How to Diversify After a Concentrated Loss (The Rebuild Math)

“Diversify more” is the most common advice given to someone who just got hurt by concentration risk, and it’s almost useless as stated. It doesn’t say how much is enough, what counts as genuinely different exposure versus the same bet twice, or how to rebuild when most of what you had is now gone. This is the version with actual numbers.

The math that got you here

Concentration risk isn’t really about being “too aggressive” in the abstract — it’s a specific mathematical property: how much of your total outcome depends on one input being right.

Say a portfolio was $40,000, with $34,000 of it — 85% — in a single position. That’s not an unusual allocation for someone who found a coin or a stock early, watched it run, and kept adding because it kept working. The remaining $6,000 sat in cash or a few smaller holdings.

Then the position drops 70%. That $34,000 is now worth $10,200. The rest of the portfolio, untouched, is still roughly $6,000. Total portfolio: about $16,200 — a 59.5% drawdown on the whole account, driven almost entirely by one line item.

Here’s the part that’s easy to miss in the moment: the 70% drop in the asset and the 59.5% drop in the portfolio are two different numbers, and the gap between them is exactly the size of the concentration problem. If that same $34,000 had been spread across ten uncorrelated positions and only one of them dropped 70%, the portfolio-level damage would have been a fraction of what actually happened. The asset did what it did. The portfolio-level outcome was a sizing decision, made well before the drop, that determined how much that one asset’s behavior could do to you.

That’s the distinction worth sitting with before doing anything else: the loss already happened, but the amplitude of the loss was a structural choice you made, not something the market did to you unprompted. That’s also good news, because it means the fix is structural too, not a prediction you have to get right next time.

Why “diversify more” fails as advice

It fails because it doesn’t specify two things that actually matter: a cap on any single position, and a check for whether your other holdings are genuinely different bets or the same bet wearing different tickers.

The cap. Without a written number, “diversify” quietly erodes back toward concentration the next time something is working. A reasonable starting framework, adjustable to your own risk tolerance:

• Core, high-conviction positions: cap around 10-15% of total portfolio each. This is already a meaningful bet — enough to matter if you’re right, small enough that being wrong doesn’t define the year.
• Medium-conviction positions: 3-7% each.
• Speculative or early-stage bets: 1-2% each, sized so a total loss on any single one is a rounding error, not a setback.

These aren’t universal constants — a 25-year-old with a long horizon and stable income can reasonably run higher single-position caps than someone close to needing the capital. The point isn’t the exact percentage; it’s having one, written down, before the next position is sized, the same way a stop-loss or drawdown rule needs to exist before you’re emotionally attached to the outcome.

Correlation. Five different altcoins are not five different bets if they all move together on Bitcoin’s beta — that’s one bet, sized five times, dressed up as diversification. The same applies to owning several growth-stage tech stocks that all rise and fall on the same rate-sensitivity story. Genuine diversification means checking what actually drives each position’s return — and deliberately including things that don’t share a driver: a different asset class, a different sector, a different macro sensitivity, or plain cash sitting outside the risk pool entirely.

Rebuilding from what’s left

Continuing the example: $16,200 total, with $10,200 (63%) still sitting in the position that just dropped 70%. Two mistakes are common at this exact point, and they pull in opposite directions.

The first is holding the full remaining stake because “it’s already down so much, selling now locks in the worst of it” — which is the sunk cost fallacy showing up in portfolio-construction language instead of trading language. The position’s future return doesn’t care what you paid or when you sold relative to the bottom.

The second is dumping the entire remaining stake immediately to “fix” the concentration in one move — which can make sense, but often isn’t necessary and can force you into a single bad price and a single tax event when a staged approach would do the same job with less friction.

A middle path that fits the caps above: trim the position down toward its target cap (say 15% of the current, smaller portfolio) over a few tranches rather than one sale, redirecting the proceeds into a small number of positions with genuinely different drivers — not five more names in the same sector chasing the same recovery narrative. If there’s no new capital to add, this trimming-and-redistributing process is the rebuild; it doesn’t require fresh money to start working.

The trap that undoes this

The most common way people re-concentrate almost immediately after a loss like this isn’t ignorance of diversification — it’s finding “the next one” that feels certain enough to justify sizing up again. That instinct has a name and a predictable shape; if it sounds familiar, the sunk-cost and conviction-inflation pattern that drives it is worth checking against your own reasoning before the next position gets sized the same way the last one did.

Write the position cap down somewhere you’ll actually see it before the next trade — not as a New Year’s-resolution level of intention, but as a specific number attached to a specific account. A rule you can look at is a rule you can follow under pressure. A rule you only remember having had isn’t one.

FAQ

Doesn’t diversifying just guarantee average, boring returns? Diversification caps your upside on any single winner — that’s real, and worth being honest about. But the trade you’re actually making isn’t “big gains vs. mediocre gains,” it’s “a portfolio that survives being wrong about one thing vs. one that doesn’t.” You can still run a high-conviction, high-upside position; the rebuild here just means sizing it so that being wrong about it doesn’t wipe out the whole account again.

If I trim my surviving positions to rebalance, doesn’t that trigger a taxable gain? In many jurisdictions, yes — selling any appreciated position to rebalance is generally a taxable disposal, separate from whatever losses you’re carrying elsewhere. Realized losses elsewhere in your portfolio can sometimes offset that gain, but the specifics depend heavily on your jurisdiction and situation, so don’t assume a specific outcome without checking your own numbers.

How many positions is actually enough to call a portfolio diversified? There’s no magic number, but somewhere in the range of 8-15 positions across genuinely different return drivers is a common practical floor for someone managing their own portfolio without full-time attention. Fewer than that and a single bad outcome can still dominate the whole portfolio; far more than that and you’re likely just adding tracking overhead without much additional risk reduction, especially if the extra names are correlated with what you already hold.

What if I don’t have new capital to diversify with — everything is tied up in the one position? Then the rebuild happens gradually, out of the position itself, rather than by adding new money. Trim the oversized position down toward your target cap in stages rather than all at once — this reduces concentration risk sooner without forcing you to sell the entire remaining stake into a single price and a single tax event.