Biotech FDA Approval Catalyst Strategy
Bio-Technology is a challenging sector for investors due to the binary nature of the product cycle. Indeed, many bio-technology firms’ futures rest upon the success of a single product. A critical stage in the product life-cycle is the FDA approval process. We examine a simple strategy that goes long a company’s stock upon filing of a 510k for a medical device approval. We then exit the position the day before the scheduled FDA decision date.
This is a pure momentum strategy that rides the enthusiasm surrounding a company’s FDA approval. The strategy calls for exiting the position just prior to the binary outcome of an FDA decision, alleviating the ‘event risk’ surrounding the FDA decision.
In this article we examine all 510k filings made with the FDA from 1999 to present. Our main conclusions are as follows:
· A basic strategy of entering positions on filing date and exiting the position just prior to decision date generates statistically and economically significant excess returns.
· The strategy is most effective for stocks of companies with poor prior performance as measured by 12-month price momentum.
· Conditioning the strategy on price momentum enhances the strategy’s strength for stocks with market cap in excess of $100mm.
· Finally, we examine the strategy conditioned on other factors such as short-term price reversal, cash-flow to price, and money flow with mixed results.
Biotech companies that are currently slated for FDA 501(k) decisions before 3/31/12 include: Echo Therapeutics (ECTE), Gen-Probe Incorporated (GPRO), and Insulet Corporation (PODD). The three companies are all small cap with market capitalizations of $76M, $2.8B, and $900M, respectively.
Companies that are currently slated for FDA 501(k) decisions before 6/30/12 include: Anika Therapeutics (ANIK), Given Imaging (GIVN), and Hansen Medical (HNSN). These three companies are all small cap as well with market capitalizations of $120M, $539M, and $135M, respectively.
As mentioned, the Biotech Catalyst Strategy is most effective for stocks with poor price momentum. Ranking the six candidate stocks by their S&P Relative Strength rankings, ECTE has the lowest ranking of (12), followed by HNSN (16), GPRO (59), GIVN (81), and ANIK (95).
Biotech Catalyst Strategy
Riding the enthusiasm surrounding biotechnology firms’ FDA filings is not new or radical, yet most research on biotech has centered evaluating areas of R&D and human capital (Hurd, 2001; Joos, 2007). This week we looked at the exploitability of a strategy centered solely these FDA filings.
We find if an investor were to pursue a naïve event approach to the strategy, i.e., purchasing all companies that file a 510k, we find that the investors would earn a statistically significant return of 5.25%, earning a positive excess return relative to all Health Care companies (GICS code 35). The strategy would be profitable nearly 55% of the time, as seen in Table 1.
Table 1: Buying All Biotechs on FDA Filing Data and Selling Decision Date -
Part of the problem in evaluating the success of our strategy is the high degree of variability in the length of the holding period, which is largely driven by the complexity of the filing and the FDA’s ability to reach a decision. As a result, we subdivided the events by holding period bins. As seen in Table 2, we find that longer holding periods are positively related to excess returns and hit rates. When the returns are annualized and an ANOVA is run, we do not find that the holding period returns are statistically different.
Table 2: Biotech Strategy by Holding Period
Generally, as an investment strategy becomes more popular, the excess returns can be crowded out through the presence of too many investors following the strategy. Given its positive performance over the entire period, potential investors must consider the efficacy of the strategy through time, paying particular attention to the most recent performance. Though 2009 was a difficult year for the approach, it does not appear as if the strategy’s popularity has diminished its strength. In Table 3, we observe strong hit rates and excess returns in both 2008 and 2010.
Table 3: Biotech Strategy Payoff by Year of Filing Date
FDA Filing as a Turn-Around Catalyst
Our results indicate a nearly monotonic inverse relationship between price momentum and the strategy’s payoff. The weaker the price momentum of a stock leading into its filing with the FDA, the greater the payoff to the strategy. Indeed, the worst quintile of price momentum provides an average holding period return of 11% as seen in Table 4. Furthermore, our t-test comparing the mean return for Quintile 5 to the mean return of Quintiles 1-4 establishes a statistical difference at the 97% confidence level. An ANOVA analysis provides the same conclusion.
