Peter Brandt: Classic Charting Breakouts and the "Risk-First" Discipline
Peter Brandt · Mid-1970s to present. A commodities and currency futures trader of more than 40 years, a Market Wizard, and founder of Factor LLC. His public chart analysis of crypto, including Bitcoin, runs from around 2015 to the present.
Published 2026.07.08
Peter Brandt is a classic chartist who has traded commodity and currency futures for more than 40 years. His method isn't about flashy indicators — it's about narrowing "which patterns you'll even look at" to an extreme, and designing the stop and position size before the entry. This article deconstructs the <b>structure</b> of the method he has left in his books and public record, for educational purposes — it is not individual entry instructions or a guarantee of profit.
- Narrow the candidates themselves to an extreme — only 8–16-week consolidation patterns with horizontal boundaries (rectangles, right-angled triangles, head-and-shoulders). He publicly excludes symmetrical triangles and standalone trendlines.
- The trigger isn't a prediction but a lagging observation — a breakout on the "daily close." Intraday spikes are treated as traps, a market that has already run away isn't chased, and a retest reinforces confidence.
- Risk-first — set the stop location first, then back-calculate the position within 0.6–1% of capital from that distance. Results come not from win rate but from a reward-to-risk of 3–4R or more and from expected value.
- The fit with crypto is high, but because of liquidation, funding, and stop-hunts, the time references and the reward-to-risk threshold have to be recalibrated more conservatively.
① Which Patterns Even Become "Candidates" — Screening
Brandt is a classic chartist in the Edwards & Magee lineage. He narrows his candidates to patterns with horizontal boundaries only — rectangles (range consolidation), right-angled triangles (ascending/descending), and head-and-shoulders. He publicly says he "dislikes" symmetrical triangles and standalone trendlines, and excludes them, because their reference lines are subjective and make the stop location ambiguous.
The shape conditions are demanding too. He prefers patterns tightly compressed into 8–16 weeks (26 weeks or less in principle) and favors those consolidated into a narrow price range (roughly 15% or less). He reckons "there are only 10–20 truly strong patterns a year" — meaning that most of the time there are no candidates, and standing aside is the default.
A horizontal line gives a clear invalidation price. Because the boundary is fixed at a specific price, "where to consider yourself wrong" is set as a number before entry. With a symmetrical triangle or a trendline, the line shifts depending on slope and starting point, so the stop location wavers from person to person — Brandt excludes them not for their predictive power but for stop clarity.
② The Entry Frame — Observational and Lagging
The key is that the entry trigger is not a "prediction" but a lagging observation. For Brandt, the signal is "has the daily close broken through the horizontal boundary and completed the pattern?" A spike that briefly pierces the boundary intraday is treated as a trap, and no move is made before the close is confirmed. The breakout itself isn't read as a buy instruction but handled as a structural signal that confirms the pattern's completion or invalidation.
- Find horizontal-boundary consolidation-pattern candidates on the weekly chart.
- Mark the boundary line and the invalidation price (the stop location) in advance.
- Wait to see whether the daily "close" breaks through the boundary and completes the pattern — ignore intraday spikes.
- After the breakout, observe whether the pullback to the boundary (the retest) holds, to reinforce confidence.
- Don't "chase" a market that has already run far with a market order.
There's always a next setup, and a better one comes along — the principle he repeats when he refuses a chasing (FOMO) entry and justifies standing aside.
③ Risk and Sizing — Risk-First
The center of gravity of Brandt's method is the stop, not the entry. Before entering, he sets the stop location first (the pattern's lower edge, the breakout bar's low, or the retest low) and decides position size by back-calculating from that distance. The stop is a value built into the pattern's structure, not a discretion to be negotiated after the fact — if the pattern is broken, that trade is automatically void.
He exposes only about 0.6–1% of account capital to risk per instrument. He sets the target with an Edwards & Magee "measured move," but only accepts a candidate when the expected range relative to the stop distance is an asymmetric reward-to-risk of 3–4x or more. When price reaches about 70% of the target, he shifts his stance to protecting the gain.
Brandt's public principle: "A trader's only asset is trading capital, and without capital the business is over." He keeps the hit rate on individual trades below half and produces results through a reward-to-risk and expected-value structure in which a few big wins cover many small losses. In other words, what he designs for is not "how often you're right" but "how much you make when right and how much you lose when wrong."
④ Invalidation and Limits
Invalidation is mechanical. He exits without attaching a reason if the pattern boundary is reversed, the daily close breaks the pre-set stop line, or the relationship between price and the 8- and 18-day moving averages inverts (trend damage). He clears losing positions within a few days to a week and makes it a rule not to carry a losing position over the weekend on a Friday.
