Leverage Is a Different Risk

The loan does not move

Margin arithmetic is surprising for one reason: the borrowed amount is fixed while the position is not. Buy $40,000 of stock with $20,000 of your money and $20,000 borrowed, and a 10% fall takes the position to $36,000 and your equity to $16,000 — a 20% loss on a 10% decline. The exposure doubled and so did the change in your capital, which is what leverage means and is not itself alarming. What makes it a different kind of risk is the maintenance requirement. Brokers require equity to stay above a fraction of the position value — 25% is a common floor — and they do not ask politely when it is breached: they liquidate. Write the condition out and the trigger is closer than intuition allows. Equity is P − loan and the requirement is 0.25P, so P − 20,000 = 0.25P, giving P = $26,667, which is a 33.3% fall from $40,000. At that point equity is $6,667, so the account has lost two thirds of its own money on a one-third decline in the stock. Recovering $6,667 to $20,000 requires +200%. Two properties of that result deserve naming. The first is the asymmetry between the market move and the account move: a 33% decline in a stock is an ordinary correction, and at this leverage it is terminal. The second is that the call is a forced sale, so it converts a temporary paper loss into a permanent one and removes the position that would have supplied the recovery. Leverage does not just magnify the P&L; it introduces a price at which someone else makes the decision. The call, in four numbers — Your equity: $20,000 · Position value: $40,000 · Trigger price of the position: $26,667 — a 33.3% fall · Your equity at the trigger: $6,667 — a 66.7% loss ← · Gain needed to get back to $20,000: +200% ← Maintenance requirements rise in volatile markets and brokers may raise them at any time, so the effective distance to a call shrinks exactly when the market is getting worse.

What leverage does to a strategy with an edge

The simulator makes the trade-off visible on a strategy that already earns money. Take the same edge, the same 200 trades and the same sequence, and compare 2% of equity risked per trade with 6% — which is what tripling exposure does, because leverage multiplies the size of each bet in R terms without touching the win rate or the expectancy. At 2% the median worst drawdown is 22.2% and ruin does not appear in the sample. At 6% the median worst drawdown is 56.1% and ruin appears. The median outcome rises, and so does the chance of losing the account, and both of those are the same fact seen from two directions. This is why leverage has to be assessed against the drawdown rather than against the return. A book that earns 20% a year at 3× leverage and draws down 60% is not a better version of one that earns 9% at 12%: for most households it is a version that gets abandoned at the bottom, which converts the return into a loss. The arithmetic of R2 applies at every level: the fall that triggers a forced sale is the fall that ends the compounding, and the recovery happens without the position that caused it. The practical rules that follow are unglamorous. Size a leveraged book off the drawdown you can hold divided by the leverage, not off the return you want. Treat the maintenance requirement as a moving target, because it rises with volatility. And recognise the specific trap of leveraged products that reset daily: a 3× fund does not deliver three times the return of its index over time, because the resets turn volatility into decay — which is the same drag from R1 with a multiplier attached. One strategy, two exposures — 2% of equity risked per trade: median drawdown 22.2% · no ruin in the sample · 6% per trade — leverage tripling the bet: median drawdown 56.1% · ruin appears ← · What rose: the median outcome · What rose with it: the probability of losing the account Leverage interacts with every rule already built: the stop distance, the fraction and the maintenance requirement are all multiplied, so a position that was 1% of risk becomes 3% and a 33% market fall becomes a call. Check the leverage first and the position second.

The rule that replaced the day-trading minimum

For years the binding constraint on small leveraged accounts was the pattern-day-trader rule: more than three day trades in five business days in a margin account, and the account needed $25,000 of equity. The number was a blunt instrument — it had nothing to do with how much risk a given trade carried, it treated a conservative $2,000 day trade and a reckless $200,000 one identically, and it created the well-known absurdity that a funded account was allowed to lose money steadily but not to trade actively. That framework has been replaced. Under the amendment FINRA adopted, the pattern-day-trader minimum equity requirement is superseded by an **intraday margin standard**: the requirement is calculated from the risk of the positions actually carried through the day rather than from a flat account-size threshold, with firms permitted to phase the change in over a transition period that runs into 2027. The practical change is that the constraint moves from **account size** to **position risk**, which is more sensible and less forgiving in a specific way. Under the old rule, an account above the threshold could day trade almost anything until the margin call arrived. Under a risk-based standard, the amount of intraday exposure you may carry is a function of the volatility and concentration of what you hold, so two accounts of the same size can have very different capacity, and a concentrated position in a volatile name will consume capacity that four diversified ones would not. What does not change is the underlying arithmetic of this lesson: the loan is fixed, the call arrives when your equity falls below a maintenance requirement, and leverage still raises the median and the probability of ruin at the same time. Two distinctions are worth holding alongside it. **Initial margin** is what you must have to open the position; **maintenance margin** is what you must keep to hold it, and it is the lower number that triggers the call. Standard margin accounts use strategy-based requirements set out in the rules; **portfolio margin** accounts, available above a higher equity threshold and subject to approval, use a risk model that can require substantially less for hedged positions and substantially more for concentrated ones. The 2026 change sits inside that framework rather than replacing it — it changes how the day-trading restriction is computed, not how margin calls work. Whatever the rule says, the number that matters is the one your broker will apply when the market moves, and that number is in the margin agreement you signed. • The pattern-day-trader minimum equity requirement is being replaced by an intraday margin standard, effective from June 2026 with a phase-in into 2027. • The new constraint turns on position risk rather than on a flat account-size threshold. • Initial margin opens a position; maintenance margin is the level that triggers the call. • Portfolio margin is a different, model-based regime available above a higher equity threshold — smaller for hedges, larger for concentration. A rule change that removes a threshold does not remove the arithmetic. An account newly able to trade intraday can still be wiped out by a move of the size this lesson computes, and the risk it can carry is now measured rather than capped — which places more of the discipline on you rather than less.

