Chief Investment Strategist
Monetary policy is currently top-of-mind, as the global monetary policy cycle is slowly transitioning from widespread accommodation toward selective normalization. What should investors expect while they are expecting the start of tighter monetary policy?
We believe the European Central Bank and the Bank of Japan will continue their easy policy over the intermediate term; however, market expectations are building around the timing of rate increases from the Bank of England and the Federal Reserve. This cycle is unlike any in the record books. Global central bank balance sheets have increased to $10 trillion in the last six years through their efforts to combat the financial crisis and the risk of deflation. The resulting ultra-low interest rate environment has led investors to take on greater credit and duration risk in pursuit of income, raising the risk of asset/liability mismatches and the resulting financial market upset should interest rates rise quickly. It is our expectation that global growth and inflation will remain moderate in comparison to historical periods, allowing the central banks to shrink their balance sheets at their own pace. The onset of a tightening cycle has historically led to a pickup in volatility, and this time should be no different especially since the drop-off in trading by investment banks and brokers has significantly reduced market liquidity. However, if the Fed raises rates at the measured pace that we expect, financial market conditions should still be supportive of risk taking. We also think the expectations of higher rates built into fixed income markets are sufficient to allow bonds to generate positive total returns over the intermediate term.
EXHIBIT 1: STOCKS HAVE HANDLED INITIAL RATE HIKES WELL
Sources: Northern Trust, Bloomberg, Barclays Capital. Intermediate bonds: BarCap U.S. Government/Credit Index; long bonds: BarCap U.S. Long Government/Credit Index. See appendix for details of the cycles.
As shown in Exhibit 1, stocks generated positive returns in four of the five tightening cycles between 1986 and 2004 with the exception being the 1988 cycle. The Black Monday crash of October 19, 1987, where global stocks swooned (the Dow Jones fell by 23% in a day), is included in this cycle as it occurred within six months of the first rate hike. Six months after the crash, the Fed started an aggressive rate cycle by raising the fed funds rate by 3.25% over the course of 12 months. More consistent with the other cycles, from this point forward the stock market generated a 32% return during this tightening period, as a 41% annualized earnings growth rate helped offset some contraction in price/earnings multiples.
Today, there is considerable debate about what conditions will be required before the Fed starts hiking interest rates. Indeed, the Fed itself has contributed to this uncertainty through their evolving guidance on the topic. The Fed had previously cited a 6.5% unemployment rate as one guidepost toward raising rates, but we breached that level in April 2014. The uncertainty about the state of the labor markets is such that it is the primary topic at this years Jackson Hole Economic Policy Symposium. In Exhibit 2, we show key economic indicators and their behavior in the six months before and six months after the initial fed funds target rate increases of the last five tightening cycles. The top two panels address the Feds goals of maximum employment and stable prices. The bottom two charts display two factors that we think are also important in the Feds evaluation of the economic cycle capacity utilization and wages, which serve as warning signs of potential inflationary pressures.
EXHIBIT 2: THE DATA DONT YET CALL FOR TIGHTENING
Sources: Northern Trust, Bloomberg. *Placement of current notch is unrelated to expected timing of the Feds first hike. Historic wage growth: production and non-supervisory workers; current wage growth: total private sector workers.
Inflation is currently well below the levels at which prior rate hikes began. We characterize several of the Fed tightening cycles as reactionary, as they began when inflation was high or accelerating (1986, 1988 and 2004). While overall inflation wasnt too high in 2004, we believe the Fed was being reactive as core personal consumption expenditures (PCE) were up 2.3% at an annual rate in the six months before the hike, and were projected to continue increasing. The 1999 cycle was proactive from an inflation standpoint, as core PCE was growing below the Feds 2% stated goal and the Fed was moving ahead of a forecasted jump in inflation and growth. Through these cycles, the Fed has found itself being both reactive and proactive toward inflation. We expect this Fed to favor growth stability over inflation risk, as they worry about the durability of the recovery, and therefore tolerate the risk of higher inflation before starting to tighten policy. We also expect a benign inflation outlook to get them plenty of cover.