A chart-pattern breakout is inherently a lagging indicator and vulnerable to false breakouts. Brandt himself has publicly said the hit rate on individual trades is low (most end in a loss), and because results concentrate in a few big wins, the whole method collapses if you slip from the discipline — it is hard to reproduce. He posts long-term directional calls on X (social media) but is known to correct himself when wrong, so such targets should be read only as "observation hypotheses," not as grounds to buy. He has gone through slumps himself — enough to publicly look back on a "hard trading year" and its drawdowns.
What is organized here is only the structure of a method one person made public — not a guarantee that copying it produces profit. It must not be read as a buy instruction for any specific stock or coin, or as a certainty about future prices.
⑤ When Porting to Crypto (Perpetual Futures)
Brandt himself has publicly analyzed Bitcoin as "one of the markets where classic charting works best," so the fit with crypto is high. But because the method was established to suit the rhythm of commodity and currency "spot/futures," it must not be ported as-is to perpetual futures and high leverage, where liquidation exists.
- Since the market runs 24 hours with no weekend, redefine time references like the "daily close" and "Friday liquidation" around the exchange's UTC close or your own trading rhythm.
- On high leverage, where liquidation exists, back-calculate from notional exposure and the liquidation price, and lower the leverage so the liquidation price does not fall within the stop distance.
- Because false breakouts, stop-hunts, and funding costs are larger than in stocks/futures, raise the retest confirmation and the reward-to-risk threshold more conservatively.
On high leverage, a 0.6–1%-of-capital risk can become several times that in effect through leverage × volatility. Forced liquidation turns averaging down into a total loss, and stop-hunts deliberately sweep the liquidity just below the boundary — a stop placed right at the boundary is easily taken out before the close is confirmed. Calculate R only from the "entry-to-stop" distance, never from leverage.
A hypothetical example (educational; not a guarantee of profit). With account capital of $100,000 and a per-trade risk ceiling of 1% of capital, 1R = $1,000. Suppose a rectangle pattern's daily-close breakout completes at 105 and the pre-set stop is 100 — the stop distance is about 4.8%. Position notional back-calculates to $1,000 ÷ 4.8% ≈ $20,800 (you do not inflate it with leverage). If the Edwards & Magee measured-move target is about 124, the expected range is (124−105) ÷ (105−100) ≈ 3.8R. That is, an asymmetric structure that risks 1 to aim for 3–4. Repeat only these 3–4R setups, and even if the individual hit rate stays below half, a few big wins can offset many −1R losses — that is the reward-to-risk and expected-value frame. This is not a specific entry instruction or a certainty about future prices.
- Does this pattern have a "horizontal boundary"? Or is it a symmetrical triangle or a standalone trendline (if the latter, it's excluded from candidates by Brandt's criteria)?
- Did I confirm the entry trigger with a "daily close" breakout rather than an intraday spike?
- Did I set the stop location first, before entering, and back-calculate position size to within 1% of capital from that distance?
- Is this a spot where the reward-to-risk opens up to at least 3–4R, or a FOMO entry chasing a market that has already run away?
- Am I ready to exit without attaching a reason if the pattern boundary is reversed?
FAQ
If I copy Peter Brandt's method exactly, will it make me money?
No. This article deconstructs the "structure" of the method he left in his books and public record for educational purposes; it is not a guarantee of profit. Brandt himself has publicly said the hit rate on individual trades is low and that results concentrate in a few big wins, and the whole method collapses if you slip from the discipline. You should read it on the premise that it is hard to reproduce.
Why doesn't he use symmetrical triangles or trendlines?
Because they make the stop location ambiguous. A horizontal boundary fixes the invalidation price at a specific number, but with symmetrical triangles and trendlines the line shifts depending on slope and starting point, so the stop wavers from person to person. Brandt picks only horizontal patterns not for their predictive power but to "nail down the risk clearly before entry."
Can I treat his X (Twitter) targets as buy signals?
No. Brandt publishes long-term directional calls but is known to correct himself when wrong. Such targets are only "observation hypotheses," not entry instructions or certainties about future prices. This entire document is an educational explanation of structure, not a buy call for any specific stock or coin.
Can I use it as-is on crypto perpetual futures?
The fit is high — Brandt has analyzed Bitcoin as a market where classic charting works well — but porting it as-is is dangerous. Because it's a 24-hour market, you have to redefine the "daily close and Friday liquidation" time references, lower leverage by back-calculating from the liquidation price, and raise the reward-to-risk threshold more conservatively because of stop-hunts and funding costs.