SEPA Risk Management: How Minervini Sizes Positions and Controls Drawdowns

Every trader who has spent time with Mark Minervini’s work eventually fixates on the same thing: the Volatility Contraction Pattern entry, the tight pivot buy point, the explosive breakout. The VCP is visually compelling, mechanically precise, and — on the surface — teachable in an afternoon. But the entry is not what separates Minervini from the crowd. The risk management layer underneath it is. Without the rules governing how much to risk, when to build exposure, and when to stand aside entirely, the VCP is just another chart pattern with a 50% failure rate. With those rules, it becomes the foundation of a system that has compounded capital across multiple market cycles.

This article focuses on that invisible framework: the Specific Entry Point Analysis (SEPA) risk management system. We will cover the hard stop rule, the logic of progressive exposure, how Minervini reads the market environment before deploying capital, the expectancy math that makes even a moderate win rate highly profitable, and the arithmetic of drawdown recovery that makes the entire system coherent. We close with a full worked example using a $50,000 account. If you are looking for the VCP entry mechanics or the Trend Template, those are covered in the companion pieces on VCP pattern identification and the Trend Template and Stage Analysis.

SEPA as a Complete System, Not Just a Pattern

It is worth being precise about what SEPA actually is, because the label gets misused. Specific Entry Point Analysis is Minervini’s full methodology for identifying, entering, sizing, managing, and exiting high-momentum growth stock trades. The entry component — the pivot buy point off a volatility contraction — is just one layer. The other layers are the ones most retail traders skip because they are less photogenic than a clean breakout chart.

At its core, SEPA is a trend-following system with a fundamental filter layered on top. This matters for understanding the risk rules. Minervini is not a pure value investor hunting for cheap stocks, nor is he a pure technician ignoring earnings. He wants both: companies with accelerating earnings, revenue growth, and expanding margins, and stocks showing the right technical structure (Stage 2 uptrend, above rising 150- and 200-day moving averages, base formation after a prior advance). The combination dramatically raises the base rate of success — you are not trading random breakouts, you are trading breakouts from fundamentally superior businesses at the right point in their growth cycle. The risk rules assume this pre-filtering is already done. If you skip the fundamental screen, the stop rules become inadequate because you are trading lower-quality setups.

This also explains why comparisons to pure momentum traders like other trading legends can be misleading. Minervini’s win rate is meaningfully higher than pure price-following systems precisely because the fundamental filter reduces the number of “false” breakouts from stocks that happen to form a nice pattern but have no underlying business catalyst.

The Hard 7–8% Loss Limit: Why a Specific Number Matters

The most famous rule in Minervini’s system is also the most misunderstood. He sets a maximum loss of 7–8% on any individual trade — not 10%, not 12%, not “just below my stop.” The specificity is not arbitrary. It serves a purpose that purely mechanical traders often miss.

When the loss limit is vague — “I’ll cut if it falls more than I’m comfortable with” — the trader inevitably negotiates with themselves in real time. The stock falls 8%, and the story sounds plausible so the stop slides to 10%. At 10% the narrative is even more compelling because now there is a bigger loss to rationalize away. The position that should have been closed at 8% turns into a 20% loss while the trader waits for a bounce that may never come. A specific number eliminates this negotiation. When the stock hits the predetermined level, the analysis is already done. There is nothing left to decide.

The second reason for a specific number is portfolio-level math. If the maximum single-trade loss is capped at 7–8%, and position sizing is calibrated correctly, a string of ten consecutive losers produces a manageable drawdown. As we will show in the drawdown table below, a 20–25% portfolio drawdown requires a 25–33% gain to recover. A 50% drawdown requires 100%. Keeping each loss small keeps the hole you dig shallow enough that normal subsequent gains can fill it.

Stop Placement: Structural, Not Mathematical

Here is the distinction that separates Minervini from mechanical stop systems: the 7–8% is a maximum, not a target. The actual stop is placed just below the base’s structural low — the lowest point of the consolidation the stock was forming before the breakout — not at a fixed percentage below the entry price. If the base low happens to sit 5% below the ideal pivot buy point, the stop goes 5% below entry (below the base low), and the 7–8% rule simply confirms that this is an acceptable trade. If the base low sits 12% below the pivot, the setup does not qualify under SEPA because the required stop is too wide; Minervini would pass on that trade entirely and wait for a tighter formation.

This structural approach does two things simultaneously. It ensures the stop is placed where the trade thesis is genuinely invalidated — if the stock falls below the base low, the whole story of “controlled consolidation before the next advance” is broken — and it automatically filters out sloppy, wide-based patterns that are structurally weaker anyway. The 7–8% cap and the structural stop placement are mutually reinforcing: only tight, well-formed bases produce stops that satisfy both criteria.

In the complete position sizing guide, we cover the broader logic of stop-first sizing in detail. Minervini’s approach is a specific application of that general principle: the stop defines the risk, and the risk defines the size.

Progressive Exposure: Building Size Incrementally

Perhaps the least discussed element of SEPA — and possibly the most important for controlling drawdowns — is Minervini’s use of progressive position sizing. He does not deploy full size on every trade. He starts small, watches for confirmation, and only adds to a position when the market is proving him right.

