by P.J. DiNuzzo March 12, 2020
Here Today Gone Tomorrow: The Impact of Economic Surprises on Asset Returns
Since the end of the global financial crisis, economic forecasts have clustered in a much narrower range than their pre-crisis dispersion, as shown in Figure 1. Structural forces such as demographic change have put downward pressure on growth, while extraordinarily accommodative monetary policy has provided a cushion. The U.S. economy's performance has been consistent with these constrained expectations, producing few surprises.
That may be changing. The Federal Reserve has continued to normalize monetary policy and withdraw excess liquidity, and fiscal policy is helping fuel a cyclical bounce above the structural limits. Record-low unemployment and rising short-term interest rates are the most visible signs that the post-crisis economic environment is in flux. And as common sense-and historical analysis-suggests, a narrow range of expectations is associated with a higher degree of surprise.
Economic surprises reverberate through the financial markets, producing short-term volatility in asset prices. The impact of economic surprises on returns varies by asset class. It also depends on the phase of the business cycle in which the surprise occurs. Although there is some correlation between economic surprises and asset returns in the short term, we find that in the long term, these surprises hardly matter. We use these relationships between economic surprises and asset returns to explore various implications for portfolio strategy.
More Surprise, More Volatility
Economic surprises have a weak but positive correlation with market returns, as shown in Figure 2. Since the 1970s, a 100-basis-point surprise in quarterly GDP growth has been associated with about a 70-basis-point change in quarterly U.S. equity returns.
Economic surprises can be broadly classified as positive or negative. An economic surprise is positive when the actual data exceed expectations. When the data fall short of expectations, the surprise is negative.
To understand how assets respond to economic surprises, we regress returns of four broad asset classes on an "economic surprise index."2 U.S. equities and commodities represent high-risk asset classes. U.S. Treasury bonds and cash (USD) serve as proxies for low-risk assets. Our economic surprise index is a measure of surprise in four macroeconomic variables-GDP, the Institute for Supply Management (ISM) Manufacturing Index, retail sales, and employment measures.
Different asset classes react to economic surprises, both positive and negative, to different degrees (see Figure 3). The safe-haven assets-cash and Treasury bonds-are relatively insensitive to economic surprises. The returns of equities and commodities, by contrast, are keenly sensitive.
The intuition is straightforward. The return of a governmentguaranteed, fixed income instrument such as a Treasury bond or bill is relatively easy to predict. The returns of stocks and commodities are more uncertain. An equity's return depends on its future profitability, which depends in part on the economic environment. Commodity returns depend on supply-and-demand dynamics dictated by future economic conditions. An economic surprise results in an immediate reassessment of the conditions that will determine the value of these assets. The nuances in asset performance go deeper. An economic surprise during a contraction in the business cycle has a different impact than a surprise during a recovery.
When our regression controls for the business cycle, we see that both safe-haven assets and risky assets react more to economic surprises in the recovery and contraction phases (see Figure 4). We measure the asset's response to economic surprises in terms of the number of standard deviations (z-scores) from its mean response in all economic cycles. In periods of contraction, cash posts the most significant response to economic surprises relative to its response in all environments. In absolute terms, however, cash returns are modest. In periods of recovery, commodities respond most strongly.
Although these results are statistically insignificant, they are suggestive of investor behavior during periods of pronounced change in the economic outlook (BenRephael et al., 2018). In periods of contraction and recovery, when the outlook is changing, investors are more sensitive to economic surprise, and there are wider asset price fluctuations. But when the outlook is more stable, investor reaction to economic surprise is muted.
Can Investors Capitalize on Surprise?
Economic surprises are just that-surprises. In Figure 5, we illustrate this through nonfarm employment surprises, obtained by regressing nonfarm payroll changes on the Vanguard Leading Economic Indicators (VLEI) series, a business cycle measure similar to those published by The Conference Board and the Economic Cycle Research Institute. There is no particular trend to the positive or negative surprises.
Of course, the belief that motivates tactical asset allocation strategies-indeed, any active strategy-is that a surprise to the consensus can be foreseen by a prescient analyst. How prescient would an investor need to be to capitalize on economic surprises?
We answer this question with a simple simulation based on economic data over the past 25 years:
• We start with a $1,000 investment in a base portfolio of 60% U.S. equities and 40% U.S. bonds.
• In advance of a positive economic surprise, we allocate 80% of the portfolio to equities and 20% to bonds.
• In advance of a negative economic surprise, we allocate 40% of the portfolio to equities and 60% to bonds.
Figure 6 displays the results. Not surprisingly, the portfolio of the omniscient investor (able to time all economic surprises) generates slightly better returns, outperforming the base portfolio (60/40 asset allocation) by 0.2 percentage points per year over the 25-year period. Absent omniscience, however, the results quickly deteriorate. An investor would need to successfully trade on 75% of economic surprises to earn returns similar to those of the base portfolio, 7.4% per year. If the investor had been no more prescient than a coin flipper, accurately trading on 50% of the economic surprises, the portfolio's returns would have fallen to 7.3% per year. And if the investor had gotten everything wrong? The initial investment of $1,000 would have received 7.2% year on-year returns, about 0.2 percentage points below the base portfolio. However, these returns do not include transaction costs incurred in rebalancing portfolios to take advantage of the economic surprises. Needless to say, those costs would further reduce these returns.
How achievable is a 75% success rate, the threshold for a successful timing strategy? Not very. The parallel is inexact, but estimates of security selection skill among equity fund managers (Sorensen, Miller, and Samak, 1998) suggest that a success rate of 54% equates to annualized excess returns of 2.61 %-5.59%. In the 25 years ended 2017, only 6% of the equity funds in Morningstar's database produced annualized excess returns in that range. In competitive investment markets, a success rate of 75% is unlikely.
In the Long Run, Surprises Don't Matter
The odds of capitalizing on a short-term economic surprise are long. But what about long-term portfolio strategies? Do short-term surprises hint at long-term risk-reward dynamics that can inform strategic asset allocation decisions? In a word: no. Surprises don't matter for long-term returns.
The accumulation of short-term surprises can change the longer-term outlook, raising or reducing an economy's growth prospects and, potentially, expected asset returns. But in the global financial markets' near-instantaneous arbitrage mechanism, the same surprises that are difficult to profit from in the short term are immediately priced into long-term expectations (Davis et. al., 2010). As Figure 7 demonstrates, expectations do not forecast stock market returns. The same surprises that have a visible impact on short-term market returns are irrelevant in the long term.
Conclusion
Since the end of the global financial crisis, economic performance has been consistent with investors' narrow expectations. As the post-crisis economic environment evolves, these expectations may be vulnerable to surprise.
Our analysis of the relationship between economic surprises and asset returns yields two insights: First, the odds of successfully trading on surprises are low. Second, what can seem consequential in the short run is irrelevant to the long-term investor. Short-term surprises are quickly priced into long-term expectations, and these long-term projections have almost no relationship to future returns.
Sincerely,
P.J. DiNuzzo, CPA, PFS™, AIF®, MBA, MSTx
President, Founder, and Chief Investment Officer