Value Line: why mechanical systems stop working

Sam Eisenstadt of Value Line created a famous computer-driven stock ranking system that worked fabulously for about a quarter century.  Then it stopped working.  Why?

1.  Any economic system is dynamic, not static.  When an innovation happens, it spurs changes in all sorts of other systematic variables.  When a competitor introduces a new product into a market, rivals don’t simply watch their market share erode.  They launch new products themselves, ranging from simple knockoffs of the original innovation to genuinely new, but different, products of their own.

To the extent that “me too” products proliferate, the value/power of the initial innovation is eroded.  In the case of Value Line, post-ERISA, rival money managers poached Value Line IT people to create duplicate systems for themselves.  In fact, a number of very successful value-oriented money managers in the 1970s-1980s were driven, so far as I can see, almost exclusively by their VL ranking system knockoffs.

The fact that many professionals began to act on the system’s predictions–in large size and as soon as the predictions were made–began to blunt the effect of the system.  It tended to make subsequent outperformance smaller in degree and duration.

In short, the Value Line system changed the world  …and then the world began to catch up.

2.  The Value Line system is based on an extensive analysis of historical data.  That was okay when computing power was expensive and when (future-oriented) stock market derivatives were few and far between.   This was also before ERISA turned money management from a backwater into a gigantic business, that is, before brokerage houses and money managers built large staffs of securities analysts churning out predictions of future earnings.

The result of these changes was to reorient Wall Street away from historical earnings to studying–and buying and selling based on–future earnings estimates.  When the actual numbers came out, the market had already reacted.  Not always, but most of the time.

3.  The system has, in my view, a number of quirks, which I’ll just state without elaboration.

–It’s biased toward smaller stocks, which is one reason the VL system did so well in the 1970s.

–I think there’s a semi-permanent underclass of 5-ranked stocks, which makes the 1 vs. 5 comparison less relevant than, say, 1s vs. the S&P 500.

–The system is bad at turning points in the economy, activity either decelerates or accelerates sharply.  In today’s world, investors react to macroeconomic news far in advance of when corporate earnings reflect such changes.

–It works better in down markets, where investors cling closer to reported earnings, than in up.

could the VL system start to work again?

Maybe.  Sam Eisenstadt is a shrewd guy, after all.  Much of the Wall Street information gathering apparatus has been dismantled during Great Recession-induced cost cutting.  We’re unlikely, I think, to experience another decade of macroeconomic and stock market disruption on the scale of the past ten years (at least, I hope we won’t).  So conditions for a system like VL’s to work look to be better than they’ve been in a long time.

The biggest issue I can see is that computing power is no longer expensive.  Most of us could do something like what VL does on our home computers.  I doubt many of us are going to take the trouble, though.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: