Arbitrage Pricing Theory (APT)

In 1976 Stephen Ross published an extremely influential article, titled “The Arbitrage Theory of Capital Asset Pricing,” in the Journal of Economic Theory. His “theory,” which is really more of an hypothesis or suggestion, is offered as an alternative to an earlier academic theory, the Capital Asset Pricing Model.

Investors in Ross’s world are the usual academic stick figures, who all have the same information and expectations.  And he appears as eager as his predecessors to use the tools of statistical regression analysis and computer power to make very precise and complex calculations.

Ross’s innovation

What’s new in Ross’s approach is his starting point, which is investor expectations, rather than price outcomes in the movement of actual securities.  

In his view, investors–or at least those who have the power to influence prices– calculate fair value for all assets, explicitly or implicitly, using multiple regression analysis on variables that all investors are in agreement about.  In other words, they look at financial assets as a straight-line function of some economic variables, like the price of gold, the inflation rate, the growth of GDP…  We don’t know in advance what these variables may be at any given time, but can presumably deduce them from the behavior of asset prices.

Investors can have different opinions about asset prices, but these will ultimately be based on different expectations for future values of one or more of the fundamental agreed-upon variables.   But investors will always engage in arbitrage when individual views diverge sufficiently.  That is, when they see an asset they regard as overpriced, they will sell it short and invest the proceeds in fairly-priced assets; when they see an underpriced asset, they will buy it, getting the money from sales of fairly-priced assets.  Although Ross waffles a bit on this point, arbitrage always reduces risk.  It always drives the overall asset market toward a relatively stable equilibrium.  The Ross model is a static one.  There is no temporal dimension.  therefore, there is no discussion of how long the arbitrage process may take, just, I think, the assumption that it is not problematic and will occur relatively swiftly.

The APT’s influence

APT has had two long-lasting consequences:

1.  It has sparked a gold rush of researchers trying to determine the key price-determining factors and to set up consulting practices to sell their results and their advisory services to institutional investors.  As yet, however, there is no general agreement about what the factors are.  This is understandable.  On the one hand, there is no way to be sure that any factors are the “right” ones.  Nor is there any good reason to believe that there is, even for a short period of time, a single “true” set of factors that all investors use.  On the other hand, if researchers all came up with the same factors, they would be selling an undifferentiated, commodity-like product that would not command a premium price.

2.  The arbitrage idea has been used to buttress the efficient market hypothesis.  The argument, still taught in business schools, is that while there may be asset mispricings in the markets from time to time these are swiftly eliminated by arbitrage.

My take

I think the examination of investor expectations is a real advance. But in rereading the article recently, what strikes me the most is the time, 1973-1976, during which this article was written and revised.

This was a period of extraordinary turmoil in financial markets.  The government of the UK effectively went bankrupt.  Stock prices had an extraordinary collapse, rivaling our experience in 2007-2009.  The top 10% of the S&P 500, the so-called “Nifty Fifty,” which had traded for a time at 5x-10x the price earnings multiple of the typical stock, were in the fourth year of what would be a decade-long swoon as their multiples reverted toward the median.

None of this fits neatly into the Ross hypothesis but no mention of these events appears in the Ross article.  It’s kind of like looking at an analysis of politics in the US during the Sixties that doesn’t mention civil rights or Vietnam or Woodstock or the War on Poverty.

Leave a Reply

%d bloggers like this: