Generating Alpha by Using Behavioral Finance Approach: Biotech and Oil

Behavioral Finance is one of the fastest growing, yet least understood of investment management fads. Unfortunately, when a new investment strategy or technique becomes fashionable it is often latched upon by practitioners looking to reword and reinterpret it in order to fit their pre-existing strategy. Especially when this pre-existing investment strategy involves earning money on commission! It’s funny how much discourse on utilizing behavioral finance techniques involves ending up in promulgating a long only investment ‘strategy’!

This article on behavioral finance has a different objective. It is intended to argue that alpha generation-in terms of isolating stock picking ability-could be produced in the biotech and oil exploration and production (E & P) sectors. Moreover, this alpha generation could be significantly enhanced by using behavioral finance techniques to produce a hedge trading strategy. ETFs would be an integral part of this strategy.

Do to this I will have to answer three questions

What is Behavioral Finance?

What Sectors or Stocks are Best Suited to Behavioral Finance?

How to Generate Alpha Using this approach?

What is Behavioral Finance?

Behavioral Finance is best encapsulated in the work of Kahneman and Tversky’s seminal volume ‘Judgment under Uncertainty’. In this work the authors develop a series of case studies which all emanate from a central starting premise or observation. This premised is the idea that when faced with decisions under uncertainty we use heuristics or ‘short cuts’ to reach a conclusion. Unfortunately, these heuristics are, quite often,  not the optimal solution to the problem.

Moreover, according to Nassim Taleb, medical research has discovered that we use the emotional parts of our brains to make decisions based on risk. Knowing this, it should come as no surprise that we often make in-optimal decisions based on greed and fear in the markets.

There is plenty of literature on the subject and I would refer to the work of Kahneman and Tversky in the first instance. However, I shall briefly mention one heuristic, namely, the availability heuristic. This relates to our tendency to over estimate the importance of near-term or ‘available’ information. For example, I believe that investors tend to irrationally discount the rest of a biotech companies pipeline, just because they have had a recent failure in the lab. I think a similar process applies to Oil E & P stocks.

Why Biotech and Oil Stocks are Suited to Behavioral Finance

The key is uncertainty. These are sectors that exhibit a high degree of individual stock volatility, but interestingly that does not necessarily mean a high degree of dispersion. This allows them to be subject to long/short strategies which I will outline in the last section. For now, I want to focus on the two sectors in more detail.

Biotech investing is fraught with uncertainty. Not only are there the usual concerns of clinical safety and efficacy, but also regulatory and pricing concerns are paramount too. There are also competitive concerns to any company pipeline. This comes not only from generics (after patents run out) but from competitors starting to trial ‘me-too’ drugs with similar modes of action to the one that is in more advanced trial stages. I will focus on the clinical and competitive issues here, because arguably the other issues will reflect across the industry as a whole. Recall that we are trying to generate pure alpha here, not call the sector higher or lower.

Probabilities of Success in FDA Clinical Trials?

Evaluating the pipeline of a biotech company inevitably involves making some assumptions over the ‘chance of success’ of a drug achieving endpoints in clinical trials and getting FDA approval. This involves uncertainty. Biotech investors can use this uncertainty to their benefit, if they understand not to overreact to a companies trial results. In addition, there are all sorts of general assumptions made over drugs probabilities of success in FDA clinical trials, which turn out to be misguided or plain wrong.

I will go into more detail on this in another article in future. For now, let me give one example. How many times have you seen an analyst research report ‘pencil in’ a 50% probability of success for a drug in Phase III? The reason they do this is because historical evidence suggests that, that is a fair estimate. However, this evidence rarely analysis the influence of the mode of action. Is the compound using a similar mode of action to a compound that has been FDA approved before? What about the trial sizes? Is this a novel class of drug? What about the time spent in clinical trials? How tricky is the target indication? There are myriad inputs and a one size ‘50%’ fits all solution will not do.

 I argue that investors can use this to their advantage. I also argue that investors tend to mentally adjust ‘probabilities of success’ based on whether the company has had success/failure recently in other trials.

Chance of Success for Oil Exploration?

Similarly, with Oil E & P, I would argue that their exploration campaigns are fraught with uncertainty. A company with recent success with the drill bit suddenly sees its existing drilling program being presented in the best possible light. However, one with a few failures gets sold off and then many investors are convince themselves that the management are idiots. Another example is how companies get bid up just because they have prospects in a region similar to that which has had success. It doesn’t seem to matter that the prospect could be in an unrelated field or be a different type of play entirely.

Generating Alpha Using Behavioral Finance

What makes Biotech and Oil E & P interesting is that the individual share price movements display a high degree of event volatility, yet they will be correlated. This means that a long/short investor can generate alpha by buying a portfolio of stocks in the sector and then shorting the relative ETF.

For example, accepting that Oil companies are priced based on a factor of their reserves, an investor could argue that the chance of success in his oil stock portfolio is 40% He and the market agree that the un-risked NPV is $1000. This gives him a risked NPV of 1000*.4=$400 In other words, he thinks his portfolio is worth $400.  However, the market is pricing his stocks at $300. The market thinks the oil reserves in his stocks are only worth $300.  There is a value discrepancy here.

 He buys the stocks at $300 then shorts $300 worth of oil. When the value discrepancy is ironed out- by his companies discovering oil- then he will make $100 because his stocks go to $400. Note that this $100 return is irrespective of whether the price of Oil has doubled or not. In other words, this is pure alpha generation using a hedged approach to risk.

