When A Storm Is In The Offing: Fundamental Growth In The U.S. Equity Market by Michael Aked, Michele Mazzoleni & Omid Shakernia – Research Affiliates
- Nearly one-third of the time, equity market momentum can be misleading in predicting future dividend growth.
- When investors rely solely on historical price returns to inform their future investment opportunities, they are at risk of allowing random price fluctuations to drive perceptions, going to cash too late and staying in cash too long. Broadening decision-making inputs to include economic data can improve forecasting ability.
- The last several years have been smooth sailing in the alignment of real stock price growth and real dividend growth, but investors would do well to check which way the economic wind is blowing; it may be time to chart a different course to avoid a storm in the offing.
“So tremendous had the force of the sea been when it broke the ship, that it had beaten one great ingot of gold, deep into a strong and heavy piece of her solid iron-work: in which, also, several loose sovereigns that the ingot had swept in before it, had been found, as firmly embedded as though the iron had been liquid when they were forced there.”1 Charles Dickens – from The Uncommercial Traveller, 1867
Today we enjoy all manner of weather prediction, from daily temperature forecasts and rain probabilities to warnings of severe storms replete with heavy winds and wave surges. Nonetheless, extreme storms that impact life and property are fairly infrequent, giving us the paradox of prediction: forecasts are mostly boringly mundane, but sometimes crucially important. The luxury of such forecasts is a relatively recent phenomenon.
Late in the afternoon of October 25, 1859, the largest storm of the nineteenth century ripped out of the Irish Sea and up the west coast of Wales, England, and Scotland. A total of 133 ships were sunk and 90 more damaged. Some 450 lives were lost on the Royal Charter, a steam clipper nearing the end of its journey from the goldfields of Australia. The Royal Charter Gale, as the storm came to be called, unleashed its fury just as the newly invented telegraph was enabling swift communications between distant places. As Peter Moore recounts in The Weather Experiment, the need to avert another such catastrophe impelled the development of modern weather forecasting, and the telegraph was the cutting-edge technology that made it possible.
The forecasting of equity fundamentals, such as earnings and dividends, suffers from the same paradox of prediction as the forecasting of weather. Long periods of sameness are punctuated with infrequent bouts of tumultuous change that can abruptly destroy livelihoods and toss accumulated wealth across the sands for others to claim as their own. During times of calm in the market and in the economy, the classic dividend discount model works, and stock prices provide significant information about fundamental dividend growth.2 Yet as we shall demonstrate, nearly one-third of the time, the steady-as-she-goes approach proves imprudent. Just as the information provided by the telegraph improved weather forecasting, macroeconomic data can be helpful in mitigating the risks of price momentum. In particular, we find that macroeconomic information can forecast when the market’s intelligence is not useful.
U.S. Equity Market – The Calm
As a means to create better storm prediction for the equity markets, we can begin by investigating the knowledge factored into market prices. In January 2016, Adam, Marcet, and Nicolini (henceforth, AMN) published a Journal of Finance article titled “Stock Market Volatility and Learning.” AMN show that the majority of stock market volatility is generated by the learning process around the future growth rate in fundamentals, such as dividends. Simply put, investors do not intrinsically know the long-term growth rate of equity earnings or dividends, but have to learn about it. Recent stock price movements prove to be a good way to model future stock price behavior. The elegant model proposed by AMN generates market-mimicking momentum and mean reversion in the stock price as a byproduct of the learning process.
Seeing recent price movements as a useful indicator of prospective dividend growth helps us think about the extent to which price changes might be fundamentally justified. Figure 1 shows that trailing real-price changes in the U.S. equity market are positively correlated (R2=0.41) with next year’s dividend growth.3 Most of the time, the market is knowledgeable and the recent price trends are the best indicators for tomorrow. But in many instances—about one-third of the time in the 1949–2015 period we study—we observe a disconnect; that is, the sign of equity market returns and subsequent dividend growth differs.
For the period 1949–2015, each percentage point increase in price of the U.S. equity market is associated with a positive 13-basis-point change in the dividend growth rate in the coming year.4 The deviation of dividend growth rates from their long-term averages is also persistent. If last year’s dividend growth was above trend, about 40% of the excess growth will persist into the following year. Figure 2 shows the inferred response in dividend growth to a one-year price change, given that growth was at trend. For example, a 20% price rise will cause the expected real-dividend-growth rate to rise from the 1.5% average to nearly 4% in the next year. The growth rate will slowly revert to trend over the following several years. Likewise, a 20% reduction in the market’s value implies a real-dividend-growth rate of ?1.25% the following year. Historically, deviations in growth from long-run averages decay over a period of years.
Clouds on the Horizon?
At times, the risk to your ship, passengers, crew, and cargo differs from what fellow mariners’ activity would suggest. Take heed. In such times, chart your own course to the nearest port. No less importantly, when weather conditions once again turn favorable—and others are still sleeping it off in the alehouse—the swiftest to move will collect the largest bounty. Similarly, when investors allow historical price returns to be their sole source of information about future investment opportunities, they run the risk of allowing random price fluctuations to drive perceptions, going to cash too late and staying in cash too long.
In this domain, however, the pursuit of knowledge is a bumpy ride. Not a well-behaved, self-limiting process, the relationship between current prices and future dividend growth is dynamic: prices predict growth, which begets price changes. AMN demonstrate that the dividend growth model is inherently unstable. Accordingly, the AMN model—just like markets—displays characteristics of bubbles and crashes, a pleasing replication of the observed market phenomena. As Figure 1 shows, mean reversion sets in up to one-third of the time. In calm periods, when investors are lulled into confidence that the sun will shine again tomorrow, market-price changes may differ in sign from future dividend growth rates. Many of these cases are small aberrations, similar to a passing shower on an otherwise-crystal-clear day. Less frequently—but far more consequentially—other cases result in unforecasted changes in the environment that can prove very expensive, either through direct loss (by staying at sea too long) or opportunity cost (by staying in port too long).
Exactly how expensive are unforecasted changes? Table 1 displays the statistical incidence and investment impact of agreement and disagreement in historical price