While the U.S. unemployment rate has fallen from 10% to 6.2% in the last five years, we believe this overstates the health of the labor markets. The quality of job creation has been weak, and the participation rate has fallen to 63%, a low not experienced since 1978. A broader measure of unemployment, the U-6 unemployment rate, stands at 12.2% above the 10.7% average for the last 20 years and above the 8.9% average from 1994 through the start of the 2008 financial crisis. Reflecting this weak environment, wage gains have averaged just 2.5% in 2014 for production workers, and a meager 2.0% for all private workers (including supervisors). The current pace of wage growth, as the black notch in the wages panel of Exhibit 2 shows, is below the level that wages were growing at the start of each of the prior five fed funds rate tightening cycles. We think sustained wage gains will be required before the Fed gets aggressive on raising interest rates. The capacity utilization figures are probably the least impactful, but they do show the current level of 79.2% approaching the 30-year median of 80.2%. Reflecting the relatively benign inflation and wage figures, the fed funds futures market first signals a hike in July 2015 and has the hike fully priced in by October 2015.
EXHIBIT 3: MOST ASSET CLASSES HAVE DONE WELL DURING RATE HIKES
Sources: Northern Trust, Bloomberg, Barclays Capital. Emerging market equities and global real estate data unavailable for 1986 and 1988.
Risk assets have typically fared well in the 12 months that straddle the first hike in a period of policy tightening. Historically, the stock market has had some level of correction in the first couple of months after the first rate hike, but resumed its upward trajectory as investors arent yet concerned that the Fed could go too far and push the economy into recession. Over the course of the full tightening cycle, earnings growth is the key contributor to stock returns as valuations have tended to compress. The negative returns from 1988 stand out, but as discussed earlier, they are attributable to the October 1987 crash as opposed to the fed funds cycle. As shown in the Appendix, price/earnings multiples have fallen on average by nearly 5% across all five cycles while earnings have increased at an annual rate of 19% reflecting the continued growth in the economy. Developed ex-U.S. equities, emerging market equities and global real estate have outperformed U.S. equities during prior initial rate hike cycles, reflecting the different monetary policy cycles. High-yield returns have been positive through tightening cycles as the improving economy has historically helped improve credit quality and therefore reduce spreads.
We believe the effect of rising Fed policy rates on fixed income returns is not always fully understood. Understanding future fixed income asset class returns requires considering future expectations for interest rates or what is already priced in to the bond market. For instance, using the current forward curves (left hand panel of Exhibit 4) we determine that a two-year bond currently yielding 0.47% is really a collection of a one-year bond yielding 0.11% and a one-year bond in one year yielding 0.89%. In other words, the market expects one-year rates to go up to 0.89% over the next year and, should this occur, our original two-year bond will maintain its price level, despite the upward shift in rates. This exercise gets more involved as we look at bonds further out the yield curve and even more involved when we consider forecasting returns for bond indexes, which have multiple bonds being issued and maturing over any given time frame but the key point is that future rate expectations are a key driver of expected returns.
EXHIBIT 4: RISING RATE EXPECTIONS PROVIDE SOME CUSHION FOR BONDS
Sources: Northern Trust, Bloomberg. Current data as of 8/22/2014.
Looking at future rate expectations, we find that the market has priced a steady level of rate increases over the next one, three and five years with the 3-month U.S. Treasury going from its current near-zero yield to nearly 3% by 2019. Lets illustrate the discounting of future interest rate expectations in another way. A 10-year bond, yielding 2.40% today, will be a five-year bond in five years. Priced into the markets is a five-year yield in five years of approximately 3.2%; meaning the current five-year rate of 1.63% could double over the next five years without a negative impact on the price of our original 10-year bond. These increases in future expected interest rates have resulted in a yield curve that is steeper than any of the previous rate cycles we studied, with the exception of 2004. This can be seen in the right hand panel of Exhibit 4, which shows the yield curve steepness of each cycle broken out by the spread between the three-month and two-year, two-year and 10-year, and 10-year and 30-year. The bigger the spread, the bigger the cushion built into the markets for rate increases. This is not to say that rate increases cannot cause negative price movements; just that, for negative price movements to occur, increases in interest rates need to outpace built-in expectations. Historically, returns have been cushioned by higher coupon levels. In 2004, the Barclays Capital Long Government/Credit Index provided a 2.3% one-year price return in the face of a Fed that increased the fed funds rate from 1% to 3% over the course of a year. Adding the 6.3% coupon to the 2.3% price return gets you the 8.6% return seen in Exhibit 1. The lower level of nominal rates reduces this cushion, but we also expect the volatility in interest rates to be lower than historical levels.
