Larry Williams’ Volatility Breakout System and Aggressive Money Management

In 1987, a futures trader named Larry Williams entered the Robbins World Cup Championship of Futures Trading with a real, audited account of roughly $10,000. Twelve months later that account was worth more than $1.1 million — a return of over 11,000% in a single year. It remains one of the most famous performance records in trading history, and it cemented Williams as a household name among short-term traders. But the lesson most people take from that record is exactly the wrong one.

The common reading is that Williams found a magical entry. He did have an entry — the volatility breakout, a deceptively simple idea he popularized in Long-Term Secrets to Short-Term Trading. But the entry alone would never have produced four-digit-percentage returns. What turned a modest, repeatable edge into a seven-figure account was aggressive money management: position sizing derived from the Kelly criterion and Ralph Vince’s fixed-fractional framework, applied at a level most professionals would consider reckless. Williams himself has said so plainly, and he has spent decades warning traders that the same sizing that made him famous also produced gut-wrenching drawdowns and could have wiped him out. This article unpacks both halves of the story — the entry and the sizing — and is honest about where the danger lives.

Part One: The Volatility Breakout Entry

Markets spend most of their time inside a range of recent volatility. Price oscillates, fills orders, builds positions, and gives no clear signal about direction. Williams’ insight was that when price thrusts beyond a meaningful fraction of recent volatility, that thrust itself is information. It signals that the “line of least resistance” — a phrase borrowed from Jesse Livermore — has broken in one direction, and that momentum is likely to continue at least for the rest of the session.

The classic formulation is built off two simple reference points: today’s open and yesterday’s range. The entry trigger for a long is:

• Buy stop = today’s open + (k × yesterday’s range)
• where yesterday’s range = yesterday’s high − yesterday’s low
• and k is a fraction, commonly between roughly 0.5 and 0.8

The mirror image applies for shorts: Sell stop = today’s open − (k × yesterday’s range). You place a resting stop order at that calculated level. If price never reaches it, you do nothing. If it trades through, you are filled and you are in the move — not predicting it, but reacting to a confirmed thrust.

Why the open matters

Many breakout systems anchor off the prior close. Williams emphasized the open as the reference because the open already incorporates overnight news, gaps, and the market’s fresh consensus at the start of the session. Measuring the breakout from the open rather than yesterday’s close means the trigger adapts to gaps: if the market gaps up hard, your buy level is already higher, and you are not chasing a move that has effectively already happened before the session began. It keeps the entry honest relative to where trading is actually starting.

Why a fraction of range, not a fixed number of points

Using k × yesterday’s range rather than a fixed dollar amount makes the system self-scaling across instruments and across volatility regimes. In a quiet market yesterday’s range is small, so the breakout threshold sits close to the open and you get triggered on modest thrusts. In a violent market the range is large, the threshold is far from the open, and you demand a much bigger move before committing. The same code works on a sleepy bond contract and a wild crypto pair because the volatility input is built into the formula. This is the same family of logic behind the opening-range and ATR-based approaches many traders use today.

The role of k

The choice of k is a trade-off, not a magic constant. A small k (say 0.3) triggers easily: you catch more moves but eat more false breakouts that immediately reverse. A large k (0.8 or above) demands a powerful thrust before entry: fewer trades, fewer whipsaws, but you give up part of the move and miss some genuine runs entirely. Williams’ published work generally lands in the 0.5–0.8 zone as a sensible middle ground, but the honest answer is that the optimal k drifts with the market and the timeframe. Beware of curve-fitting it to historical data; a k that looks perfect in a backtest is often perfect only for that sample.

The Supporting Filters: TDW, Seasonality, and Cycles

Williams rarely traded a raw breakout in isolation. A large part of his research was about when the breakout was worth taking. Three filter families recur throughout his writing.

Trade Day of Week (TDW). Williams documented that, historically, certain weekdays had a statistically different tendency in certain markets — for example, taking long breakouts only on specific days of the week improved results in his testing. The mechanism is partly structural: weekly cash flows, options expiration patterns, and institutional rebalancing are not evenly distributed across the week. TDW is a filter that says “only act on the breakout signal on days when the odds have historically been with you.”

Seasonality. Commodities in particular have recurring annual patterns driven by harvest cycles, heating-oil demand in winter, planting and weather windows, and tax-year flows. A breakout that fires with a strong seasonal tailwind is a different proposition from one that fires against it. Williams used seasonal studies to bias which direction he was willing to take signals in a given window.

Cycles and market structure. He also layered in cycle work and simple structural conditions — for instance, only taking a long breakout when the recent swing structure was supportive. None of these filters create the edge by themselves; they concentrate the breakout into the conditions where it has historically paid, raising the win rate or the average winner enough to matter.

