Oil s tipping point $150 per barrel would likely be necessary for another U.S. recession Vanguard commentary April Executive summary. Rising oil prices are arguably the greatest risk to the global economy. We consider various oil-price scenarios through, using an econometric model that not only incorporates Vanguard s proprietary leading indicators but also explicitly captures oil s asymmetric and nonlinear impacts on growth and inflation. Given the U.S. economy s building momentum, we find that West Texas Intermediate (WTI) crude oil prices would likely need to persist at $150 per barrel to generate a U.S. recession, although WTI oil prices even at $120 would likely engender a weaker-than-expected recovery (or soft patch ) later this year. Conversely, annualized rates of core personal consumption expenditures (PCE) inflation tend to rise above 2% for a time in every oil-price scenario we have considered. A direct implication of this scenario analysis is that a near-0% federal funds rate may no longer be optimal going forward, regardless of the future direction of oil prices. Authors Joseph H. Davis, Ph.D. Roger Aliaga-Díaz, Ph.D. Connect with Vanguard > vanguard.com
Oil above $100, again Since their collapse in 2009, oil prices have risen sharply, at first because of stronger demand from a rebounding global economy, and more recently out of concerns over future supply disruptions in the Middle East. On April 1,, WTI crude oil prices closed near $108 per barrel. Figure 1 illustrates that global energy markets (as of April 1, ) expected the price for WTI crude oil to remain near these levels through before declining to about $103 by the end of. 1 Although WTI oil prices in the United States have yet to approach their July 2008 highs of more than $145 per barrel, they could push markedly higher with any combination of further Middle Eastern unrest, supply disruptions in other parts of the world, or stronger-than-expected global growth. As a result of these and other wild cards, oil prices are notoriously difficult to predict and thus are a primary risk factor in any economic or investment outlook. Indeed, as economists have long observed, oil-price spikes of at least 50% or more have preceded nearly every U.S. economic recession in the past 50 years. As we detailed in a 2008 Vanguard paper, Oil, the Economy, and the Stock Market (by Davis and Aliaga-Díaz), sharp supplydriven spikes in oil prices have tended to coincide with bear markets in equities, have sparked higher inflation and government bond yields, and have presaged rising unemployment. Given oil s track record, investors are asking: At what point would higher oil prices derail the U.S. recovery? and What would that mean for inflation and short-term interest rates? We address these questions here. Assessing oil s potential economic impact Having considered multiple oil-price scenarios through, we focus on six scenarios in Figure 1: Three hypothetical scenarios in which WTI oil prices adjust permanently in the second quarter of (to $90, $120, and $150 a barrel, respectively); and Three hypothetical scenarios in which WTI oil prices temporarily spike (to $120, $150, and even $200 a barrel, respectively) in the second quarter of before eventually falling back to $105 by the end of. We then quantify how those oil-price scenarios may affect the U.S. economy by estimating a smallscale vector autoregressive (VAR) econometric model. It s important to note that this VAR model explicitly captures the fact that oil s observed impact on U.S. growth and inflation has been both asymmetric an x% increase in oil prices hurts growth more than an x% price decline helps it and nonlinear the effect of an oil price shock on growth increases proportionally more than the oil price shock itself. We also account for the fact that the U.S. economy has become somewhat less oilintensive per unit of gross domestic product (GDP) over time. Overall, the model s structure and estimation procedure are similar to that used in Davis and Aliaga-Díaz (2008), although we supplemented that model here with our proprietary Vanguard leading economic indicator (or, VLEI) to better capture the economy s current momentum. 2 Notes on risk: All investments are subject to risk. Past performance is not a guarantee of future results. 1 The U.S. Energy Information Administration () has a similar outlook for future WTI crude oil prices. 2 See Vanguard s recent paper, U.S. Economic Outlook: Cautious Optimism (Davis et al., ), for a detailed description of our leading indicators. 2