It is not hard to see how this approach would work with biotech stocks too. However, biotech stocks tend to be less correlated with the economy than Oil stocks. Also, I would expect their results to be more dispersed, because their end product is not a commodity as would be the case for oil producers. Biotech is subject to more technological obsolescence by medical science developments. This makes it more feasible to be used as a non-correlated concentrated portfolio than Oil E & P.

Nonetheless, both are great sectors to generate alpha utilising a behavioral finance approach. As a proxy for shorting Oil or Biotech a relevant ETF should give good exposure.

Source:

Kahneman, Daniel and Tversky, Amos ‘Judgment under Uncertainty: Heuristics and Biases’ Cambridge University Press, 1982

Kahneman, Daniel and Tversky, Amos ‘Choices, Values and Frames’, Cambridge University Press, 2000

Taleb, Nassim Nicholas ‘Fooled by Randomness’ Random House, 2008

Generating Returns in an Increasingly Correlated Financial Market

The proliferation of growth of Exchange Traded Funds ETFs and the increasing economic inter linkage of the global economy are believed to be behind some fundamental changes in financial markets. In particular, correlations are seen as being on the uptrend. Stock correlations with their index and sector are seen as rising, as is the correlations between asset classes.

Historically, Individual stock correlations with the index usually rise when the market falls as investors sell out of stocks and sectors, however there may be more pervasive reasons why correlations could go up in future years.

ETFs and Correlation

ETF growth could be a cause, as ETFs are taking an increasing share of stock and bond market investment. ETFs tend to be passively managed. They buy and sell stocks and bonds, in order to mirror market cap weightings rather than their underlying fundamentals. Similarly, sector ETFs tend to follow the same approach. All of which, should create an environment whereby individual stocks and sectors tend to move in step with each other. Furthermore, it will make achieving stock diversification a lot harder.

The Effects of Globalisation on Markets

As the global economy and its markets become ever more integrated, the fundamentals that drive prices will become increasingly integrated. The business cycle will become more synchronised as will policy decisions by ruling parties. The resulting lack of diversification may cause superposition within the global economy and- given the implied directionality-may increase the risk of meltdown, when it all goes wrong. Indeed, Taleb (2007/2010) in "The Black Swan: The Impact of the Highly Improbable" develops a similar line of argument.

Indeed, each successive recession seems to have got deeper in magnitude and required lower and lower interest rates in order to stimulate the economy. Therefore, a potential increase in correlations could be seen as, merely reflecting the underlying reality of how the global economy is changing.

The Role of Leverage and how Correlations Work Historically

Alternatively, increasing correlation could be a merely an extenuation of previous market conditions. Historically, as markets fall, the only thing that has risen is correlations, liquidity problems and panic, cause investors to indiscriminately sell out of investments. As leverage becomes an increasing part of global trading, this attribute of market trading could be accentuated. In this scenario, Delta trading or stock market timing, becomes the only game in town, as correlations go up in the fall and stay high, even given a bounce.

If this argument is correct than low correlations can be seen to be contingent upon relatively stable market conditions. So, investors should integrate these changing market conditions into their strategies, provided correlation levels are a factor in their performance. Market conditions cannot be relied upon to stay stable.

Correlations Affect Investment Strategy

With increasing correlation, alpha generation becomes a pointless exercise as individual assets move together in line with an index. For equities, stock picking and fundamental research is a waste of time. Investment is reduced to delta trading, or in other words, being in and out of markets and assets classes at the right time.

Furthermore, certain hedged strategies are affected. Equity market neutral strategies are affected if they are making assumptions as to their underlying beta exposure. Stock beta will move around unusually. For example, in a falling market long stock/short index strategies may find them under exposed on the short side as correlation suddenly jumps. Similarly, long stock/short sector will be exposed to the same sort of risks.

In this scenario, the correct thing to do would be to run an equal weighted correlated stock portfolio alongside a short position in the index. The trouble is that doing this will tend to reduce the opportunities for significant alpha generation as any semblance of stock/sector preference is ironed out by market weighting.

Tactical Asset Allocation Strategies

Tactical asset allocation strategies seek to eliminate the attempt to generate alpha in favour of trading asset classes against each other. However, even these strategies won’t work in this scenario because increasing correlation within asset classes will create difficulties. So what to do if increasing correlation is a persistent part of the investment world?

Statistical Arbitrage Strategies

Statistical arbitrage or stat arb, strategies tend to use mean reversion techniques in order to isolate unusual trading positions. They rely on statistical signals that provide entry and exit levels into trades. If correlation is on the increase, then the signals given for, say two dislocated stock market sectors will be more relevant, because the signal is seen as reflecting greater statistical outliers. Therefore, these outliers or price dislocations could be closed easily as markets trend towards correlation.

Stock Market Investment Strategies

Another strategy that could work is to find sectors that are characterised by high degrees of uncertainty and the opportunity for significant volatility on their earnings prospects. For example, as the whole sector gets sold off indiscriminately, the stock price of individual companies-who might have with significant upside to their earnings - will be sold off. If the significant upside event occurs, then these stocks will outperform.

Two sectors that exhibit these characteristics are biotech (where clinical trial results will guide prospects) and Oil exploration (well results) and the prospect for alpha generation with a long stock/short sector approach may, in fact, be raised by increasing correlation.

Source:

Taleb, Nassim Nicholas (2007/2010). The Black Swan: The Impact of the Highly Improbable. New York: Random House.