Why Independence Matters

Imagine you own five businesses. All five sell winter coats. All five are located in the same city. When winter is cold, you do well. When winter is warm, all five suffer together.

You don't really own five businesses. You own one business with five doors.

Now imagine a different version. One business sells winter coats. One runs a software company. One operates a laundromat. One grows avocados. One manages rental apartments. A bad month for coats has nothing to do with avocados. A slow quarter for software doesn't touch the laundromat.

This is the most important idea in portfolio construction. And it is the most misunderstood.

Most people think investing is about picking winners. Find the best strategy. Find the highest return. Concentrate your capital on the thing you believe in most. This logic is intuitive. It's also wrong, at least for a portfolio.

Here's why.

Two strategies that each return ten percent but move independently of each other will, when combined, produce a portfolio that is dramatically more stable than either one alone. The returns add up. The volatility doesn't. Because when one zigs, the other is doing its own thing. The bad months don't line up. The drawdowns don't stack.

Meanwhile, a single strategy returning fifteen percent sounds better. But it carries all its risk in one basket. When it draws down, the whole portfolio draws down. There's no offset. No cushion. Nothing to smooth the ride.

The math here is almost too elegant. When two return streams are genuinely independent, meaning they have zero correlation, combining them improves the risk-adjusted return of the portfolio far more than adding the same amount of return from a correlated source. It's not additive. It's multiplicative. Two modest, independent engines create a portfolio that is stronger than either one on its own.

This is correlation. Not as an academic concept, but as a lived reality for anyone trying to compound capital over years.

And yet most portfolios ignore it entirely.

Walk into any traditional fund and look at what they run. A long equity strategy based on value. Another based on momentum. A third based on quality. They'll tell you they're diversified. Three strategies. Different names. Different factor exposures.

But watch what happens when the stock market drops five percent. All three fall. Maybe by different amounts. But they all fall. Because all three are, at their core, expressions of the same bet: that equities will go up. The diversification is on the surface. Underneath, the positions are moving together.

These are five coat shops in the same city.

Real diversification, the kind that actually protects capital, requires strategies that are genuinely independent. Not just different names or different factors, but different sources of return. A strategy that profits from equity trends and a strategy that profits from the way CEOs talk on earnings calls are not the same kind of bet. They draw from different information. They respond to different forces. When one is having a bad month, the other has no reason to care.

When the system evaluates a new strategy, the first question is not "how much does it return?"

The first question is: does it move independently from everything we already run?

A strategy with modest returns and zero correlation to the existing portfolio is worth more, in portfolio terms, than a high-returning strategy that moves with the market. This feels counterintuitive. It runs against every instinct that says "find the best performer and give it the most capital."

But portfolios don't work on instinct. They work on math. And the math is unambiguous.

Think of it this way. If you add a strategy that returns ten percent with zero correlation to your portfolio, you've added ten percent of return and almost nothing in terms of additional risk. The portfolio gets better on both dimensions: more return, better stability.

If you add a strategy that returns fifteen percent but correlates heavily with what you already own, you've added fifteen percent of return but also a proportional amount of risk. Worse, you've added risk that concentrates in the same direction. The drawdowns will be deeper. The bad months will be worse. The compounding gets interrupted more often.

Over a year, the high-return correlated strategy looks better on paper. Over a decade, the modest uncorrelated strategy wins. Because compounding rewards consistency. Every deep drawdown costs you time. Time spent recovering instead of growing. The smoother the ride, the faster the compounding. This is the quiet advantage that most investors never see.

There's a practical implication here that matters a great deal. It means that some of the best strategies in a portfolio are the ones that look least impressive on their own.

We run strategies whose standalone returns are unremarkable. Decent, not spectacular. If you looked at them in isolation, you might wonder why they're in the portfolio at all. But they move independently from everything else. When equity markets sell off, these strategies are doing their own thing, driven by their own signals, responding to their own data. They don't know the stock market is falling. They don't care.

That independence is what makes the portfolio resilient. Not any single strategy's brilliance. The fact that the whole is structurally different from a collection of things that move in lockstep.

This principle also explains why we kill strategies that most funds would keep. We've built strategies with genuinely strong returns. Clean equity curves. Sound economic logic. And we've killed them because their correlation to the existing portfolio was too high. The marginal contribution to the portfolio was small, because the risk they added was the same risk we already had.

More of the same isn't diversification. It's concentration with a different label.

The hardest part of this discipline is saying no to something that looks good. A strategy with an attractive return is hard to turn away from. The instinct is to find room for it, to carve out allocation, to add it to the mix. But if it moves with something you already own, you're not adding a new engine. You're making an existing engine louder.

Louder isn't better. Different is better.

There's one more thing worth saying about independence, and it has to do with crisis periods specifically. In calm markets, correlations between strategies tend to be lower. Everything is operating on its own fundamentals. The independence is easy to maintain.

In crisis markets, correlations spike. Strategies that seemed independent start moving together. This is the moment that reveals whether your diversification was real or cosmetic. If your "independent" strategies all fall during the same week, they weren't independent. They were all quietly tied to the same underlying risk, and the crisis revealed the connection.

True independence holds during stress. A strategy built on a genuinely different source of information, responding to genuinely different market forces, has no reason to fall just because equities are falling. It might. By coincidence, on any given day. But structurally, it's disconnected. And structural disconnection is the only kind that matters when you need it most.

We test correlation not just in normal markets but specifically during stress periods. The correlation between two strategies in the calmest year on record is interesting. The correlation between them during the worst month is essential. If it stays low when everything else breaks down, the independence is real.

This is why we obsess over correlation the way most funds obsess over returns. Returns are the thing everyone sees. Correlation is the thing that determines whether you survive long enough to collect them.

Independence isn't a feature. It's the foundation.