One common financial fallacy is for people to divide the stocks outstanding by their volume and call that the average holding time. This makes today's average holding period of a stock seem very small, and the resulting chart can be used by those who want to paint today’s stock market as more short term than ever.
The below tweet is not an atypical example of this phenomena.
When I found the source of this chart, a newsletter by Alan Newman, the argument that it was buttressing was not surprising.
“Thus, we are making the case that every year from 1997 may fairly be presumed manic in nature.”
Except that these charts are absolutely the wrong way to look at things. The rise of high frequency trading and the decrease in trading costs thanks to electronic trading has meant that market makers are trading a lot more than they have in the past. Thanks to the improvement in technology, transaction costs for investors have fallen across the board. Measured stock market volatility has also gone down, though there are likely tail risks associated with this volatility decrease such that the worrywarts are not completely wrong. But either way, the rise of high frequency trading says nothing about the changing time preference of today’s investors.
So how should we measure the time horizon of today’s investors? One way is to look at the median holding period of non-market making investors. It turns out that someone has already done this work, it just doesn’t get as much airtime as the average holding period because the data doesn’t tell a crazy story about investors. Martijn Cremers and Ankur Pareek have saved us a lot of trouble, because in 2014 they published a paper introducing the concept of Stock Duration.
Stock Duration is a measure of how long institutional investors, who have to report their quarterly holdings to the SEC, have held a stock. The paper focuses more on quantitative investment strategy, but they included a measure of Stock Duration over time at the end.
Source: Martijn Cremers and Ankur Pareek, Short-Term Trading and Stock Return Anomalies: Momentum, Reversal, and Share Issuance April 17, 2014. Figure 1, Panel A.
Their sample looked at the median duration of the largest 1300 stocks most commonly held by institutional investors. And the median stock duration increased during the period when high frequency traders started drastically increasing trading volume.
In another duration focused paper, they analyzed how duration and other factors drove the returns of US mutual funds with over $10 million AUM. In order to accomplish this, they calculated a Fund Duration metric. Fund Duration measures the average holding period of each stock in a fund and requires that a fund have at least two years of history. The aggregate of this metric again counters the narrative that investors are becoming more short term oriented.
Source: Martijn Cremers and Ankur Pareek, Patient Capital Outperformance: The Investment Skill of High Active Share Managers Who Trade Infrequently, 2015. Appendix A, Table A3
When we look directly at mutual funds, they are if anything becoming more long term oriented.
Source: Martijn Cremers and Ankur Pareek, Patient Capital Outperformance: The Investment Skill of High Active Share Managers Who Trade Infrequently, 2015. Appendix A, Table A3
It’s possible that mutual funds are now furiously trading in and out of their stocks between SEC quarterly reporting periods, but the relatively stable fund turnover ratio suggests otherwise. So while investors will always do many stupid short sighted things and many still have incentives driving them towards short term oriented thinking, the somewhat widespread idea that more investors than ever are thinking short term is probably not correct.
I recently came across a paper by Cremers, Pareek and Sautner on Stock Duration, Analyst Recommendations, and Overvaluation. In this paper they look at how stock duration, the holding period of institutional investors, interacts with analyst recommendations.
One of many interesting findings was a twist on the classic finding of stock underperformance or outperformance driving the analyst rating, rather than vice versa:
But interestingly enough, when stock analysts are late to the party of a stock with long term holders, the stocks still tend to do well.
However, when fast money is holding the stock then an analyst's extreme buy recommendation has on average marked the peak of the stock’s outperformance. The return of fading these optimistic short term investors were found to be larger if positions are not entered for at least 3 months after the signal.
On the downside, extremely bearish analysts are often too pessimistic, but the effect is larger when there is a lot of fast money involved.
So while investors have many reasons to value contrarianism, it’s important to note that fighting against long term bulls does not necessarily result in a favorable outcome. It’s better to be bullish in the face of pessimists. But if you must be bearish make sure that the people on the other side of the trade are the types who are trying to make a quick buck.