Why 'Diversification' Doesn't Protect You In Crashes
TL;DR
Owning twenty tech stocks isn't diversification — it's the same trade in twenty costumes. We break down correlation: how to measure real portfolio risk, why correlation spikes toward one in crashes (when you need diversification most), and what actually protects a portfolio.
“Owning twenty tech stocks isn't diversification — it's the same trade in twenty costumes. We break down correlation: how to measure real portfolio risk, why correlation spikes toward one in crashes (when you need diversification most), and what actually protects a portfolio.”Click to post on X ▸
Where this fits in the Confluence Method
This lesson lives in the Stack step of the Confluence Method, where you confirm momentum before a setup qualifies as a trade. It also reinforces the risk and psychology that let the edge compound over many trades.
Read the full method ▸Full transcript
7 sections0:03Most retail investors think they're diversified because they own twenty different ticker symbols. In reality they own one trade — long technology, long growth — twenty times. In a normal market the twenty positions seem to move somewhat independently and create the illusion of diversification. In a crash — when diversification is the only thing that matters — correlations spike toward one. Every position drops together, the 'diversification' disappears, and the portfolio behaves like one massive bet on the sector. Today: the correlation math that defines real diversification, why correlations spike in crashes, and what actually protects a portfolio when conditions go bad.
0:40Here's the math nobody shows you. Diversification is defined by CORRELATION, not by the number of positions. Two stocks with correlation close to one act like one trade — they move together, they fall together, owning both adds no risk reduction over owning one. Two assets with correlation close to zero act independently — owning both genuinely diversifies because one can rise while the other falls. The number of names you hold tells you nothing; the correlations between them tell you everything. And in crashes, correlations across risk assets historically spike toward one — the exact moment you needed diversification, the diversification disappears.
1:23Watch this in a synthetic crash. An investor holds twenty tech stocks — Apple, Microsoft, Nvidia, twenty different names. In good times the portfolio looks diversified; individual names sometimes lead, sometimes lag. Then a crash hits. All twenty positions drop together, deeply, simultaneously. The portfolio loses thirty percent in three weeks. The diversification provided zero protection because the underlying correlation was near one. Twenty positions, one trade. The investor felt safe because they 'spread the risk' across twenty names, but the correlations meant they were always concentrated in one bet.
2:00Here's what real diversification looks like. Different ASSET CLASSES, not different names within an asset class. Equities, bonds, commodities, cash, possibly currencies or real estate. These tend to have low or even negative correlations with each other — when stocks crash, bonds often rally; when inflation spikes, commodities outperform. The classic sixty-forty portfolio works specifically because of low equity-bond correlation, not because of name diversification within equities. The goal isn't owning many things; it's owning DIFFERENT things — measured by correlation, not by sticker count.
2:32Now with real diversification. The portfolio splits across uncorrelated assets — equities, bonds, commodities, cash. Same crash hits. Equities drop thirty percent, but bonds rally as rates drop. Commodities stay roughly flat. Cash holds value and earns. Net portfolio drawdown: maybe eight to twelve percent instead of thirty. The recovery is faster because the equity drawdown is partially offset by the bond rally. Same risk capacity, completely different outcome. The reason isn't financial sophistication — it's just correlation math. Different things behave differently because they ARE different. Twenty tech stocks behave the same because they ARE the same.
3:12In real practice, audit your portfolio. Calculate pairwise correlations between your positions using ninety-day rolling data. If most pairs sit above zero point seven, you have a concentrated bet wearing a diversified costume. If many sit below zero point three, you have genuine diversification. Add asset classes — bonds, commodities, possibly currencies — until your average correlation drops meaningfully. Position count is irrelevant; correlation is everything. Most retail portfolios fail this test badly and don't realize it until the crash exposes the concentration.
3:50So: diversification is defined by correlation, not by position count. Owning twenty tech stocks is one trade twenty times. Real diversification requires uncorrelated asset classes — equities, bonds, commodities, cash — measured by their historical and stress-period correlations. Audit your portfolio for hidden concentration before the next crash makes the audit involuntary. Subscribe for the full method, and trade your own plan. Education, not financial advice.
