Buy Borrow Die, Explained: The Catch Is in the Middle Leg

Buy borrow die is a three-leg strategy: buy assets that compound, borrow against them instead of selling, and hold them until you die, when the tax treatment of the estate is what makes the whole thing famous. The appeal is funding a life without ever realising a gain. What almost nobody says out loud is that two of those three legs are automatic. You do not operate them. The middle one you operate for decades — and it is the only leg you operate that can end the strategy early.

What is buy borrow die?

Three legs, in order. Buy appreciating assets and hold them, so gains accrue unrealised. Borrow against those assets when you need cash, because a loan is not a sale and a sale is what triggers the gain. Die holding them, at which point the estate's tax treatment of those unrealised gains is the thing the strategy is named for.

Look at what you are actually asked to do in each. Buy: one decision, then nothing. Die: no decision, and no exertion. Borrow: a live, callable obligation sitting against a portfolio that moves every day, for however many decades separate the first two.

I run a leveraged book myself, so let me be plain about the asymmetry. The two legs everybody discusses are the two you do not operate. The one nobody discusses is the one you do.

The leg you do not operate

The third leg is a tax mechanic — the treatment of cost basis in an estate. It is what makes buy borrow die famous, and it is the reason the phrase circulates at all.

It is also tax law: it changes, it varies by jurisdiction and by the particular estate, and it needs a real CPA or an estate attorney looking at your actual situation. Incomestead has nothing to say about it and is not qualified to. You will find no threshold, rate or structure advice here, and you should be suspicious of a blog that offers one.

Which is convenient, because it is also the leg you do not operate. The one you operate is the middle one.

The benefit at leg three only pays if you are still holding the assets when you get there. Leg two is what decides whether you are.

How do rich people borrow against stock?

They pledge the shares they already own as collateral — a margin loan inside the brokerage, or a securities-backed line of credit outside it — and draw cash against them without selling. The mechanics, and the underwriting a real lender would apply, are set out in the brokerage family bank. This post takes the facility as given and asks a different question: does it survive the holding period the strategy specifies?

Here is the book we will test. Illustrative, shaped like mine, single currency: $3,750,000 of gross securities value, against a margin loan of $975,000. On this book there is no cash line — the loan is the negative settled balance.

Our reading is margin loan ÷ gross securities value. Not net of the loan, not including outside cash. On this book that is 26.00%, which sits in Clear — the calmest of the four zones we govern to, and inside the 30% line we treat as the ceiling of Clear, our own convention for how much margin is safe.

Our two-week persistence rule requires a breach to appear in two consecutive weekly readings before it drives any action, so a single volatile Friday cannot conscript you into selling. A book in Forced is the one exception — it is acted on in the week it is first read, because a second opinion costs a week of exposure in the one zone where a week is expensive.

What follows are two entirely separate ways this Clear book stops being Clear. No figure below combines them — the first holds the loan constant, the second holds the market flat, so treat each as a floor rather than a forecast.

What if the market falls and the loan does not move?

This is the one readers miss, and it is not subtle once you see it. The loan is the numerator. It does not fall with the market. The collateral is the denominator, and it does. So the reading rises without you doing anything — no trade, no drawdown taken, no decision made.

On this book, holding the loan at $975,000 exactly where it is: a fall of 13.34% takes it out of Clear. A fall of 25.72% reaches Freeze. And a fall of 35.00% reaches Forced.

Market fall	Securities	Loan	Reading	Zone
None — today	$3,750,000	$975,000	26.00%	Clear
20%	$3,000,000	$975,000	32.50%	Harvest
35%	$2,437,500	$975,000	40.00%	Forced
50%	$1,875,000	$975,000	52.00%	Forced

Note the third row. Under our convention 40.00% is Forced, not Freeze — the boundary rule is set out in the margin pillar. It carries a second consequence: a book that reaches Forced is capped at 34 on the Governance Score, however disciplined the rest of the operation is.

So: is a 35% fall the sort of thing that happens? Apply the arithmetic to every S&P 500 bear market since 1990. Peak-to-trough figures come from Yardeni Research, Stock Market Historical Tables: Bull & Bear Markets, as of its 21 January 2024 edition. This is arithmetic applied to a hypothetical book, not a forecast, and not a claim that this book tracks the index.

Does the book pay its own interest?

Now the second path, which needs no crash whatsoever. Hold the securities flat at $3,750,000 and watch the loan alone.

At an assumed rate of 6.00%, the first year's interest on $975,000 is $58,500. The question is who pays it — and the answer is not "you, by selling." Dividends, coupons and option premium are cash generated inside the book without a sale, and they credit against the debit automatically, so living by the rule "never sell" does not force the interest to capitalise. Premium is the weaker form of that income, and the family bank pillar covers why.

Which leaves one line worth knowing exactly — and it is exact, because it is algebra rather than projection:

Break-even yield = your reading × your rate. Below that yield, your debit grows every year on its own.

