Two Mediocre Stocks, Daily Rebalancing, $7 Million

A Stanford professor took two ordinary stocks. Not good ones. Not ones he had inside knowledge about. Mediocre ones. He rebalanced them every day — mechanically, no predictions, no intuition — and turned $100,000 into $7 million.

That number sounds wrong. So let's be precise about what's actually happening here.

The strategy doesn't beat the market by being smarter about which stock goes up. It beats the market by exploiting volatility itself. The stocks bounce around. You rebalance. The bouncing is the fuel. This is called volatility harvesting, and it runs completely against the way most people think about investing.

Most people think volatility is the enemy. You endure it. You hedge against it. You wait for it to settle down.

This strategy says the opposite. Volatility is the point. The more the stocks swing relative to each other, the more you extract.

The man behind the math is Thomas Cover — information theorist, Stanford professor, one of the people who built the mathematical foundation for how we think about communication and uncertainty. His 1991 paper introduced what he called the Universal Portfolio. It's not famous outside academic circles. It should be.

Cover published the original proof in 1991. You can read the full paper — it's dense, it's mathematical, and it's worth at least skimming to see where this comes from.

The core claim is this: a portfolio strategy that constantly rebalances to fixed proportions — say, 50/50 between two assets — will, over time, outperform buy-and-hold on either asset alone, as long as the assets have any volatility and aren't perfectly correlated.

No forecasting. No signal. No edge in the traditional sense.

The math works because rebalancing forces you to sell what's gone up and buy what's gone down. Mechanically. Every time. You're always selling relative strength and buying relative weakness. Over enough cycles of volatility, that compounds in your favor.

The portfolio selection framework that Andrew Charles Jones wrote up breaks the theory down visually — it's one of the cleaner explanations of how Cover's proof translates into actual portfolio construction. If the original paper is too heavy, start there.

The key word in Cover's framing is universal. The strategy doesn't need to know anything about the stocks in advance. It works across the class of possible outcomes. That's what makes it remarkable.

The video touches briefly on mean variance theory — the dominant framework in academic finance since Harry Markowitz published in 1952. The idea: maximize expected return for a given level of risk. Build a portfolio that sits on the efficient frontier. Classic stuff.

Cover's work cuts at that from a different angle.

Mean variance is a single-period model. It asks: given what I know now, what's the best allocation? Cover's universal portfolio is multi-period. It asks: what strategy keeps compounding over time, regardless of what I know?

Those are different questions. And the second one is the one that actually matters if you're a real investor with a real time horizon.

The distinction connects to the Kelly Criterion — the bet-sizing formula that tells you how much of your bankroll to put at risk each round to maximize long-run growth. Kelly is also about repeated decisions under uncertainty, not single-shot optimization. Cover's approach lives in the same intellectual neighborhood.

The video frames it simply: you're choosing a portfolio B, and you want a good basis for action. Not the best action in one moment. A good basis — something that holds up over time, across outcomes you can't predict.

Here's what volatility harvesting actually looks like.

You hold Stock A and Stock B, 50/50. Stock A goes up 20%. Now your portfolio is roughly 55/45. You rebalance: sell some A, buy some B, back to 50/50. Next week Stock B goes up 20%. You rebalance again. You keep doing this.

Each rebalance locks in a small gain from the move. The stocks don't have to trend upward for this to work. They just have to keep moving.

The five rules from this breakdown of Cover's approach make it concrete — it's the clearest practical translation I've seen of what the academic proof actually implies for how you run a portfolio. Rule one is essentially: rebalance. Not because it feels disciplined, but because that's where the mathematical edge lives.

This is counterintuitive. Most investors think of rebalancing as a risk-management chore — a way to stop the portfolio from drifting. Cover says no: rebalancing is the return-generating mechanism. The drift is what you're harvesting.

Two mediocre stocks. No predictions. $7 million. The mechanism is simple. The implications are not.

Cover wasn't primarily a finance person. He was an information theorist. His most important book — Elements of Information Theory, co-authored with Joy Thomas — is one of the standard graduate texts in the field. Entropy, channel capacity, data compression. That's his world.

The universal portfolio result is information theory applied to markets. He was asking: how much information does a portfolio strategy need to perform well? His answer: none, if you rebalance correctly.

That framing changes how you think about what an edge actually is. Most market participants think the edge is information — knowing something others don't. Cover's result says there's a structural edge available that requires no information at all. Just the right process, repeated.

This doesn't mean the strategy is magic. Transaction costs matter. Tax drag matters. The two-stock example is an illustration, not a trading manual. But the underlying proof is real, and the intuition it builds — that process can compound without prediction — is genuinely useful.

It also makes you wonder how many other edges look like information advantages but are really just structural advantages dressed up in a story.

A 40-second video references a 1991 Stanford paper, a Wikipedia page on information theory, a Kelly Criterion explainer, and a graduate textbook. Most viewers catch none of that. The links are there, on screen, for a second — and then gone.

That's the problem VidLink is built around. Video surfaces things — papers, tools, frameworks, books — and normally gives the viewer no way to act on them. You'd have to pause, screenshot, search, and hope you find the right thing. Most people don't.

Cover's work deserves better than a screenshot hunt. The 1991 paper is publicly available. The explanations exist. The connections to Kelly, to information theory, to practical rebalancing — all of it is one click away, if the video is built to make that possible.

This video is exactly the kind where the links are the content.