A large number of names in our sample were small market capitalization stocks and bring with them higher costs of trading for institutional investors. In order to assess if the inverse relationship between our Biotech strategy and price momentum held for more liquid names, we re-ran our analysis limiting our sample to those companies whose market capitalization was in excess of $100 million on filing date. As seen in Table 5, we find that price momentum Quintiles 4 and 5 of this large-cap sample provide similar returns as the ‘all-cap’ screened universe.
Our results for companies with market capitalization in excess of $100mm again indicate a nearly monotonic relationship between price momentum and the strategy’s payoff. That is, the weaker the price momentum of a stock on filing date, the greater the payoff to the strategy. Indeed, the worst quintile of price momentum proves an average holding period return of just under 11%, see Table 5. When we examined the t-test comparing the mean return for Quintile 4 and 5 to the mean return of Quintiles 1-3, we found a statistical difference at the 99% confidence level.
Given the significant improvement from the strategy’s interaction with price momentum, we examined the strategy in conjunction with 3 other factors: 1-Month Money Flow, Cash Flow-to-price, and Short-term reversal. Though we were optimistic that each would provide additional insight, we were unable to improve upon the basic strategy in a convincing way.
Next we ran a portfolio simulation to gauge how an equal weighted portfolio would have fared using our strategy. The simulation was run using the following settings:
· Each position is equal weighted
· The position is held 30-days (Not until decision date)
· The portfolio is rebalanced daily. As new filings are made, they are added.
· All companies that made 501k filings and had pricing were purchased
· Domestic and International companies were used. Dollar based returns cited.
In our simulation, we treat each outstanding filing independently. If a company has multiple filings outstanding contemporaneously, we will overweight our position in that company relative to its number of filings. Additionally the number of instances in a given month is variable. Therefore, the size of the position for one outstanding filing is proportional to the number of filings in a given month (not a constant). In an average month, we typically observe 40-50 stocks/month in our general approach and only 10-15 stocks/month when conditioned and constrained based on price momentum quintiles.
The average monthly return to the basic Biotech strategy is 51bps while the price momentum screened approach yields 111 bps. The monthly hit rates are 59% and 52%, respectively. Chart 1 illustrates the cumulative excess returns to the basic strategy and price momentum screened strategy, 565% and 1,343%, respectively.
Chart 1: Cumulative Return to the Basic and Price Momentum Screened Approaches
Dec 99 to Oct 10; All Health Care stocks; Daily Rebalancing
We fortunately have significant resources accustomed to doing just this kind of identifier mapping. A data and analytics team in India matches up approximately 50% of the company names in the FDA data to trading IDs. While this percentage may seem low, one must remember that a large number of the names in the FDA data set were private and not publicly traded. Still others were subsidiaries, in which case we matched to the parent trading ID.
A further data challenge was obtaining good pricing. This challenge applies mainly to the very smallest capitalization companies. Large reverse stock splits resulted in aberrantly large prices. In order to ensure the requisite data quality, we cross-checked and validated these prices against our own stock split data and verified their accuracy.
A Comment on FDA Data
The primary data challenge became matching up the FDA data with our own. Specifically, the FDA cares little about stock identifiers. Matching company names in a text format to a company identifier proved a challenging project.
When a biotechnology firm files a 510k form with the FDA seeking approval for a medical device, we find that the filing event alone serves as a catalyst for the stock. Our research shows that holding these names during the period in which the FDA is considering approval results in significant economic profits. This biotech strategy is robust in that it appears to work through time and across market capitalization.
In looking to enhance our biotech strategy, we also found that conditioning on price momentum serves to almost double the return of the strategy. Specifically, by buying the most beaten down stocks of companies, i.e., those with the lowest price momentum leading into the FDA filing event, investors can significantly improve the strategy.
Exploration into other conditioning factors such as Cash-Flow-to-Price, Money Flow, and Short-term Reversal did not yield noteworthy results.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.