Product and Strategy Notes: H1N1 Update, November 2009; Corporate Management Success: Luck Versus Skill; New Products; and Who Knew?

H1N1 Update, November 2009

This is another update to our original assessment of the potential economic and asset allocation implications of H1N1 Swine Flu that was published in our May 2009 issue. At that time, we noted a number of warning indicators we would be monitoring. In some cases, important changes have or may be occurring in these areas.

The first observation is that worldwide H1N1 attack rates (i.e., the percentage of the population that becomes infected) may reach 50%, compared to 10% to 20% attack rate for normal seasonal influenza (reference: U.S. Defense Intelligence Assessment DI-1812-1555-09 dated 10Jun09, "Worldwide 2009-H1N1 Virus Might Have Substantially Higher Health Impact Than Typical Seasonal Influenza"). The World Health Organization has produced a lower H1N1 attack rate estimate, at 22% to 33%, and the UK's Planning Assumptions issued on 3 Sep 09 used 30%. In part, higher attack rates for H1N1 are due to the late start of vaccination programs in most countries, coupled with shortages of H1N1 vaccine. All else being equal, higher attack rates will lead to higher hospitalization rates (H1N1 related hospitalizations per 100,000 population) and deaths.

The second observation is related to H1N1's virulence, or its ability to cause severe illness and death. In May, we noted that we would be looking for "reports that it is associated with viral pneumonia, and cases of severe inflammation (which produce so-called 'cytokine storms', in which inflammation sets off a positive feedback loop, sending the body' immune system into overdrive, and filling the lungs with white blood cells and other fluids). This may be associated with an unusually high death rate for 19 -- 64 year olds, relative to the death rates for younger and older infected patients" and "reports that the virus is characterized by unusually high replication rates in a host." Unfortunately, those reports are now starting to come in. The virus continues to disproportionately affect young people, with those 24 and younger statistically most at risk. There also appears to have been an important change in H1N1's ability to bind to tissue deep in the lungs and cause severe cases of viral pneumonia (technically you may see this change referred to as a new development in the H1N1 gene at position D225G). This is similar to a change which was thought to have occurred in the 1918 influenza virus that preceded its most deadly wave (see "Quantitative Biochemical Rationale for Differences in Transmissibility of 1918 Pandemic Influenza A Virus" by Srinivasan et al). There are also reports that the change at position D225G reduces the effectiveness of the H1N1 vaccine, reduces detection rates (because the viral load is higher in the lungs, but lower in the upper respiratory tract), increases viral loads, and increases reinfection rates. These developments can help to explain why the number of reported H1N1 infections has declined more slowly than epidemiology models first predicted.

We also note a rising number of reports of Tamiflu-resistant variants of H1N1. This is consistent with the spread of Tamiflu-resistance across seasonal influenza viruses, leaving Relenza as the most potent currently approved antiviral for use in fighting influenza.

In sum, the probability that we will experience another wave of H1N1 influenza infections that is more severe than what has been seen up to now appears to have increased. From an asset allocation perspective, this further raises the probability that financial markets will experience a longer-period of high uncertainty than many investors may currently expect. As a result, asset classes that deliver high relative returns when uncertainty is high (short term government bonds, volatility and gold) may be undervalued today. On the other hand, recent H1N1 developments have negative implications for those asset classes which perform best in the Normal Regime (such as equities and high yield bonds) that have staged a rather remarkable recovery in 2009 from their previous post-crash lows.

We will continue to closely monitor H1N1 developments.

Corporate Management Success: Luck Versus Skill

About five years ago, a researcher from Stanford (now at Said Business School at Oxford) named Jerker Denrell did some very creative work on the relative impact of luck versus skill in corporate management success. The results were undoubtedly disquieting to many, and Dr. Denrell probably paid a career price for his efforts. In "Random Walks and Sustained Competitive Advantage", he concluded that "sustained interfirm profitability differences may be very likely even if there are no a priori differences between firms...[Specifically], a random resource accumulation process is likely to produce persistent resource heterogeneity and sustained interfirm profitability differences." In a subsequent paper ("Should We Be Impressed with High Performance?"), Denrell noted that to the extent skill was involved in high performance, the tendency of management researchers to overemphasize the experiences of successful firms relative to the much larger number of failed firms could also be misleading. He noted that, "although it is reasonable to believe that more capable firms will achieve higher performance, several other factors influence firm performance, including luck. As a result, high performance is, at best, a very noisy signal of superior capabilities. Moreover, because it is a rare event, high performance is more likely for firms that engage in practices that produce high variability in outcomes. If such practices lead to lower average performance, exceptionally high performance will in fact be a signal of incompetence rather than competence." While we found these papers fascinating (and, we admit, intuitively in line with years of experience as line managers and consultants), they didn't get much traction in the larger management studies community. This was a great shame, because the analogies to and connections with active investment management are clear.

