A recent Motley Fool article discusses reverse stock splits and their implications for future performance: “Investors have to wonder: Will reverse splits do any good, or are they basically the kiss of death for a company?” With Palantir Finance, we can easily examine the short-term performance of companies that undergo reverse stock splits. This question lends itself particularly well to a common process in Palantir Finance where we create a group of instruments based on a selection of filters, generate an index based on that group and compare the performance of that index to a benchmark.
The Explorer tool allows us to add a set of filters to isolate only the instruments we are interested in. The following set of filters returns stocks in the NYSE, Nasdaq and AMEX with a market cap larger than $10 million that have undergone a reverse stock split.
We can also set a look-back period on each of the metrics. By specifying the scope of the splits metric to include data from only the past two weeks, we get all companies that satisfy the above criteria, and have very recently undergone a reverse stock split. We can export this Explorer Group into the index tool to create a dynamic index of all these stocks. The Index is rebalanced weekly, so every week, it recalculates its membership based upon the rules laid out in the Explorer tool. Each week, a stock is added to the Index if it underwent a reverse split. It is kept in the index for two weeks, then removed. The Individual Instrument Constributions segment on the bottom half of the screenshot graphically depicts the Index’s composition. Each color represents a single instrument. The mosaic pattern indicates that the composition changes quickly and often, and usually contains anywhere from one to three instruments at a time.
The performance of this Index is very heavily influenced by the early results. Since the Index performed so poorly in the latter half of 2003, the value of the Index dropped significantly, making any movements in the Index after 2003 look muted. We don’t want our results to be so heavily weighted by this period, so we instead look at the performance of these instruments in the Strategy tool. The following strategy uses the Index we created above. With a starting Net Asset Value of $10,000, the strategy checks the Index membership each day and buys $1,000 worth of any new stocks that are included. These stocks are then held for 10 business days. Effectively, we are testing the short term performance of stocks immediately following a reverse stock split. The results panel shows us the performance of our strategy. Following this strategy would have lost over 60% of our starting Net Asset Value.
We can compare the performance of the strategy to the performance of the S&P 500 over time using the Chart tool. Normalizing the y-axis to the start date of the strategy, January 1, 2003, shows that our strategy performed significantly worse than the S&P 500 over almost the entire time horizon.
It may be unfair to compare the performance of these stocks to the overall market, since stocks that undergo a reverse stock split are generally already in financial trouble. Theoretically, a reverse stock split doesn’t alter anything about a stock other than price and number of shares outstanding — the overall value of the company remains the same. Still, investors may perceive a reverse stock split as a signal of desperation from a company’s management, which could further damage a stock’s performance. On the flip side, an extremely low stock price could garner the label of “penny stock” and scare away potential investors. Further analysis can help us isolate the effect of just the reverse split and determine in general which of these factors prevails.
To begin building a control group, we need to identify some common properties shared by all stocks. We can build off the previous claim that these stocks are all in some sort of financial trouble and their recent performance has followed a downward trend. The following histogram plots the 6-month return of all stocks that underwent a reverse stock split on the day that the split occurred.
The average 6-month return of this group is 13.09%, as labeled by the green dotted line on the histogram, but the distribution of returns is heavily right tailed with four outlying observations pulling the average 6-month return of this group into the positive. Since the mean is so heavily influenced by outliers, it might be better to look at a more robust measure of centrality such as the median or the trimmed mean. The yellow dotted line shows us that the median 6-month return of a stock on the day that it’s reverse stock split took effect was -11.29%. We can now build a control group around this fact. Using methodology similar to the above, we can create a control index of stocks with a 6-month return of between -6% and -16% on each day (excluding the reverse split stocks), and compare a strategy that holds these control stocks to the strategy that holds our reverse split group.
The orange line indicates the performance of the control group we created. Our reverse split index underperformed compared to the control index. This evidence suggests that a reverse split may have an adverse effect on the performance of a stock, if we assume stocks that performed poorly in the recent past are likely to perform poorly in the near future. Notice that for a long period of time, the control index did quite a bit better than the S&P 500, so this assumption may not hold up in practice. There are many other characteristics one could use to create a control index to compare against. This type of quick and dirty analysis is a nearly trivial task within Palantir Finance. A more concrete, quantitative analysis can be done through the Regression tool, but that’s a topic we’ll visit in another post.