The Virtuous Investor: Rule 15

Don’t go for short-term gain – Weigh long-term outcomes with short-term results

This post is part of a series on The Virtuous Investor. For an overview of the series and links to the other parts, click here.

“See thou compare not the grief of the fight with the pleasure of the sin: but match me the present bitterness of the fight with the bitterness of the sin hereafter which followeth him that is overthrown.”

Erasmus of Rotterdam

We all know we should be long-term oriented in our investments. Yet, there are constant temptations to give in to the sin of short-termism, especially if long-term outcomes are less than certain or if we are in a bear market. And many companies have a vested interest in providing seemingly attractive short-term solutions for investors. A financial service provider that shall remain nameless, who promotes analyst recommendations for individual stocks as part of its business tried to show that despite the poor track record of analysts’ recommendations there is value in looking at them.

In order to do so, they looked at both the recommendations of a large group of sell side research analysts and their buy and sell recommendations as well as their price targets. Unsurprisingly, they found that the buy and sell recommendations of the analysts were biased and typically too optimistic. Similarly, target prices of analysts had little correlation with, let’s call it “reality”. These biases, together with the bias of analysts to recommend glamour stocks and chase price momentum have been well-documented for years. Thanks to regulatory changes, the excessive optimism of analysts has declined after 2000 but their recommendations are still far from being unbiased.

Nevertheless, the research paper argued that while the buy and sell recommendations, as well as price targets, may be biased, the change in recommendations and the change in the gap between price target and share price contain valuable information. And that might indeed be true. After all, analysts are no dummies but experts in the sectors and companies they cover. Thus, they tend to be able to predict changes in the business outlook of an industry or an individual company and these changes in outlook are then reflected in changes in their recommendations and price targets. The research paper then went on to show that with the help of a little data mining to improve the results, a long-only strategy investing in the stocks with the most positive recommendation changes earned an active return of 3.65% per year from 1998 to 2019. Similarly, the paper found that a strategy based on changes in target price gaps delivered an active return of 4.77% per year on a long-only basis. Thankfully, the paper also showed that all of these results are statistically significant on a 1% level.

Which got me thinking…

If you get such high returns (with the additional benefit of low volatility), why have I never read about these strategies in an academic paper? And why have I never heard of a fund manager actively exploiting this strategy? In the case of this research report, it turns out that the answer to my question could be found in the appendix. Normally, these kinds of reports don’t provide crucial information like portfolio turnover or transaction cost assumptions at all, but in this case, the authors thankfully did so in the appendix, which allowed me to estimate the impact of things that academics euphemistically call “frictions”. 

First, the appendix of the paper showed that the return from the proposed active strategies largely disappeared if there was an implementation lag between calculating the signal and buying the corresponding portfolio of one or two months. This factor decay is so fast, that I suspect returns might already be significantly lower if one has an implementation lag of a week or so which would be normal in the case of a traditional mutual fund or a segregated managed account (though probably not in the case of a hedge fund).

Second, the study was done using the Russell 3000 index of stocks. This index includes not only large cap stocks but also small and midcaps and this means that transaction costs are likely substantial due to the reduced liquidity of these stocks. It also means that the portfolios they constructed contained a large number of stocks that needed to be traded each month to keep the strategy going. Indeed, the appendix of the report stated that the long-only portion of the portfolio had a monthly turnover of 28% for the strategy following changes in buy and sell recommendations and 48% for the strategy following changes in price target gap. The long-short version of the strategies had roughly twice the turnover per month. 

If I assume relatively typical transaction costs of 1% round trip for an institutional investor (see an analysis of the effective trading costs for institutional investors that includes direct and indirect costs here), then the portfolio turnover reduces the monthly return by 0.28% to 0.42% or from 3.36% to 5.04% per year. The entire reported active alpha is gobbled up by transaction costs! 

Unfortunately, too many investors are tempted by such short-term oriented strategies and their seemingly high returns. Yet, the virtuous investor takes all the costs into account to compare the present bitterness of the fight with the bitterness of the sin hereafter which followeth him that is overthrown to use Erasmus’ words.

Active strategies before and after transaction costs