We were therefore very excited recently to see that the role of luck in corporate management success has now been taken up by another set of researchers from Deloitte Consulting. Raynor, Ahmed and Henderson (the latter is at the University of Texas) sum up their analysis in "A Random Search for Excellence: Why 'Great Company' Research Delivers Fables and Not Facts." We were already familiar with Andrew Henderson's writing, as he had previously co-authored "How Quickly Do CEOs Become Obsolete?", which found that CEOs ability to add value declined faster the more dynamic the industry. So we had high hopes for the work Henderson had done with his co-authors from Deloitte. We were not disappointed. In the present paper, the authors note, "researchers who think they are studying successful companies are usually studying the winners of a random walk." They then ask a pointed question, and give an equally blunt answer: "What does this mean for the soundness of some of the most popular and influential management research? The bottom line: you can't trust it." They elaborate, "since there are many more lucky companies than good ones, the inputs to every success study we can lay our hands on are very likely the wrong inputs. This has material consequences for the confidence we can have in the advice offered, for no matter how rigorous the data collection, no matter how Artistotelian the logic, to deviate a bit from the old aphorism, 'randomness in, randomness out.' Because these studies fail as science, managers cannot hope to reliably achieve the results they are told to expect."

Using Return on Assets as their performance measure, and a long set of corporate performance data, the authors find that under one percent of firms deliver superior performance that is statistically different from what luck alone could produce. However, they also take pains to note that management quality is still important at companies that are not in this elite group: "None of this should be taken to suggest that management doesn’t matter in firms with statistically unremarkable profiles. Rather, we're arguing that there is nothing demonstrably different, based purely on an examination of performance, about what management achieved in those firms. Remember, performance that is defensibly attributable to nothing other than common causes is still caused. But those causes are available, in a real sense, to all comers. The players in the drama of competition will certainly feel that they are working hard...because they are. But they are working no harder, and, more to the point, no more effectively, than the norm." The authors also note, "we do not take the paucity of firms with clear changes in performance as evidence that very few firms have ever changed their performance. We take it as evidence that very few firms have ever changed their performance enough to be distinguishable from the roar of white noise arising from the volatility endemic in a dynamic and unpredictable marketplace."

Last but not least, it is critical that investors recognize what happened when the authors replaced Return on Assets with Total Shareholder Returns as their key performance measure. They begin by noting that "shareholder returns are a function of the capital market's estimate of future performance. A good fraction of TSR tells the story of changing hopes for the future rather than delivering on past promises. Consequently, strong returns over time are often largely the result of consistent upside surprises that serve to ratchet up expectations, which is then made manifest in a rising stock price." Using TSR as a performance measure, the authors find no evidence of consistent superior performance beyond what would be expected due to luck alone. They conclude "markets rapidly bid up [the price] of any firm that is delivering exceptional returns so that it very quickly is no longer delivering exceptional returns."

The implications of these findings for active investment management are clear and harsh. If most cases of superior corporate performance are due to luck rather than skill, and if the stock prices of superior performers are quickly bid up to the point that additional superior returns become nearly impossible to achieve through skill (but not through an extended period of good luck), how many skilled (which is not synonymous with successful) active investment managers probably exist? This point is reinforced by another new paper, "The Alpha Uncertainty Principle" by Sassan Zaker from Banque Julius Baer. Zaker begins with an excellent review of the relationship between an active manager's Information Ratio and the years of data needed to distinguish between luck and skill, which is vastly longer than most managers' track records. He then offers the important and original insight that there is a further tradeoff between the breadth of strategies and techniques being employed to generate the alleged alpha and the amount of uncertainty regarding its authenticity (i.e., the statistical likelihood that it is based on skill rather than luck). In a version of Occam's razor, Zeker shows how, for a group of active managers with a given level of alleged alpha, we should prefer the manager who achieves it with the fewest degrees of freedom (i.e., different strategy elements, such as leverage, illiquidity, and/or dealing in multiple markets).

With respect to portfolio construction, we have three key takeaways from these papers. First, we have even more evidence that consistently successful active investment management (that creates skill-based value for investors after costs and taxes), must be very, very rare. Second, allocations to expensive active management strategies should therefore be used sparingly. Third, successful active strategies are more likely to be based on some form of arbitrage that exploits predictable behavioral tendencies rather than simple long-only security selection.

New Products

UBS has launched an exchange traded note (ETN) that tracks the Dow Jones UBS (formerly AIG) Commodities Index. Annual expenses are only 50 basis points, compared to 75 basis points on the competing iShares product (DJP). However, both of these products also require an investor to accept the credit risk of the note issuer. As we have noted in the past, in 2009 we switched from a long-only allocation to commodities futures to a new index product (LSC) that tracks the S&P Commodities Trend Indicator, which takes both long and short positions in different commodities futures. Our logic was that the flood of investor money into long-only commodities products had changed the structure of the market, raising futures prices relative to spot prices, and thereby reversing the positive roll returns that have been a key contributor (along with price surprises) to historical returns on commodities futures based index funds. In October, IndexIQ launched a similar equities based product (ticker GRES). It will track an equally weighted custom index of companies with operations in eight different commodities sectors, take long and short positions in them and offset its exposure to the overall equity market (presumably leaving a pure commodities exposure). Annual expenses will be 75 basis points.  It is an interesting approach, and we look forward to seeing how it performs. Most importantly, it is another example of the move towards relatively low cost, uncorrelated alpha products that deliberately offset their beta exposure.

