Journalists and asset managers: similar objectives, similar methods.
Luca Sofri, publisher of ilpost.it, has written a wonderful book on sloppy and otherwise compromised journalistic practices in Italy and elsewhere (Notizie che non lo erano). In reading it I could not help to find similarities between journalism and investing.
One of the main points in Luca’s book is how with time journalists have become lazier than ever in scrubbing the sources and making sure the facts are right. The real target today is not to inform people but to get more “hits”, “likes”, “views” on the web or whatever. These are not necessarily undesirable things unless you achieve them by compromising your standards.
In a similar vein, as I have said before, the main objective of the investment management industry today is to accumulate assets regardless of the benefits or improvements delivered to people’s lives. An element of this strategy is the constant dissemination of relatively useless (if not simply inaccurate) information. Three brief examples illustrate the problem.
Economic growth: in the long run (years, decades) it is terribly important to investors but in the short run (quarters, years) it’s mostly pure noise. Yet there isn’t a reputable investment house that does not forecast growth in the next months and quarters. What is the utility of this information from the markets’ point of view? The historical (over the last 100+ years) correlation between growth and equity returns is slightly negative. Some say this is because expectations tend to be built in market prices; but that is indeed the point: if markets move before growth figures are released, then why bother?
Or take the use of price/earnings (p/e) ratios to judge whether markets are expensive or otherwise. Here the problem is not the concept but the how. Most research houses use a p/e where the “e” is either last year’s actual or next year’s forecasted earnings. Aside from the logical question of who in his right mind would value a business on the basis of a single year’s worth of earnings (actual or forecasted), it turns out the statistic has very little use for knowing what equities will do over the next 5, 10 years or any reasonable equity investment horizon. But if you use a form of cyclically-adjusted earnings for your p/e calculation (such as trend earnings or a moving average of the last 10 years’ earnings), the result will startle you: some versions will yield an R2 with future returns of 0.50-0.70.
And finally, technical analysis: it’s true that a few years ago MIT professor Andrew Lo found that the conditional behavior of prices (conditional on certain common technical analysis patterns) is not random, but even he doubted you could make any money out of this observation when you accounted for transaction costs. Yet you would be pressed to find a major investment house whose research does not carry a technician on the payroll.
Inappropriate use of information by journalists and investment managers: who would have guessed these professions had so much in common?