. al, 2019)15. Global growth momentum had slowed at that point and, unlike the present, the Fed was embarking on an asynchronous hiking cycle relative to other central banks. And finally, the U.S. economy is in a far better place at this point in the cycle relative to the 2015 hike and high inflation is a global problem prompting action from many central banks.
Given the underlying strength of the U.S. economy, the Fed is behind the curve i.e., the policy rate is far below neutral, relative to similar points in past hiking cycles. Chairman Powell admitted as much at the March 2022 FOMC press conference, “in hindsight, yes, it would have been appropriate to move earlier”. This is a deliberate result of both the Fed’s decision to adopt a flexible average inflation targeting regime in 2020 as well as underestimating both the durability of supply chain issues and the lagged effects of easy fiscal and monetary policies. Focusing on the former, as we hope forecast errors will not be so one-sided going forward, the 2020 strategy review stemmed from both a reduction in policy space due to the decline in real interest rates combined with the Fed’s reticence to use negative rates, and the uncertainty associated with the aforementioned ‘navigating by the stars’.
The main conclusions of the review were that the Fed would not act pre-emptively when unemployment fell below estimates of NAIRU and tighten policy. Furthermore, due to the existence of the zero lower bound on interest rates, the Committee decided to target inflation moderately above 2 percent after inflation has been running persistently below 2 percent—as it did for much of the post Global Financial Crisis period—to achieve inflation that averages 2 percent over time (FOMC, 2020)16. Now a new policy framework is added to the list of unprecedented features of the economy as the Fed lifts off from zero.
With the Fed behind the curve, current market pricing suggests the FOMC will hike interest rates rapidly to a similar peak i.e., terminal rate, as the 2015 cycle (Figure 5). Short-term interest rates indicate roughly 9 rate hikes in 2022 and potentially three more in 2023, exceeding the hike per meeting cadence of the 2004 hiking cycle and far above the quarterly hikes that characterized the second half of the 2015 cycle. These calculations are assuming that market pricing reflects the market’s true expectations, which may not be the case. They ignore the impacts of risk premia, which are forms of compensation that can exist for several reasons including hedging demand to offset riskier assets like equities, macroeconomic uncertainty, liquidity needs, mandate constraints, regulatory incentives, among other possible factors (Bernanke, 2015)17. For the last few hiking cycles, the ex- post realized risk premia has generally been positive as the Fed underdelivered on interest rate hikes relative to market pricing (Figure 6). An exception to this was the end of the 2018 cycle when the Fed concluded at a higher rate than markets were pricing.
Leaving aside the current or appropriate amount of risk premia in interest rates for now, we think the evidence points to a similar or higher terminal real rate for the current cycle relative to 2015. On the structural determinants of neutral rates outlined above there are compelling arguments on both sides about whether they are higher or lower since the pandemic, leading us to make the conservative assumption that they net to zero. The productivity enhancing investments in IT, the potential for greater attachment to the labor force through remote work, higher investment due to the onshoring of supply chains, and the renewed commitment to countercyclical fiscal spending across the globe are just a few of the pandemic-driven changes that could support a higher real neutral rate.
On the other hand, higher household savings and wealth, exits from the labor force due to the pandemic, and larger debt levels increasing the economy’s interest rate sensitivity could yield a lower neutral.18 We think the risks of a higher terminal rate stem from the cyclical strength of the U.S. economy due to the magnitude and duration of easy monetary and fiscal policy rather than any structural changes to the neutral rate since the pandemic.
There are 2 reasons why a higher terminal rate may be necessary this cycle: (1) the high level of realized inflation mechanically requires higher nominal rates to achieve the same real outcomes (2) the FOMC may need to move interest rates into restrictive territory to return inflation to target. The first reason conveys the idea of the Fed ‘leaning against the wind’ i.e., the central bank adjusts nominal interest rates by more than the changes in inflation to affect real interest rates. Using the last cycle as an example, the Fed’s target range for the policy rate peaked at 2.5 percent in December 2018 with core PCE inflation at 2 percent, implying a 0.5 percent real rate. Based on the March 2022 SEP projections, the same real rate requires an additional interest rate hike this cycle given higher inflation. However, it is the second scenario yielding a higher terminal rate that we find concerning, the potential need to raise rates into restrictive territory.