Elsewhere on the new product front, in the U.S. Schwab has launched a range of very low cost equity ETFs, while Vanguard recently launched a range of UCITS products (covering both equity and fixed income) that have been gaining traction in Europe. We were also very interested to see the U.S. launch of iShares Diversified Alternatives Trust ETF (ticker ALT). With annual expenses of slightly more than 1.00%, this new product aims to provide retail investors with access to a wide range of alternative strategies. Coming from the iShares shop, we know that it is well designed, and will be well executed. However the critical and as yet unanswered question is the degree to which ALT's returns will be correlated with the returns on broadly defined asset classes -- in other words, is this really an uncorrelated alpha product? Time will tell. But we'll be watching with interest.

Finally, we call your attention to a very interesting new whitepaper from S&P ("A Beta for Sentiment?") that discusses the construction of a new investable index to track market sentiment. It makes for very interesting reading, and could lead to an equally interesting product at some point in the future. Ideally, we would like to see an inverse design that would cause returns to rise when sentiment declined, as it would offer investors another hedging instrument to potentially use in a portfolio.

Who Knew?

Over time, we read a lot of research. Some of these papers we set aside in the hope that we will one day be able to work them into an article. Eventually, this pile grows large enough that it can serve as the basis for an article on its own, usually very interesting, merits. Well, the pile has once again reached that height. So herewith is a short summary of some fascinating recent research.

Using an extremely extensive data set from Finland, Grinblatt, Keloharju, and Linnainmaa ask an aged old question: "Do Smart Investors Outperform Dumb Investors?" Well, the definitive answer is now in. They do. And now for the question on every parent's mind: How much of this is due to genetics? In "Genetic Variation in Financial Decision Making", Ceasrini, Johannesson, et al analyze Swedish data and conclude that "approximately 25% of individual variation in portfolio risk is due to genetic variation." Coincidentally, another recent paper studied the same issue. In "Nature of Nurture: What Determines Investor Behavior?" Barnea, Cronqvist, and Siegel use data on 40,000 Swedish twins and "find that up to 45% of the variation in stock market participation, asset allocation and portfolio risk choices is explained by a genetic component." Interestingly, twins that are raised apart were found to have similar portfolios. Of course, nurture is also important. In "Growing Up in a Recession: Beliefs and the Macroeconomy" Giuliano and Spilimbergao find that "individuals growing up during recessions tend to believe that success in life depends more on luck than on effort, support more government redistribution, but are less confident in public institutions." Sure looks like interesting times ahead, eh?

Moving on to gender differences, Sapienza, Zingales, and Maestripieri find that "higher levels of testosterone were associated with lower risk aversion among women, but not men" and that the effect of testosterone levels on both genders was non-linear. Moreover, both men and women "with higher levels of testosterone and lower levels of risk aversion were more likely to choose risky careers in finance." ("Gender Differences in Financial Risk Aversion and Career Choice Are Affected by Testosterone"). In another paper, ("Menstrual Cycle and Competitive Bidding"), Pearson and Schipper find that "women bid significantly higher than men in an auction in their menstrual and premenstrual phase, but do not bid significantly different in other phases of the menstrual cycle." They hypothesize that evolution has "genetically predisposed women to behave more riskily during the fertile phase of their cycle in order to increase the probability of conception, quality of offspring, and genetic variety." On the other hand, in "Two Heads are Less Bubbly Than One", Cheung and Palan find that in an experimental stock market, having investment decisions made by a team reduces the change of bubbles forming, particularly when these teams are made up of either two women or a man and a woman. In contrast, all male teams are characterized by "more extreme, though not consistently more profitable behavior." Who knew?

Finally, we call your attention to "The Way You Make Me Feel: Evidence for Individual Differences in Affective Presence" by Eisenkraft and Elfenbein. The authors ask, how much do individuals influence the way that other people feel? Specifically, they try to disentangle the impact of our own disposition from the impact of another person's presence on the feelings we experience. After controlling for a range of other influences, the authors find that our positive feelings are 31% explained by our own individual traits, while our traits explain 19% of the negative feelings we experience. In contrast, the presence of another person explains 10% of our positive feelings, but 23% of our negative feelings. Think about that for a second. Much of our happiness seems to come from within, while other people can really tick us off. Once again, we can only marvel -- "Who knew?"

| Table: Market Implied Regime Expectations and Three Year Return Forecast | Uncorrelated Alpha Strategies Detail | Global Asset Class Returns | Table: One Year Asset Class Valuation Conclusions and Recent Momentum | Product and Strategy Notes: H1N1 Update, November 2009; Corporate Management Success: Luck Versus Skill; New Products; and Who Knew? | November 2009 Economic Update | Market Phase Change Risk Analysis | November 2009 Issue: Key Points | Feature Article: Predicting Changes in Investor Behavior | This Month's Letters to the Editor: What is Index's Perspective on Risk-based Factor Modeling? and Index Definition of Alpha and Beta | Global Asset Class Valuation Updates Detail |