Returning to the current environment, returns from fixed income are still expected to be low by historical standards because of the low starting point of interest rates. However, we do not anticipate material adverse price impacts from increasing rates. In fact, we continue to believe that the rate normalization process will occur at a slower pace than the market believes. We have a relatively benign outlook for asset classes out the credit spectrum as well; and view concerns over current yield levels in High Yield as somewhat misplaced. It is true that the current 5.25% yield on the Barclays Capital High Yield index is near historical lows. However, this has much more to do with the low interest rate environment, rather than irrational exuberance regarding appropriate credit risk compensation. Even before the recent uptick, high-yield credit spreads were at approximately 3.25% versus lows of approximately 2.30% in 2007 and approximately 2.35% in 1997. Current high-yield fundamentals also appear solid. Default rates, currently at 1.8%, should continue to be low given cash balances and limited refinancing needs over the next few years. High Yield has historically done well in rising rate environments, and we would probably need to see a stagflationary environment (slow growth, high inflation) for this not to hold true.
EXHIBIT 5: A GOOD DEFENSE WINS CHAMPIONSHIPS
Sources: Northern Trust Quantitative Research. 1999 data excluded for scaling reasons; full data in appendix. Quality factor as defined by Northern Trust Quantitative Research.
In addition to examining the returns of different asset classes during interest rate-hike cycles, we have analyzed the performance of different types of stocks during the cycles. We focused on the most commonly utilized equity factors size, value, momentum, low volatility, dividend yield, and quality which are gaining increasing traction in portfolio management. In general, the more defensive factors (low volatility, dividend yield, quality and value) have outperformed in the year surrounding the start of the Fed rate-hike cycle, while size (small) and momentum have underperformed.
With the exception of 1999, being long these defensive factors and short size (small-cap) and momentum (what we call the tightening trade) has been a winning strategy. We believe the effect of this strategy was delayed in 1999 due to the tremendous momentum of the technology stock bubble but the strategy delivered extremely robust returns in 2000 as the tech bubble burst.
EXHIBIT 6: A STRATEGY FOR FED TIGHTENING
Source: Northern Trust Quantitative Research. *The 1999 market response was delayed; average factor return during 2000 shown. Returns six months before to six months after first rate hike. Return represents average return of low volatility, dividend yield, quality and value factors less average return of size and momentum factors.
The historical data is instructive but never fully predictive of what will happen during the next cycle. The 2004 cycle may be the closest comparison to todays environment, which suggests that a focus on defensive factors may be beneficial once the market begins to believe a hike in rates is in sight. However, we do need to take account of the current environment when considering how closely this cycle may resemble prior ones. We do expect the Fed to operate very deliberately, indicating the market is unlikely to be surprised by rate hikes. In addition, the sustained level of low rates may limit the upside to the strategy. High dividend-yielding stocks (and to a lesser extent, low-volatility stocks) are popular today as fixed-income alternatives, and have become expensive. Low volatility stocks are less expensive as they are valued near historical averages, but they have typically been cheap prior to tightening cycles. But if the Fed does find itself behind the curve, and the markets are truly surprised by the pace of Fed tightening, then the tightening trade should be more lucrative.
Investors will struggle with the potential impact of changing monetary policy in the years to come, as the pace of normalization is unclear and the impact of higher interest rates on the economy and financial markets is subject to debate. We have shown that historically both stock and bond markets can perform well once they have passed through an initial adjustment period. The key variable is the strength of the economy and the upside risk to inflation, as this is the scenario that could cause the Fed and other central banks to move sooner and with greater magnitude than the market expects. We think we are more than a year away from the first potential hike in the fed funds rate, but with the Feds bond buying program scheduled to end this October, we expect the rate debate to take center stage. Our current view of the economy leads us to think the Feds eventual measured pace of rate normalization should not be disruptive to the equity markets, and the current expectation of rising rates in the fixed income markets gives sufficient cushion to the potential of rising rates over the next five years. Finally, our work on equity factor performance shows that a focus on the tightening trade makes sense once the rate cycle begins.
Special thanks to Avantika Saisekar, Investment Analyst, for data research.
APPENDIX: DETAILS OF THE CYCLES