The Concentration Problem: Why Your Portfolio Probably Has a Single Point of Failure
Open your brokerage app. Look at the pie chart. Five colors, six colors, maybe eight. Feels balanced.
It's not.
The top 10 stocks in the S&P 500 now make up 37% of the entire index. Highest since the dot-com bubble. If you own VOO or SPY, a third of your money sits in ten companies. Apple, Microsoft, NVIDIA, Amazon, Meta, Alphabet, Broadcom, Tesla, Eli Lilly, Berkshire. That's not diversification. That's a mega-cap tech bet wearing a trench coat.
Zoom out to your full portfolio and it gets worse.
How concentration sneaks in
RSU vests you never sold. An engineer at NVIDIA who vested $200K over two years and held? That position is worth $600K+ after the run-up. Total portfolio: $900K. NVIDIA alone: 67%. One guidance miss and a third of their net worth is gone before market open.
I've seen portfolios where a single stock is 40%, 50%, 70% of total value. The owner always knows it's risky. They've just never run the numbers on how risky.
Winners that drifted. You bought NVIDIA at $150. It ran. You didn't trim because selling winners feels wrong and triggers taxes. Now that $10K position is $45K and represents 25% of your portfolio. You didn't choose a concentrated bet. The market chose it for you.
This is portfolio drift. It happens to every buy-and-hold investor. Winners grow, losers shrink, and your careful allocation slowly becomes a momentum bet on whatever ran hardest.
The index fund illusion. "I'm diversified, I own a total market fund." VTI holds 3,600+ stocks. The top 10 represent 33% of the fund's value. The bottom 3,000 represent less than 20%. If Apple drops 20%, your "diversified" fund drops 2.3% from that single holding. If the top 10 all sell off 20% in a correlated risk-off event, you're down 6.6%. That's not a tail risk. That's the base case during real market stress.
Inherited positions you can't bring yourself to sell. 5,000 shares of J&J from your parents. Cost basis: $12/share from 1985. Current value: $800K. Selling triggers a $740K capital gain. So you hold. And one stock becomes 40% of your portfolio because the tax bill feels scarier than the concentration.
How to measure it
Pie charts lie. You need actual math.
The Herfindahl-Hirschman Index (HHI). Same formula the DOJ uses for antitrust. Square each position's portfolio weight, sum them.
HHI = Σ(weight²)
Under 0.15 = diversified. 0.15 to 0.25 = concentrated. Over 0.25 = one bad week away from real pain.
Twenty equally-weighted positions: HHI of 0.05. One stock at 50%: HHI above 0.25 no matter what else you hold. The math doesn't care how many small positions you use as filler.
The 30% drawdown test. Take your largest position. If it drops 30% tomorrow — a normal drawdown for an individual stock — what happens to your total portfolio? If the answer is "down more than 5% from one name," that's a concentration problem.
The sector check. AAPL, MSFT, GOOG, NVDA, META. Five tickers. One sector bet. If you're 60%+ in tech across all positions, a sector rotation doesn't hit one holding. It hits everything at once. Diversification means your positions don't all move in the same direction on the same day.
What it costs
$500K portfolio, well diversified (HHI 0.08). Severe market downturn. Expected max drawdown: about 25%. You lose $125K. It hurts. You recover.
Same $500K but 45% in one stock (HHI 0.23). Same downturn hits that stock harder. Max drawdown: 40%+. You lose $200K.
The extra $75K is the concentration tax. You pay it during every major drawdown. It doesn't fund anything useful. It just disappears.
Fixing it
See the full picture first. Most people have 2-3 accounts. A brokerage, a 401(k), maybe an IRA. Each one looks fine alone. Combined, the concentration becomes obvious. Your brokerage won't show you this view because they only see what they hold.
Tax-loss harvest before you rebalance. Got underwater positions elsewhere? Sell those to generate losses that offset the gains from trimming your winners. A $15K loss somewhere in your portfolio turns a $15K taxable gain from selling concentrated stock into a $0 tax event. I wrote a longer breakdown of this here. There's also a free calculator if you want to estimate the savings.
Set threshold rules. "If any position exceeds 25% of my portfolio, I trim to 20%." Write it down. Put it on a sticky note. Rules work because they remove emotion. Emotion is the reason you're concentrated in the first place.
Replace, don't just sell. Trim NVIDIA and buy SMH (semiconductor ETF). You keep sector exposure while killing single-stock risk. Same principle works for any concentrated position — find the ETF that covers the same sector or theme.
Monitor it. Concentration isn't a one-time fix. Prices move every day. The position you trimmed to 20% last quarter could be back at 28% after one good earnings print. You need something watching for you, not a quarterly reminder to open a spreadsheet.
What I'd do this week
Open your largest account. Find your single biggest holding. Calculate what percentage of your total portfolio it represents — not just that account, your total across everything.
If it's over 20%, you have work to do.
If you want to skip the spreadsheet, Helm connects to your brokerage via Plaid and calculates your concentration risk score across all accounts automatically. The portfolio tools are free — not trial-limited, actually free. Tax-loss harvesting automation is $14.99/mo if you want that too.
But the spreadsheet works fine. The tool doesn't matter. What matters is that you look at the number. Because the expensive mistake isn't holding a concentrated position. It's holding one and not knowing it.