The danger with filters is the same as with k: each added condition that improves a backtest also adds a degree of freedom, and degrees of freedom are how you fool yourself. A robust filter has an economic story behind it (why should Tuesdays differ?). A fragile one is just a pattern the optimizer found in noise.

A Worked Breakout Example

Suppose you are trading a futures contract and yesterday it traded a high of 4,180 and a low of 4,120. Yesterday’s range was therefore 60 points. You choose k = 0.6. This morning the contract opens at 4,150.

• Buy trigger = 4,150 + (0.6 × 60) = 4,150 + 36 = 4,186
• Sell trigger = 4,150 − (0.6 × 60) = 4,150 − 36 = 4,114

You place a resting buy stop at 4,186 and a resting sell stop at 4,114. Mid-morning, a burst of buying carries price up through 4,186 and you are filled long. Now the critical question is not “where might this go” — it is where are you wrong, and how much do you risk if you are. A common Williams-style stop is a fixed dollar amount per contract, or a move back below the open or below a fraction of the range. Say you set your stop at 4,168, an 18-point adverse move. That 18 points is your risk per contract, and it is the single most important number in everything that follows. The entry got you into the move. The stop, combined with sizing, decides whether the trade can hurt you.

Part Two: The Money Management That Actually Won the World Cup

Here is the part the headlines leave out. A volatility breakout with sensible filters might win, say, 50–60% of its trades with a reward-to-risk ratio somewhere above 1. That is a respectable edge. It is not an 11,000%-a-year edge. No realistic entry system produces four-figure annual returns on its own. The returns came from betting an enormous fraction of the account on every trade and letting the math of compounding and reinvestment do the rest.

Williams sized positions using the logic of the Kelly criterion and Ralph Vince’s fixed-fractional and “optimal f” frameworks. The core idea of fixed-fractional sizing is simple and powerful: you risk a constant percentage of your current equity on each trade, not a constant number of contracts. As the account grows, the absolute size of each bet grows with it; as it shrinks, the bets shrink too. This is the engine of geometric (compounding) growth — and, when the fraction is large, the engine of catastrophic drawdowns.

How fixed-fractional sizing works

The mechanics are straightforward:

• Decide the fraction f of equity you will risk per trade (e.g., 2%).
• Risk dollars = f × current equity.
• Position size = risk dollars ÷ (risk per unit, i.e. distance from entry to stop).

On a $10,000 account risking 2% with our 18-point stop (worth, say, $18 per contract at $1/point), risk dollars = $200, so size = 200 ÷ 18 ≈ 11 contracts. After a winning streak takes the account to $20,000, the same 2% rule now risks $400 and buys ~22 contracts automatically. You never re-decide; the rule scales the bet for you. This is exactly the calculation our position size and risk calculator performs, and it is the disciplined, survivable end of the spectrum.

Where Kelly comes in — and where it bites

The Kelly criterion answers a precise mathematical question: what fraction of capital should you bet to maximize the long-run growth rate of your bankroll, given your edge? For a known win rate and payoff, full Kelly produces the fastest possible compounding. The trouble is that full Kelly is brutally volatile. It is mathematically optimal for growth but routinely produces drawdowns of 50% or more, and it assumes you know your true win rate and payoff exactly — which in trading you never do. We cover the full derivation and its pitfalls in our guides on the Kelly criterion and probability and on expectancy and Kelly sizing.

What Williams effectively did in 1987 was bet at or beyond a large fraction of Kelly. When a high-edge, high-payoff system meets a near-Kelly bet size and a string of winners arrives in sequence, the equity curve goes vertical. That is the seven-figure result. But run the same sizing into a normal losing cluster — which every system has — and the account can give back 70%, 80%, or more. Williams has been candid that his real account suffered enormous swings on the way to that championship, and that he would not recommend most people trade that way. The record is real; the path was terrifying.

The Edge Came From the System; the Returns Came From the Sizing

This is the single most important sentence in the whole story, so it is worth stating cleanly: the entry and filters gave Williams a positive expectancy; the aggressive fixed-fractional sizing converted that expectancy into both the spectacular returns and the spectacular risk. Sizing is a multiplier. It cannot create an edge — bet aggressively on a negative-expectancy system and you simply go broke faster. But applied to a real edge, the size of the fraction dials the outcome anywhere from “steady professional” to “lottery ticket.”

