How to Use Leverage Safely
To use leverage safely, size every position from the stop distance and your risk figure first, then treat the leverage your account permits as a boundary you never approach. Leverage multiplies the exposure without improving the method it is applied to.
Leverage lets a position be larger than the cash behind it. It does not change the method, the edge or the odds — it scales whatever those already are, and its effect on drawdown is harsher than the multiple implies.
Before you start
A distinction between the leverage available and the leverage used, because they are different numbers. An account offering thirty times is not an instruction to use thirty times.
The arithmetic of leverage on drawdown, not the intuitive one. The worked figures below are not what doubling suggests.
A position size derived from the stop distance rather than from what the account permits. Do that first and the leverage question mostly disappears.
The steps
1. Size from the stop, always
Your risk amount divided by the distance to invalidation gives the position. Whether that requires leverage is a consequence of the calculation rather than an input to it.
2. Work through the arithmetic
Borrow as much again as you own and you hold twice the index: a 50% fall erases the account. At 3x, a 33% fall does it. The S&P 500 fell 56.8% from October 2007 to March 2009, so a 2x position held through it would have been gone before the bottom. On the way up the multiple falls short: held at a constant multiple, an index growing 10% a year at 20% volatility gives roughly 16% at 2x and 18% at 3x, not 20% and 30%.
3. Compute the leverage you are actually using
Total position value divided by account equity. Most people have never calculated it and are surprised by the answer, particularly when several positions are open at once.
4. Add your positions together
Three correlated positions at modest leverage each is one large position. The exposure that matters is the combined one, and it is the one nobody computes.
5. Treat the available limit as a boundary you never approach
The broker’s maximum exists for their risk management, not yours. Trading anywhere near it means a normal adverse move produces an abnormal loss.
6. Account for the cost of carrying it
Margin interest accrues daily on borrowed funds. A leveraged position held for months pays that continuously, and it comes out of whatever the position eventually makes.
7. Reduce it during a losing stretch
Sizing from a percentage of the current balance does this automatically. The S&P 500 once waited 7.5 years, from 1973 to 1980, to regain a record, and compounding a fixed percentage of a falling balance is what makes a stretch like that survivable.
How to tell it worked
Position size came from the stop distance in every case, with 0 exceptions.
Total leverage across all open positions was computed, not just per trade.
Usage stayed well below the account’s maximum, checked at least 1 time per week.
And the risk figure recalculates from the current balance, so it shrinks in a drawdown.
Why the drawdown grows faster than the return
Because losses compound against a smaller base. A 10% loss requires an 11.1% gain to recover; a 20% loss requires 25%. Leverage moves you up that curve, and the curve is not linear.
The standard approximation makes it concrete: a daily-reset position grows at about L × r − (L² − L) × σ² / 2, so the shortfall rises with the square of the multiple and with volatility. The shortfall is not a market opinion; it is arithmetic that applies to any leveraged exposure.
When leverage is genuinely reasonable
When the instrument requires it structurally. Futures and currency positions are leveraged by construction, and using them at a sensible size is not the same as seeking leverage.
When the position is small and the stop is tight. A correctly sized position on a tight stop can require leverage arithmetically while risking a modest amount — that is the calculation working.
And never as a way to make a small account grow faster. That is the use case leverage is marketed for and the one where the drawdown arithmetic does the most damage, because a small account has the least room to absorb it.
The one calculation that settles it
Total position value divided by account equity, across everything open at once. One division, done weekly, and it produces a number most people have never seen for their own account.
Under 2 is conservative for most methods. Between 2 and 5 is where a bad week starts to matter. Above that, an ordinary adverse move produces a loss the account cannot easily absorb.
Correlated positions have to be added together. Four currency pairs against the same currency, or five technology shares, are one exposure wearing several names — and the combined figure is the one that behaves like leverage when the market moves.
Do it on a Sunday, when nothing is open and nothing needs deciding. A number computed calmly once a week is worth more than a rule you intend to follow in the moment.
The original data
Of the 24,971 unique videos in the search study behind this site, 55 mention leverage in the title, at a median of 36,446 views across 50 channels, and 58% of those titles are instruction-shaped. Margin appears in 34 at 28,191 and risk management in 410 at 4,079. A video that turned up in more than one search is counted once.
55 videos at 36,446 against 410 on risk management at 4,079. An eighth of the coverage and nearly nine times the audience per video — leverage attracts far more attention than the discipline that makes it survivable.
The question: The account allows 20x. The setup is excellent. Size up?
The answer is that a good setup does not change the arithmetic. The drawdown at higher leverage exceeds the multiple regardless of how the entry looks — and the setups that look most compelling are the ones where sizing up feels most justified.
When it fails
The failure is leverage used to make a small account meaningful, and the arithmetic is against it from the start. The reasoning is understandable: a modest account produces modest amounts, and leverage appears to fix that. What it actually does is move the account up a non-linear drawdown curve, so an ordinary losing stretch — the kind that happens several times a year — produces a hole that requires a disproportionately larger gain to fill. The account was never too small; it was being asked to produce returns that its size does not support.
The second failure is sizing from the available limit. That is the broker’s number.
A third is ignoring combined exposure. Correlated positions are one position.
A fourth is a fixed risk figure. It should shrink with the balance.
A fifth is ignoring the carry cost. It accrues daily.
And a sixth is expecting the multiple. Held at a constant multiple, returns fall short of it.
Related
Leverage trading covers the mechanism. Margin account is where it is administered. And position sizing is the calculation that makes it a non-issue.
Let the arithmetic decide this rather than an argument. Doubling a daily-reset exposure does not double the return over time, because choppy days take a slice out of it, and it does double every loss along the way. Less than the upside, all of the downside.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.