The call in practice: who liquidates what, and when

The arithmetic of the maintenance call is four lines long, and the sequence it sets off is where the real risk lives. When equity falls below the requirement, the broker issues a **maintenance call** asking for additional equity — but the phrase “asks” is doing a lot of work, because the margin agreement gives the broker the right to liquidate positions immediately and without prior notice, and most agreements say so explicitly. In practice brokers operate two thresholds: a soft one at which they notify and give you a few days, and a hard one, often a further few percentage points, at which the risk desk begins closing whatever it chooses. The choice of what to sell belongs to the broker, and its rule is not to sell what you would sell; it is to reduce the deficiency with the least operational fuss, which usually means the most liquid position, and often a proportional trim across the book rather than a single liquidation. Three features of that process are worth planning around. The first is **timing**: a call usually arrives when the market has already fallen, so the liquidation lands in the weakest part of the move and the price realized is the worst available rather than the average. Being forced to sell into a decline is the definition of a bad fill, and it is a feature of leverage rather than a broker’s shortcoming. The second is **concentration**: brokers apply higher maintenance requirements to concentrated and volatile positions — a single-name position may require forty or fifty percent of its value in equity rather than the twenty-five percent floor — so the same account can be comfortable with four diversified names and deficient with one, and the requirement changes without any change in the market. The third is the **interaction with stops**: a stop and a maintenance requirement are two exits with different owners, and the broker’s liquidation can arrive before your stop is reached, removing the position without triggering your order and leaving you with the loss and a decision you did not make. Underneath all of it sits the price of the loan, which is where the compounding enters. Margin interest accrues on the borrowed balance, which is charged at a spread over a benchmark rate and compounded as it accrues, and it is paid whether or not the position makes money. The arithmetic is not dramatic at a typical spread, but it is persistent: at a seven percent borrowing rate, a two-to-one position pays interest on half its market value, so the equity has to earn about three and a half percent a year before leverage contributes anything at all — and in a flat market that is the entire return. The honest statement of what leverage offers is therefore a comparison of three quantities: the asset return, the borrowing cost, and the probability of a forced sale. The first two decide the expected outcome; the third decides the distribution, and it is the reason the median outcome rises with leverage while the ruin probability rises faster. Sizing a leveraged book is not a question about the return; it is a question about what fraction of the account can be lost to a decline that is plausible in the next twelve months without a call arriving, which is why the earlier rule of dividing the holdable drawdown by the leverage is the right one. • The margin agreement allows liquidation without notice; the choice of what to sell belongs to the broker. • A soft notice threshold and a hard risk-desk threshold are different events. • Liquidation usually lands after the fall, so the fill is the worst price rather than the average. • Concentration raises the requirement — the same book can be fine with four names and deficient with one. • Margin interest is paid in all outcomes: at a 7% borrowing rate, leverage must earn ~3.5% before it contributes. The connection to the policy lesson is the one sentence worth writing into a plan: leverage changes who owns the exit. With a stop, you do; with a maintenance call, the broker does — and the broker’s exit happens at the moment the market is least willing to pay for your position.

Sources

• Regulation T, maintenance margin and the mechanics of a margin callFINRA margin requirements; broker disclosure documents
• Leverage and the amplification of drawdownsRalph Vince, "The Mathematics of Money Management"
• The 2021 GameStop collateral calls as a leverage eventSEC staff report on the January 2021 equity and options market events
• Why forced sales convert paper losses into permanent onesBrunnermeier & Pedersen, "Market Liquidity and Funding Liquidity" (2009)