This also explains why so many traders failed to reproduce Williams’ results even when they had his exact rules. They copied the entry and used prudent sizing — and got prudent, unremarkable returns. Or they copied the aggressive sizing onto a weaker edge and blew up. The famous number lives at the intersection of a genuine edge and a near-suicidal bet size, plus a year in which the sequence of trades happened to break favorably. Change the sequence, keep everything else, and the same strategy could have ended the year down 90%.

Risk of Ruin: Why the Fraction Is Everything

Risk of ruin is the probability that a string of losses drives your account below a threshold from which you cannot recover — practically, to zero or to a level where you stop trading. It depends on three things: your win rate, your reward-to-risk ratio, and crucially the fraction you bet. Hold the edge constant and raise the fraction, and risk of ruin climbs non-linearly. Past a certain fraction — even with a real, positive edge — ruin becomes a near-certainty over enough trades, because a long enough losing streak will eventually occur, and a large fraction turns that streak into a crater you cannot climb out of.

The cruel asymmetry of percentage losses makes this worse. A 50% drawdown requires a 100% gain just to get back to even. An 80% drawdown requires a 400% gain. Aggressive sizing produces exactly these deep holes, and the deeper the hole, the more the recovery math works against you. This is why prudent traders cap per-trade risk at a small fraction even when their edge would “justify” more — survival has to come before optimization. Our position sizing and risk management guide walks through the drawdown-recovery table in detail.

Aggressive vs Prudent: A Side-by-Side

The table below illustrates the trade-off with a stylized, repeated-bet model. Assume the same positive-edge system applied to a $10,000 starting account, and compare how different per-trade fractions tend to behave. These figures are illustrative of the shape of the outcomes, not a forecast — the entire point is that high-fraction paths have enormous dispersion.

Per-trade risk fraction	Growth potential	Typical max drawdown	Risk of ruin	Who it suits
0.5–1% (conservative)	Slow, steady compounding	~10–20%	Negligible	Most traders, any account size
2% (the classic standard)	Solid long-run growth	~20–35%	Very low with a real edge	Disciplined retail and pros
5–10% (aggressive)	Fast growth in good runs	~40–60%	Meaningful; recovery painful	Experienced, high-conviction only
20%+ (near-Kelly, “World Cup”)	Explosive or catastrophic	70–90%+	High over many trades	Almost no one — danger zone

Notice the pattern: moving from 1% to 2% roughly doubles your bet and modestly raises drawdown for a real boost in growth — a good trade. Moving from 2% to 20% does not give you ten times the sane growth; it gives you a small chance of a fortune and a large chance of devastation. The relationship between fraction and outcome is wildly non-linear at the top end. That non-linearity is precisely why the 1987 record is so hard to repeat and so dangerous to chase.

Sizing the Worked Trade Properly

Return to our breakout: filled long at 4,186, stop at 4,168, so 18 points of risk per contract. Let us size it three ways on a $25,000 account, with each point worth $1 per contract (so $18 risk per contract).

Approach	Risk fraction	Risk dollars	Contracts (risk ÷ $18)	What happens on a 5-loss streak
Prudent	1%	$250	~13	Down ~5%, easily recovered
Standard	2%	$500	~27	Down ~10%, recoverable
Aggressive	15%	$3,750	~208	Down ~55%+, needs +120% to recover

Same entry, same stop, same edge — three completely different businesses. The prudent and standard rows are sustainable. The aggressive row is how you both win a championship and how you join the long list of traders who detonated trying to. Decide your fraction before the trade, write it down, and let a calculator — not adrenaline — convert it into a contract count.

What to Actually Take From Larry Williams

There is a version of this story that is genuinely useful and a version that is destructive, and they look almost identical from the outside. The useful version:

• Build a real, simple, robust edge. The volatility breakout — open plus a fraction of yesterday’s range — is elegant precisely because it has few moving parts and a clear logic. Resist over-fitting k and filters.
• Anchor entries to volatility, not fixed points, so the system adapts across markets and regimes.
• Understand that sizing is a separate, deliberate decision from entry, and that it is the lever that determines your real-world results.
• Choose your fraction for survival first. A 1–2% fixed-fractional rule captures most of the compounding benefit with a fraction of the risk-of-ruin. Reserve anything larger for the rare cases where you genuinely understand the math and can stomach the drawdown.
• Treat the 1987 record as a warning as much as an inspiration. Williams himself frames it that way. The edge is reproducible; the bet size that produced the headline number is not something a sane risk manager would run with money they need.

The deepest lesson is the cleanest: an edge keeps you in the game, and money management decides what that game is worth — and whether you survive to keep playing it. Get the entry right and you have a tool. Get the sizing right and you have a career. Get the sizing wrong, even with a great entry, and the only question is when the account hits zero.