Investors have forgotten the scientific method
Applying it is important because market wisdom can be more fallacious than wise
[NEW YORK] The scientific method has been the gold standard in analysis for at least 400 years, but you wouldn’t know that from reading much financial research. Untested market truisms, spurious correlations and cherry-picking abound, which can confuse investors and reduce market efficiency.
Oxford Languages defines the scientific method as “observation, measurement and experiment, and the formulation, testing and modification of hypotheses”.
Most of us learned about this systematic process in middle school, but, sadly, one rarely sees it employed in financial research created by holders of advanced degrees.
Let me state up front that in critiquing “financial research”, I am not talking about an equity analyst recommending a stock based partly on subjective assumptions about a company’s future earnings. That’s reasonable.
Rather, I am discussing statements about how securities markets work. For example, claims that certain geopolitical events, such as an election or war, will have a specific impact on future stock returns, credit spreads or commodity prices.
We’ve long heard financial analysts claim that buying bonds is a good hedge against falling equity prices or that value stocks will outperform growth after long periods of lagging. And now we frequently hear the claim that the price of Bitcoin will keep rising as countries debase their currencies.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
The sad truth is that such pronouncements are often backed by nothing as concrete as observation, measurement and experimentation.
Instead, financial “experts” sometimes selectively seek facts that support a predetermined conclusion that simply reflects the financial interests of their employer.
I recall a Wall Street salesman once characterising his firm’s chief economist’s perennial bullishness as a virtue. It may have been, but not for their clients.
Buy signal?
In my early days as a fixed income strategist, I wondered whether certain bits of received wisdom that I had encountered earlier as a corporate bond trader truly described how the market worked.
I had been told, for example, that elimination of a common stock dividend constituted a buy signal for the company’s bonds. The plausible explanation was that halting distributions to shareholders conserved cash, putting the company on a sounder financial footing to the benefit of debtholders.
Unwilling to accept that description of market dynamics on faith, I studied 31 dividend eliminations from the preceding decade. I measured the subsequent changes in the affected bonds’ risk premiums, defined in that study as the spread, or the difference between their yields and the Moody’s Aaa corporate yield.
If the risk premium decreased following the dividend elimination, this would suggest that such an event did signal that the bond was about to experience a favourable relative price move.
That was my hypothesis, but my test of the data found the exact opposite. In 20 out of 31 cases, the bond’s risk premium increased between one month prior to the dividend cut and one month after the event.
In technical terminology, my analysis showed, with 90 per cent confidence, that investors were better off acting contrary to the market “rule of thumb”. That is a bit shy of the 95 per cent level statisticians regard as justifying strong confidence, but the results certainly offer no justification for the long-touted strategy of buying a bond after a dividend elimination.
This is only one example, but the key point in revisiting this application of the scientific method four decades ago is to show how market wisdom can be more fallacious than wise.
Applying the method
There is much more to applying the scientific method in financial research than what I’ve noted above, of course.
One important procedure is adjusting for “confounding variables” that could make a false relationship look valid. Another tactic is conducting an “out-of-sample test”, which guards against accepting a misleading result that can arise from special conditions that existed only during the specific observation period.
Since the 1980s, technological advances have made data more widely available, facilitating abundant, sound research that has expanded our knowledge of how the market functions. And some of this research has involved rigorous testing and advanced statistical analysis.
Investors must nevertheless continue to be discriminating with respect to allegedly expert advice. Protecting their financial well-being depends on conscientiously differentiating between mere hypotheses – flat-out assertions based solely on opinion – and conclusions grounded in rigorous analysis that follows the scientific method.
Unfortunately, there still appears to be a lot less of the latter. REUTERS
Share with us your feedback on BT's products and services
