Diversification across four asset classes

Diversification has been called the only free lunch in finance, which is roughly true and widely misunderstood. The free part is real: combining assets that do not move together reduces the violence of your equity curve without necessarily reducing its long-run growth. The misunderstood part is everything else — what counts as diversification, what it protects you from, and what it quietly fails to protect you from at the worst possible moment.

This lesson treats diversification as a trader’s tool rather than a retirement-brochure slogan. The core idea fits in a sentence: spreading risk is not about owning many things. It is about owning things that fail for different reasons.

What diversification actually does

Hold one asset that swings ±20% and your account swings ±20%. Hold two assets that each swing ±20% but do not move in lockstep, and the combined swing is smaller, because one is often zigging while the other zags. Your expected return is the average of the two; your volatility is less than the average. That gap — same return engine, smoother ride — is the entire free lunch.

• It reduces the damage from any single position being wrong, which no amount of analysis can fully prevent.
• It smooths the equity curve, and smoothness compounds: smaller dips need smaller recoveries, a point the drawdown lesson makes painfully precise.
• It protects against single-asset catastrophe — the fraud, the delisting, the overnight gap through your stop.

What it does not do

Diversification does not improve the expected return of anything you hold. Ten mediocre positions, diversified, are still mediocre — you have simply arranged to lose money more smoothly. If a strategy has negative expectancy, diversification gives it a quieter ride to zero, which is arguably worse because it takes longer to notice.

• It does not protect against market-wide crashes, when correlations converge and nearly everything falls together — the next lesson covers why.
• It does not substitute for position sizing. Twenty positions at reckless size is reckless twenty times.
• It does not turn a losing process into a winning one. It only changes the shape of the results, not their direction.

The naive diversification trap

Here is the trap that catches most crypto-native traders. A portfolio of ten different coins feels diversified — different teams, different technology, different narratives. Then the market turns and all ten drop 15% on the same day, because they all share one underlying risk driver: appetite for crypto risk. You did not have ten positions. You had one position with ten names on it.

The same illusion shows up everywhere once you know its shape:

• Ten cryptos is one bet on crypto risk appetite, with cosmetic variation.
• Eight high-growth tech stocks is largely one bet on rates and risk sentiment.
• Long EUR/USD, long GBP/USD and long AUD/USD is mostly a single short-dollar position wearing three hats.
• Gold, silver, platinum and palladium together are mostly one metals bet — silver and platinum add industrial exposure, but the cluster still moves as a family.

The unit of diversification is the risk driver, not the ticker. Count your bets by asking what would have to happen for each position to lose. If several positions share an answer, they are one bet.

Diversifying across four asset classes

Genuinely different risk drivers tend to live in different asset classes. Stocks respond to earnings and interest rates. Forex pairs respond to rate differentials and trade flows between two economies. Metals respond to real yields, industrial demand and fear. Crypto responds to liquidity conditions and risk appetite, with a culture all its own. None of these relationships is stable — the correlation lesson is honest about that — but they are different enough, often enough, that spreading across them beats stacking within one.

Practically, this used to be a logistics problem: four classes meant four accounts, four interfaces, four fee schedules. On Obsidiate the same account spans 50 cryptos, 30 stocks, 15 forex pairs and 4 metals — 99 instruments — so the barrier to cross-class diversification is no longer operational. The thinking, unfortunately, is still your job.

Concentration is also a choice

None of this means concentration is always wrong. A concentrated book amplifies whatever you bring to it — skill and error alike — and traders with a demonstrated, well-sampled edge in one market sometimes concentrate deliberately. The key word is deliberately: concentration should be a decision you make with smaller per-trade risk to compensate, not a default that happens because you only ever look at one corner of the market. Concentrated books live faster and die faster. Know which trade-off you are making.

A practical structure

• Cap the share of your open risk in any one asset class — a common rule is no more than half of total open risk in a single class.
• Count tightly correlated positions as one position for sizing purposes: three coins at 1% risk each is closer to one trade at 3%.
• Cap total simultaneous open risk across everything — many traders keep it under 4–6% of equity.
• Review your book monthly by risk driver, not by ticker, and ask what single event would hurt the most positions at once.

Key takeaways

• Diversification reduces volatility and single-asset catastrophe risk; it does not raise expected returns or fix a losing strategy.
• The unit of diversification is the risk driver, not the ticker — ten cryptos is one bet.
• Real spreading usually means crossing asset classes, because that is where different risk drivers live.
• Treat correlated positions as one position when sizing, and cap total open risk.
• Concentration is a legitimate choice only when it is an actual choice, made with smaller risk per trade.