In 1991, a Stanford professor buried a paper that Wall Street would spend the next three decades quietly pretending not to have read.
His name was Thomas Cover, and he wasn't a stock picker. He co-wrote the textbook on information theory that engineers still learn from today, the same field built by Claude Shannon. Cover's paper, "Universal Portfolios," pointed that machinery at something nobody expected: a way to profit from stocks that individually go nowhere, using nothing but the fact that they move at all.
The mechanism is called volatility pumping. Take two stocks that swing hard but out of sync with each other, meaning when one spikes the other tends to dip, and neither one is trending strongly in any direction. Rebalance between them constantly: trim whichever one just popped, add to whichever one just dropped. Do this long enough, and the portfolio can out-grow either stock alone, purely from harvesting the back-and-forth.
The reason sits in one idea most investors never learn: maximize the expected logarithm of your wealth, not the expected value of your return. Ordinary expected value tells you to bet everything on whichever asset looks best on average. The logarithm, which is what compounding actually runs on, tells you something different: spread your bets and rebalance, because volatility that would otherwise look like pure risk becomes raw material for growth once you're harvesting it on purpose.
Cover pushed the idea further, building a portfolio strategy that requires no forecasting and no model of the market at all, yet mathematically guarantees it tracks the performance of the best possible fixed rebalancing strategy, the one you could only have known in hindsight, arbitrarily closely as time goes on.
The paper is free, sitting on Stanford's own site since 1991. Realizing that a stock going nowhere over twenty years isn't necessarily a stock doing nothing, that's the part almost nobody sits with long enough to notice.
Source: OkMan
Read more and join the conversation on Facebook.
SIGNAL TITANS
Leave a Reply