Figure 1. Some potential scenarios for oil prices Average quarterly prices per barrel of WTI crude oil $200 150 100 50 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Actual prices/ market expectations (NYMEX futures) Per barrel prices $90 oil: Permanent rise $120 oil: Permanent rise $150 oil: Permanent rise $120 oil: Temporary spike $150 oil: Temporary spike $200 oil: Temporary super-spike Sources: Vanguard, NYMEX, and Federal Reserve Bank of St. Louis. Figure 2, on page 4, presents the impact of the various oil-price scenarios vis-à-vis the current consensus forecasts for real GDP, employment, the unemployment rate, and core PCE inflation. Given that the U.S. economy has been building cyclical momentum through early, oil prices would likely need to persist at $150 per barrel to generate another U.S. recession, as indicated by the median path for real GDP and employment growth in Figure 2. In this framework, oil would need to approach $200 per barrel to return the unemployment rate to 10%. Less-severe oil-price shocks should have more moderate effects on the U.S. economy. A rise in WTI oil prices to $120 a barrel in the coming months, for example, would likely engender a soft patch of weaker-than-expected growth for real GDP and employment later this year. With WTI oil prices close to $110 in early April, that is not a large margin of error. Figure 2 also reveals that the Federal Reserve s preferred measure of core inflation has likely already bottomed, albeit from a low, sub-1% rate. Looking forward, the median path for core PCE inflation tends to rise in every oil-price scenario we have considered. Annualized rates for core PCE above 2% (and hence above the Fed s long-term price target) are likely to occur later this year, even if WTI oil prices stay at present levels. That said, core inflation above 3% tends to occur most frequently only when oil prices are above $150. 3 3 The fact that core inflation does not tend to approach 1970s-type rates, even under more extreme oil-price scenarios, is consistent with Vanguard s previous research on the evolution of U.S. inflation dynamics and the reduced pass-through of higher commodity prices on other goods and services in the economy. See, for instance, Evolving U.S. Inflation Dynamics: Explanations and Investment Implications (Davis, 2007) as well as Recent Policy Actions and the Outlook for U.S. Inflation (Davis and Cleborne, 2009). 3
Figure 2. Simulated median macroeconomic forecasts, a. Real GDP growth (annualized rate) b. Nonfarm payroll growth (average of monthly changes, in thousands) 5% 250 4% 200 3% 150 2% 100 50 1% 0 0% 50 1% 100 2% 150 c. Unemployment rate d. Core PCE inflation (year-over-year percentage change) 10.0% 3.5% 9.5% 3.0% 9.0% 2.5% 8.5% 2.0% 8.0% 1.5% 7.5% 1.0% 7.0% 0.5% 6.5% 0.0% Per barrel prices Actual measures/ $90 oil: consensus expectations $120 oil: $150 oil: Permanent rise Permanent rise Permanent rise $120 oil: Temporary spike $150 oil: Temporary spike $200 oil: Temporary super-spike Notes: Our VAR model is estimated using quarterly data since 1960 with four lags on six variables: (1) Vanguard s proprietary leading economic index, (2) annualized real GDP growth, (3) annualized inflation rate of the core PCE index, which excludes food and energy prices and is closely monitored by the Federal Reserve as a proxy for underlying trend inflation, (4) average monthly change in nonfarm payrolls, (5) unemployment rate, and (6) an (exogenous) asymmetric oil-price term equal to the percentage change between the current WTI oil price and the maximum price over the past 12 months or zero, with price declines down-weighted. We also include an exogenous and contemporaneous squared term of the (exogenous) asymmetric oil-price term to capture the potential nonlinear affects of an oil-price shock on growth and inflation. Before estimating the model, the series for this squared oil-price term is multiplied by the ratio of U.S. petroleum consumption to real GDP to capture the fact that the U.S. economy has become somewhat less oil-intensive per unit of output over time. The simulated median under each oil-price scenario was then benchmarked versus consensus expectations under the assumption that the consensus forecasts reflect an average WTI oil price of $100 over each quarter of the forecast horizon. The paths displayed represent the median from the 10,000 stochastic simulations drawn from the model; as noted, considerable forecast dispersion exists around the median paths. For further details and sources pertaining to this methodology, see our previous Vanguard paper (Davis and Aliaga-Díaz, 2008). Sources: Vanguard calculations, based on data from Philadelphia Federal Reserve, Bloomberg, U.S. Energy Information Administration, U.S. Bureau of Labor Statistics, U.S. Census Bureau, and the Federal Reserve. 4
