Finance 101 says leverage raises risk. Finance 101 assumes you started diversified. Start with a concentrated position and the math flips: you can add leverage to a single-stock fortune and end up with less risk than you had, because concentration risk falls faster than leverage risk rises.
The setup. You’re sitting on $10 million of one stock, basis near zero. You know you should diversify. Selling means a ~37% combined federal and California capital gains haircut, so you do nothing, and every earnings report becomes a personal macro event. The concentration is the risk - but the tax bill is the cage.
The unlock is borrowing instead of selling. A box spread is an options construction that behaves like a loan: sell the box, take cash today, owe a fixed amount at expiration, at an effective rate near Treasuries - cheaper than any margin loan, and the interest can be deductible against investment income. No sale, no gain, no tax.
Then the flywheel. Deploy the borrowed cash into a long/short strategy built from subsector ETFs - diversified market exposure that realizes losses while tracking the market. Those losses offset gains, which means you can now sell slices of the concentrated position tax-free. The sale proceeds become fresh dry powder for the same strategy, which generates new losses, which frees the next slice. Every cycle shrinks the concentrated position and grows the diversified one. Over time you recycle out of the single stock entirely - while the leverage stays on.
That’s the free lunch. Leverage added, risk reduced, because the diversification you bought with borrowed money matters more than the borrowing. Diversification is the only free lunch in investing. This is how you pay the toll without paying the IRS.
The caveat is execution. An individual can run this with subsector ETFs - no manager, no fee layer - but box spread leverage and tax-loss harvesting demand meticulous execution, the kind most professionals handle with automation tools. That doesn’t change the direction of the math: for the concentrated, the riskiest position is the one you’re already in, and sometimes the way out runs through more borrowing, not less.
Assumptions
The claim is conditional: diversification can reduce a concentrated portfolio’s risk enough to outweigh added borrowing. Leverage alone does not make a portfolio safer.
- Starting position and tax: $10 million in one stock, near-zero basis and roughly 37% combined capital-gains tax are illustrative inputs. If the entire value were gain, that rate would imply roughly $3.7 million of sale tax. Actual basis, gain character, state, income and usable losses change it.
- Borrowing: A short box exchanges cash now for a fixed payment at expiration. Its rate must be calculated from actual net proceeds, maturity payoff, dates, commissions and spreads. “Near Treasuries” is a pricing hypothesis, not a live quote or proof it beats every margin loan. Broker collateral, margin calls and refinancing are part of the cost and risk.
- Tax treatment: No stock sale is assumed when borrowing. Treatment of the options contracts and any deduction is separate. Section 1256 classification, capital-loss netting and investment-interest limits must be checked; do not count the same financing amount as both an interest deduction and a capital-loss benefit.
- Risk and flywheel: A test needs the loan size, long/short exposures, stock and ETF volatility, correlations, return paths, margin rules and drawdowns. Those inputs are not supplied here, so the article does not quantify a risk reduction. Loss harvesting requires an actual loss, an eligible replacement and usable gains. It can stall in a rising market; zero tax on a stock sale requires sufficient allowable offsets. Interest, trading costs and taxes cannot be assumed away.
Have better inputs, a missing cost or evidence that changes the result? Share it . We will update our assumptions and conclusions to reflect the most accurate representation.