It’s All In The Positive Skew
I recently read an academic paper By Hendrik Bassembinder titled Do Stocks Outperform Treasury Bills? on the long-term performance of U.S. equities that was so interesting (yes, it does happen!) that I thought it important to share an overview of it with the readers of Canadian MoneySaver.
Bessembinder’s paper (which is theoretically still in draft form), while definitely “academic” in style, is thankfully largely free of higher-order mathematics, and is worth reading in full. In this article, I intend to highlight some of his key conclusions and inferences.
Importantly, while he used a number of databases, much of the paper is based upon a 90-year time series (1926 to 2015) called Centre for Research on Security Prices (CRSP), which encompasses all stocks quoted at some point during that period on the NYSE, NASDAQ or Amex stock exchanges. There are some 26,000 in all, which were responsible for an aggregate lifetime shareholder wealth creation of USD 32.1TN.
As I have written before, so-called “active” money management tends to underperform “passive” or index-based alternatives over time. Any active manager who reads this paper should be deeply afraid (or will become convinced that only he/she has the skills to beat the market.)
So, on to the key findings:
- The median listing period is just over 7 years—so survivability matters. Only 36 stocks were present for the entire 90-year period.
- Just under 48% of the available monthly stock returns exceeded the return on 1-month U.S. T-bills (i.e., 52% did not.)
- Slightly more than half of stocks had a negative lifetime return, with the median being (-3.67%).
- Barely 42% of stocks had a lifetime return exceeding the comparable return on 1-month U.S. T-bills- more bluntly, 4 out of 7 stocks do not outperform 1-month U.S. T-bills returns available over their quoted existence.
- The good news is that stock returns exhibit what is called “positive skewness”—i.e., large positive returns are more frequent over longer periods than large negative returns—and such returns tend to compound.
- However, given the median stock return is below the mean, this tends to emphasize the benefits of portfolio diversification. Standard active managers, who are not diversified, tend to underperform (with rare exceptions.)
- It is, theoretically possible to achieve lottery-like returns over time by picking the “right” stock (BRK anyone?), but the probability of doing so ex ante is vanishingly small. The median (aggregate!) single stock return was 13.1%, versus a comparable return of 1,923% for 1-month U.S. T-bills (yes, you read that correctly). More pertinent, over their lifetime, barely 28% of individual stocks outperform 1-month T-bills.
- Longer holding periods tend to produce higher returns (i.e., trading or trying to time the market is problematic) and the compounding effect of the above-mentioned “positive skew” tends to work in their favour.
- Interestingly, there is some evidence that unleveraged companies outperform more leveraged ones, if not significantly.
- While small-cap stocks with a decade holding period tend to underperform, they are more likely to produce lottery-like returns.
- Receipt and reinvestment of dividends matters. For example, GM went bankrupt in 2009, so a buy-and-hold return would have been close to -100% in principal terms. However, over its lifetime as a public company it generated USD 64BN of dividends.
- A mere 30 stocks accounted for 31.2% of the total wealth created over the 90-year period; 86 for 50%; 282 for 75%; and 983 for 100%- i.e., less than 4% of the stocks in the database were responsible for 100% of the market excess return, with the other 96% producing a return in aggregate equivalent to 1-month U.S. T-bills. Put another way, 0.33% of stocks created half the wealth.
- The single stock that generated the most wealth in absolute terms was ExxonMobil Corp (USD 940BN); while the largest cumulative return was 2 million times (apparently that is not a typo) over the 90-year period. Perhaps not surprisingly, Apple was second in value creation at USD 677BN.
So, what tentative conclusions might an investor draw from the above?
First, that you must be exceedingly fortunate to generate excess returns. Second, having found a “winner”, you should stick with it. Third, owning a broad market equity index, whether capitalization, value, or equally weighted is probably the best approach for most of us over the long term.
In other words, trying to be clever, time the market, or failing to diversify are unlikely to produce the returns that are often “promised”.
As I said in my preamble, the paper is worth reading in full.
David Ensor: Risk Management Expert;