Figure 3. Simulated federal funds rate using a Taylor rule, by oil-price scenario 7% 6% 5% 4% 3% 2% 1% 0% 1% QE1 QE2 2% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Federal funds target Actual Market s expected future With simulated Taylor rule Per barrel prices, using Taylor rules $90 oil: Permanent rise $120 oil: Permanent rise $150 oil: Permanent rise $120 oil: Temporary spike $150 oil: Temporary spike $200 oil: Temporary super-spike Notes: The actual formula for the Taylor rule we employed is available on the Bloomberg terminal (type the command TAYLOR<GO> and follow the default settings). Here, we defined the NAIRU (non-accelerating inflation rate of unemployment) as the trailing ten-year average of the unemployment rate, and we specified the Fed s core PCE inflation target as 2%. Sources: Vanguard calculations, based on data from Philadelphia Federal Reserve, Bloomberg, U.S. Energy Information Administration, U.S. Bureau of Labor Statistics, U.S. Census Bureau, and the Federal Reserve. Implications for short-term interest rates Overall, Figure 2 suggests that the U.S. economy now would be somewhat more resilient to an oil-price shock than it would have been just one year ago, although certainly not immune to an oil-price shock near or exceeding $150. Core inflation also appears to be bottoming, and is likely to rise over the course of and into. How should the Fed react? To assess the appropriate federal funds rate under these alternative oil-price scenarios, we calculated so-called Taylor rules (Taylor, 1999), based on our simulated median paths for (1) core PCE inflation and (2) the unemployment rate. Although the Taylor rules have limitations, they are widely used in the investment community to generally assess the optimal federal funds rate based on current and anticipated future trends in inflation and full employment, the Fed s two primary objectives. Of course, there can be (and indeed have been) a host of reasons why the Fed s actual target rate differs from that produced by an admittedly simple Taylor rule. Figure 3 presents the simulated Taylor-rules-based federal funds path and compares that to what the federal funds future market anticipated it to be as of April 1,. A direct implication of Figure 3 s analysis is that a near-0% federal funds rate may no longer be optimal going forward, regardless of the future direction of oil prices. Indeed, our simulated optimal federal funds rate reaches or exceeds 2% by the middle of, whether oil is below $100 (since the unemployment rate is lower) or markedly above $100 (since core 5
inflation rises above the Fed s preferred 2% rate). In this framework, the anticipated termination of QE2 (i.e., the Fed s second round of quantitative easing) in June would appear appropriate. Our calculated optimal federal funds rate paths for, on the other hand, center in the 2% 3% range. This simulated range for short-term interest rates is close to that now priced by the bond and money markets, and is notably below the average federal funds rate of 5.4% since 1954. Overall, we cautiously interpret our simulated results as consistent with an increased bias toward a rising short-term interest rate environment over time, yet at a pace that may leave the federal funds rate below its historical average for an extended period. References Davis, Joseph H., 2007. Evolving U.S. Inflation Dynamics: Explanations and Investment Implications. Valley Forge, Pa.: The Vanguard Group. Davis, Joseph H., and Roger Aliaga-Díaz, 2008. Oil, the Economy, and the Stock Market. Valley Forge, Pa.: The Vanguard Group. Davis, Joseph H., Roger Aliaga-Díaz, Andrew J. Patterson, and Charles J. Thomas,. U.S. Economic Outlook: Cautious Optimism. Valley Forge, Pa.: The Vanguard Group. Davis, Joseph H., and Jonathan Cleborne, 2009. Recent Policy Actions and the Outlook for U.S. Inflation. Valley Forge, Pa.: The Vanguard Group. Davis, Joseph H., Daniel W. Wallick, and Roger Aliaga-Díaz,. Vanguard s Economic and Capital Markets Outlook. Valley Forge, Pa.: The Vanguard Group. Taylor, John B., 1999. A Historical Analysis of Monetary Policy Rules. In Monetary Policy Rules, edited by John B. Taylor. NBER Business Cycle Series, vol. 31. Cambridge, Mass.: National Bureau of Economic Research. U.S. Energy Information Administration,. Short-Term Energy Outlook. March 8; available at www.eia.doe.gov/steo/. 6
P.O. Box 2600 Valley Forge, PA 19482-2600 Connect with Vanguard > vanguard.com Vanguard research > Vanguard Center for Retirement Research Vanguard Investment Counseling & Research Vanguard Investment Strategy Group E-mail > research@vanguard.com The Vanguard Group, Inc. All rights reserved. Vanguard Marketing Corporation, Distributor. ICROIL2 04