On this book: 26.00% × 6.00% = 1.56%. That is the income yield, on the whole book, at which realised income exactly meets the interest bill — the same $58,500. It assumes nothing about the market and nothing about time, and it is stated before tax.

At or above break-even the debit is static or shrinking at the rate you are paying today. Below it the shortfall is added to the debit and next year's interest is charged on the larger balance — so the reading climbs even in a completely flat market, with no crash and no decision. That is the whole of the second path.

How many years it takes to reach Forced is the question everyone asks next, and it is the one to be most careful with. On this same book, five defensible sets of assumptions give year 5, year 8, year 16, year 25 — and never. The width is the finding, and the full bracket, and why nobody can honestly hand you a single year, is worked through separately.

One more thing the tidy separation hides: the two paths are correlated. A deep fall is also when dividends get cut and premium-writing capacity shrinks. Keeping them apart is pedagogy, not a claim that they are independent.

The Middle-Leg Test

Three steps, on numbers you already have.

If your interest in all this is funding a retirement rather than a lifestyle, the borrow-don't-sell retirement model and its own catch takes the same mechanic into decumulation.

Is this anywhere near a margin call?

No. Every zone in this post is our own convention, self-imposed and set far above any regulatory trigger — and because the two are inverse measurements, the translation matters: a 40% loan against gross securities leaves 60% equity, against FINRA's general minimum that "Generally, a customer's equity in a margin account that holds margin stocks only must not fall below 25 percent of the current market value of the long securities."

It is not our line that liquidates anyone, either — a broker's own house requirement is what does that, and those can be raised without advance notice. And the broker does not have to warn you — FINRA, Margin Calls is explicit: "Firms don't have to issue a margin call before selling securities in your margin account to meet a margin call and may sell enough securities to completely pay off your margin loan, not just meet the margin call." The family bank pillar sets out the full comparison.

The catch, stated once

Buy borrow die is real, legal, and widely practised. The catch is not that it is a trick. The catch is that the famous benefit sits at the end of a holding period the middle leg has to survive, and whether it survives turns on a question almost nobody has answered on their own book: does the income cover the interest?

And there is a closing irony that belongs in plain sight. If the fall comes and a book is liquidated to bring the reading down, that sale realises — at depressed prices — precisely the capital gains the entire strategy existed to defer. The failure mode of leg two does not merely interrupt leg three; it undoes it, in the worst possible week, at the worst possible prices. That is why the middle leg deserves more attention than the two famous ones get.

I cannot tell you the year — and anyone who quotes you one has quietly fixed the market, your rate and your withdrawals. What a weekly reading gives you is not the year. It is finding out before the arithmetic does — with one honest limit, and it is mine as much as anyone's: a weekly reading is weekly. It cannot see an intraday move, and a book that gaps through a zone between two Fridays is read after the fact, not during.

Frequently asked questions

What is buy borrow die?

Buy borrow die is a three-leg approach to wealth: buy assets that appreciate, borrow against them rather than selling so no gain is realised, and hold them until death, when the estate's tax treatment of those unrealised gains is what the strategy is named for. The third leg is tax law — it changes, it varies by jurisdiction and estate, and it needs a CPA or an estate attorney rather than a blog. The practical point is that buying and dying are automatic, while borrowing is operated continuously for as long as the strategy runs.

How do rich people borrow against stock?

They pledge shares they already own as collateral — a margin loan inside a brokerage account, or a securities-backed line of credit outside one — and draw cash without selling, so the position keeps compounding and no gain is realised. Opening the facility requires a signed agreement and the lender's approval; after that, each drawdown typically needs no application and no underwriting, which means the borrower has to supply the underwriting themselves.

What is the catch with buy borrow die?

The benefit at the third leg only pays if the assets are still held when you get there, and the borrowing leg decides that. Two things can end it early. A loan does not fall when the market falls, so a decline raises the loan-to-collateral reading with no action by the borrower. Separately, if the book's own realised income does not cover the interest, the shortfall is added to the debit and compounds, and the reading rises even in a completely flat market. There is also an irony worth naming: a liquidation forced by either path realises, at depressed prices, exactly the gains the strategy existed to defer.

What margin zones does Incomestead use?

Our reading is margin loan divided by gross securities value — not net of the loan, and not including outside cash. Incomestead's definition of the four zones is: Clear below 30%, Harvest from 30% to under 35%, Freeze from 35% to under 40%, and Forced at 40% and above. These are our own convention, self-imposed and set far more conservatively than broker or regulatory triggers, and they say nothing about what is appropriate for your circumstances.

Stefano Starkel is a practitioner, not a licensed adviser, and runs a leveraged portfolio under the governance process described here. This article is education, not personalised financial, tax or legal advice. Margin borrowing amplifies losses as well as gains and can result in the forced sale of your securities without prior notice; in a sharp decline it is possible to lose more than you invested and to owe the broker the difference. Figures are illustrative and computed for this article; the interest rate and the yield shown are assumptions, not quoted figures. Nothing here is a recommendation to borrow any particular amount, or at all.

Incomestead recommends. You decide.