Bailout

When stocks become manipulable

By Felix Salmon

Have you ever wondered how big a stock needs to be before it can be efficiently arbitraged? The Fundometry blog has done an investigation along those lines and the answer would seem to be somewhere around $400 million in market cap.

Here’s the chart:

What you’re looking at here is four different ETFs, split into quintiles according to the size of their components. So the bar on the far left is the smallest 20% of stocks in the IWC microcap ETF, while the bar on the far right is the largest 20% of stocks in the S&P 500.

The y-axis, meanwhile, measures correlation: the degree to which each of the stocks in that quintile is correlated with the index as a whole.

Clearly correlation is very high these days across the market and in general stocks seem to have roughly a 70% correlation with their index. But the smallest stocks — the bottom quintile of the small caps and most of the microcaps — have lower correlation, probably because they’re so small and illiquid that it’s hard to arbitrage them against their respective index.

Microcap stocks, in this chart, are stocks with a capitalization between $50 million and $500 million, while small caps are capitalized between $300 million and $1 billion. So judging by the chart alone, I’m thinking that about $400 million is the point at which you can expect your stock to be arbitraged as a matter of course against its index.

This issue is related to Harold Bradley’s theories about manipulation of the relationship between microcap stocks and obscure ETFs: below about $400 million or so in market cap, it seems there’s a certain amount of inefficiency in the market and therefore room, in theory, for stocks to be manipulated. (Yes, there might be fundamental reasons why very small stocks have lower correlations than their larger brethren, but even so, the fact remains that these stocks are hard enough to trade that manipulation can be profitable.)

Are these findings, then, grist for Bradley’s mill? Do they demonstrate that ETFs shouldn’t include microcap stocks? No. A lot more work needs to be done on that front. But at least now we have an idea of where the manipulation is likely to be taking place, if it’s happening at all.

Full post as published by Macroprudential on October 21, 2010 (boomark / email).