Some additional detail on the Phillips Curve is warranted as it is the central mechanism that might cause the FOMC to move policy into restrictive territory. Simply, the Phillips Curve describes the relationship between wages or prices, and unemployment with higher wages stemming from lower unemployment and vice versa. The intuition for this relationship is that greater demand increases the price of labor, wages, which then affect the overall level of prices commonly referred to as a wage-price spiral. This relationship is augmented with inflation expectations wherein more anchored inflation expectations imply less sensitivity between the two i.e., inflation’s beta to unemployment is reduced. Numerous empirical studies using data from recent decades show that both the slope of the Phillips Curve declined, and inflation expectations became more anchored.19 Successful inflation targeting now increasingly relies on well-anchored inflation expectations rather than deviations in unemployment from NAIRU. This is a reason why the Fed felt comfortable changing its approach to achieving full employment to address shortfalls rather than deviations from maximum employment.
The Fed may find itself a victim of its own success in diminishing the relationship between economic slack and inflation if inflation expectations become unanchored. Different measures of long-term inflation expectations, while higher, appear to remain well anchored around 2 percent (Figure 7). However, the potential fragility of these expectations in the face of once-in-a-generation inflation is unknown.20 Indeed, recent research at the Fed argues that well-anchored inflation expectations may be coincidental to falling inflation rather than an explanation for it. What follows is an alternative hypothesis that argues that there is a threshold below which inflation is no longer salient in wage negotiations. One implication for policy is that low inflation is then self-reinforcing (Rudd, 2021)21. The risk is that the current elevated level of inflation could imperil this equilibrium.
Before constructing a case for why restrictive policy may be necessary, here is a definition for what we mean. Restrictive policy means the FOMC increases interest rates sufficiently above neutral to slow growth meaningfully to reduce inflation. What does that mean in the context of this hiking cycle? First, the labor market is very strong. In Powell’s own words at the March 2022 meeting, “[it’s] a very, very, tight labor market, tight to an unhealthy level, I would say.” Second, inflation is, in Powell’s words again, “running above the level that would be consistent over the long run with 2 percent inflation.” The initial conditions for this hiking cycle are then an economy overshooting on both of the Fed’s goals.
In our definition of restrictive policy, there is ambiguity in what sufficiently above neutral means and how much growth will slow with the risk being a Fed-induced recession. The optimistic scenario today assumes inflation expectations are both well-anchored and central to the inflation process. Therefore, as the supply and demand imbalances created by the pandemic fade, inflation should return to 2 percent without requiring a substantial tightening from the Fed. This is the Fed’s baseline scenario drawing from the March SEP. Through a series of hikes that leave it at most 1 hike above neutral, the Fed returns core PCE inflation to 2.3 percent by 2024 with unemployment increasing only 0.1 percentage points and growth remaining above trend.
The pessimistic scenario arises when inflation expectations are either above levels consistent with 2 percent inflation or are not a major determinant of realized inflation. In the former case, the Fed takes aggressive action to lower inflation. Given inflation’s lower sensitivity to labor market slack that we established above, this scenario requires policy rates well above neutral to effectuate an increase in unemployment sufficient to slow inflation.22 We think this means as many as 4 hikes above neutral and an unemployment rate above 4 percent. Applying the Sahm rule23, this large of an increase in unemployment is likely to result in a recession.24 In the other case, the implications of a missing causal link between inflation expectations and inflation are less clear as we do think there is a relationship. Our best guess is that it would force the FOMC to move policy further above neutral and cause a deeper recession unless an exogenous shock, e.g. an AI revolution that causes a productivity boom, returned inflation to an unremarkable level.
Any interest rate view depends on the expected path of policy, the distribution of outcomes around the modal path, and the value placed on each outcome. Our expected path for rates is like the Fed’s view. We think they can accomplish their goal of a soft-landing by taking real rates near 0.5 percent equating to a nominal rate between
2.5 and 3 percent in 2023. We do not expect higher inflation to persist at a sufficiently high level to reverse the low and stable inflation paradigm of the last few decades. Despite our base case, we are sympathetic to the view that inflation may be more volatile going forward with the pandemic re-arranging supply chains for resiliency and the transition to green energy seeing bumps along the way. However, we think the market is mispricing both the distribution of policy rate outcomes and the portfolio benefits of holding long interest rate positions at the front of the curve. Either case suggests front-end rates should increase further.
While it is impossible to separate expectations from risk premia with certainty, our current read on the front-end is that markets are pricing in a lower terminal rate and risk premia is slightly positive. We also think markets are underestimating the probability of higher rate outcomes (Figure 8). As to the other leg of our higher front-end rates view, we have conviction that risk premia at the front-end should be elevated because we see higher rates associated with worse economic outcomes. Scenarios where front-end rates increase more than our modal path entail higher unemployment, lower equity returns, and higher chance of recession as the Fed fights inflation. This runs counter to the past two decades where deflation has been the paramount concern during recessions and risk premia in bonds has been low or negative (Campbell, et. al, 2019)25. If a recession is likely to be coupled with high inflation this cycle, fixed income securities will provide fewer diversification benefits. As a result, we think investors should be compensated for bearing this risk with higher yields at the front-end.
Valuation at the long end of the curve depends on how sustained high inflation is, and the timing and duration of the next recession. On the former, we believe the Fed remains willing and able to dampen inflation. We think the probability of a recession in 2022 is low given the underlying strength of the U.S. economy and the Fed’s distance from neutral. However, the risk of a recession in 2023 and/or 2024 is rising as the Fed faces the difficult task of reigning in inflation without over tightening. In terms of scenarios, we assign a 55 percent probability to a soft landing with a terminal rate between 2.75 and 3 percent. We assign a 40 percent probability to a Fed- induced recession caused by a terminal rate around 3.5 percent. Finally, we assign a 5 percent probability to a significant deceleration of inflation to below trend levels resulting in a terminal rate near 2 percent. Putting it all together, we are moderately underweight long-end rates as the risk of a recession remains low in the near-term with upside risks to inflation.
An assertive Fed and above target inflation is a tough setup for risk assets. While financial conditions have tightened significantly since the start of the year, we think the distribution remains skewed towards further tightening. The risk of a recession is higher now and the Fed’s willingness to offset any shock in the near term, the so-called Fed put, is far away until inflation falls towards 2 percent. As a result, we favor stocks with growing and less volatile earnings profiles. In credit, we expect more volatility ahead, and we remain underweight as equities provide more leverage to earnings growth. Emerging market assets are at attractive levels as local central banks face a more daunting inflation outlook than the Fed and have tightened policy accordingly. If U.S. inflation declines without a sharp slowdown in growth, we expect emerging markets to disproportionately benefit through easier financial conditions and growth spillovers. Zooming out, we encourage investors, like the Fed, to remain nimble with a rapid hiking cycle ahead.
Legal Notices & Disclosures
1 The expectations theory of interest rates describes how long rates are equal to the average of expected short-term interest rates, set by the Fed, plus a time-varying risk premium (Shiller, Robert J., John Y. Campbell, and Kermit J. Schoenholtz, 1983, “Forward Rates and Future Policy: Interpreting the Term Structure of Interest Rates,” Brookings Papers on Economic Activity, 1:1983).
2 We focus on the short-run neutral rate as it determines the relative stance of monetary policy. For a discussion of the long- term neutral rate see Roberts, John M., 2018, “An Estimate of the Long-Term Neutral Rate of Interest,” Board of Governors of the Federal Reserve System FEDS Notes, September 5.
3 Ball, Laurence, Sandeep Mazumder, 2015, “A Phillips Curve with Anchored Expectations and Short-Term Unemployment,” IMF Working Paper, 15/39, February.
4 The level of potential real GDP is approximately equal to productivity growth plus labor force growth.
5 Williams, John C., 2003, “The Natural Rate of Interest,” Federal Reserve Bank of San Francisco Economic Letter, no. 32, October 31.
6 This stylized example describes an economy in the long run. In the short run, the central bank can affect real outcomes due to the existence of nominal price rigidities e.g., wages are slow to adjust to increasing inflation and result in declining real wages.
7 Powell, Jerome H., 2018, “Monetary Policy in a Changing Economy,” Remarks at the Federal Reserve Bank of Kansas City’s Symposium on “Changing Market Structure and Implications for Monetary Policy” August 24.
8 Laubach, Thomas, John C. Williams, 2003, “Measuring the Natural Rate of Interest,” Review of Economics and Statistics, 85, no. 4, November.
9 Hamilton, James D., Ethan S. Harris, Jan Hatzius, and Kenneth D. West, 2015, “The Equilibrium Real Funds Rate: Past, Present, and Future,” NBER Working Paper, No. 21476, August.
10 Carvalho, Carlos, Andrea Ferrero, and Fernanda Nechio, 2016, “Demographics and Real Interest Rates: Inspecting the Mechanism,” Federal Reserve Bank of San Francisco Working Paper Series, no. 5, March.
11 Goodhart, Charles, Manoj Pradhan, 2017, “Demographics will Reverse Three Multi-decade Global Trends,” BIS Working Papers, no. 656, August.
12 Bernanke, Ben S., 2005, “The Global Saving Glut and the U.S. Current Account Deficit,” Remarks at the Sandridge Lecture, Virginia Association of Economists, March 10.
13 Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier Gourinchas, 2017, “Rents, Technical Change, and Risk Premia: Accounting for Secular Trends in Interest Rates, Returns on Capital, Earning Yields, and Factor Shares,” NBER Working Paper, No. 23127, June.
14 Rachel, Lukasz, Thomas D. Smith, 2017, “Are Low Real Interest Rates Here to Stay?” International Journal of Central Banking, September.
15 Crump, Richard K., Stefano Eusepi, Marc Giannoni, and Aysegul Sahin, 2019, “A Unified Approach to Measuring u*,” Brookings Papers on Economic Activity, Spring.
16 Federal Reserve Open Market Committee, 2020, “Statement on Longer-run Goals and Monetary Policy Strategy,” August.
17 Bernanke, Ben S., 2015, “Why Are Interest Rates so Low, Part 4: Term Premiums,” Brookings Institution Blog, April.
18 For a historical view on how pandemics affect real interest rates see Jorda, Oscar, Sanjay R. Singh, and Alan M. Taylor, 2020, “The Long Economic Hangover of Pandemics,” IMF Finance & Development, vol 57, no. 2, June.
19 Hazell et. al (2020), Williams (2006), Del Negro et. al (2020)
20 Malmendier and Nagel (2015) find that differences in experienced inflation strongly predict differences in inflation expectations.
21 Rudd, Jeremy B., 2021, “Why Do We Think that Inflation Expectations Matter for Inflation? (And Should We?),” Board of Governors of the Federal Reserve System Finance and Economics Discussion Series, September.
22 This scenario does not preclude changes in the slope of the Phillips Curve as result of the increase in inflation expectations. Presumably, inflation expectations would be more responsive to realized inflation in this scenario and steepen the Phillips Curve as result. However, we think the costs of unanchored inflation expectations would be significant regardless of the slope assumption.
23 The Sahm Rule identifies signals related to the start of a recession when the three-month moving average of the national unemployment rate (U3) rises by 0.50 percentage points or more relative to its low during the previous 12 months.
24 The start of a recession occurs when the three-month moving average of unemployment rate rises at least 0.5 percentage points relative to its 12-month low.
25 Campbell, John Y., Carolin Pflueger, and Luis M. Viciera, 2019, “Macroeconomic Drivers of Bond and Equity Risks,” Journal of Political Economy, 128, August.
The views expressed herein are those of Harbor Capital Advisors, Inc. investment professionals at the time the comments were made. They may not be reflective of their current opinions, are subject to change without prior notice, and should not be considered investment advice. The information provided in this presentation is for informational purposes only.
Performance data shown represents past performance and is no guarantee of future results.
Investing entails risks and there can be no assurance that any investment will achieve profits or avoid incurring losses.
Harbor Capital Advisors, Inc.