The process typically works in stages:

• Pilot position: A fraction of the intended full size — often 25–50% — taken at the initial pivot breakout. This limits the damage if the breakout fails immediately (which roughly half of all breakouts do, at least on the first attempt).
• First add: If the stock holds above the breakout pivot, volume confirms, and the broader market shows favorable action, Minervini adds another tranche. This might come on the first or second day of strong follow-through, or at the next tight consolidation within the emerging move.
• Full size: The position reaches full size only when the trade is clearly working: the stock is moving in the right direction with the right characteristics (expanding volume on up days, contracting on pullbacks), and the overall market environment is cooperating.

The contrast with how most retail traders operate is stark. The typical pattern is: identify a setup, go full size on the breakout, watch it fail 50% of the time at full size, take a large loss, repeat. Progressive exposure inverts this. When you are wrong, you are wrong at a fraction of your intended size. When you are right, you build into the winning position. Over time this creates an asymmetry: losses are taken at reduced size and winners are held at full size.

This is closely related to the pyramiding principle that appears across nearly every great trader’s methodology — Jesse Livermore used it, Nicolas Darvas used it, William O’Neil built it into his CAN SLIM system. The direction of sizing is always the same: add to winners, not to losers. Minervini formalizes this instinct into a specific process tied to confirmation signals rather than arbitrary price targets.

The Market Environment Filter: When to Size and When to Step Back

The most dangerous period for a trend-following growth trader is a confirmed market downtrend. In a bear phase, breakouts fail at an abnormally high rate regardless of how perfect the individual setup looks. The stock might have a great VCP, strong earnings, and ideal volume characteristics — and still fail because the tide is going out. Minervini’s system explicitly accounts for this.

The logic flows from the statistical reality of market phases. In a confirmed uptrend, a well-selected SEPA setup might succeed 50–60% of the time (Minervini has cited historical win rates in this range in his public work). In a confirmed downtrend, that same setup might succeed 25–30% of the time. If your expectancy model is calibrated to a 50% win rate and you trade with full size in a downtrend where the effective rate is 25%, you will lose money even if your individual trade management is perfect.

Minervini distinguishes between market environments using several lenses simultaneously:

• Index trend: Are the major indices (S&P 500, Nasdaq) in a confirmed uptrend — making higher highs and higher lows above rising 50-day and 200-day moving averages?
• Distribution days: Is there abnormal institutional selling showing up as down days on heavy volume?
• New highs vs. new lows: Is the advance broad, with many stocks making new highs, or is it narrowing while lows accumulate?
• Follow-through day: Has there been a confirmed institutional-quality rally day (a major index up 1.5% or more on above-average volume) after a prior correction? This concept, developed by William O’Neil, is the signal Minervini uses to move from minimal exposure back to active sizing.

When the environment turns hostile, Minervini reduces exposure aggressively — moving to cash or near-cash — rather than stubbornly defending positions. Most retail traders treat every environment the same: they are fully invested in bull markets, corrections, and bear markets alike, and they absorb the full cycle drawdown. Professional traders treat cash as a position with a defined purpose. Sitting in cash during a hostile phase is not indecision; it is a conscious decision to protect capital until conditions favor deployment.

Batting Average, Win/Loss Ratio, and the Expectancy Framework

Understanding why a system is profitable requires more than knowing the win rate. A trader who wins 80% of the time but loses 5x their average winner on the 20% of losses is not making money. The metric that captures both dimensions simultaneously is expectancy — the average profit or loss per trade, expressed as a multiple of initial risk.

The expectancy formula is:

Expectancy = (Win Rate × Average Win) − (Loss Rate × Average Loss)

Minervini has discussed in his public work that his historical batting average is in the range of roughly 50%, meaning he wins approximately half his trades. His average winner is meaningfully larger than his average loser — reflecting both the asymmetric stop placement (tight structural stops, wide open upside targets) and his practice of letting winners run while cutting losers quickly. Even rough public estimates suggest his average win has historically been approximately 20–25% while his average loss was held closer to 7–8%.

The power of the expectancy framework is that it shows exactly how much the win/loss ratio can compensate for a lower batting average. Consider the table below:

Win Rate	Average Win (R)	Average Loss (R)	Expectancy per Trade	Result
50%	3.0R	1.0R	+1.00R	Highly profitable
40%	3.0R	1.0R	+0.80R	Profitable
33%	3.0R	1.0R	+0.32R	Marginally profitable
25%	3.0R	1.0R	−0.25R	Unprofitable
50%	2.0R	1.0R	+0.50R	Profitable
40%	2.0R	1.0R	+0.20R	Marginally profitable
50%	1.0R	1.0R	0.00R	Break-even (before costs)

The critical insight from this table: a 40% win rate is fully viable if average winners are three times average losers. The trader who insists on a 60% win rate but takes 1.5R wins and 1.5R losses is break-even before transaction costs. Minervini’s system achieves positive expectancy by combining a reasonable win rate with a highly asymmetric win/loss ratio — a product of the tight structural stop and the practice of letting winners run into multi-R moves.

The implication for your own trading: tracking only your win rate is actively misleading. You need both the batting average and the win/loss ratio to compute expectancy. A system with a 35% win rate and 4R average wins is more valuable than a system with a 65% win rate and 0.8R average wins. Minervini understands this arithmetic at a deep level, which is why he does not optimize for “being right” — he optimizes for the size of wins relative to the size of losses.

Pyramiding Into Strength: Adding to Winners

Minervini’s approach to adding to positions shares DNA with the methods described by Jesse Livermore in Reminiscences of a Stock Operator and later refined by William O’Neil. The principle is always the same: confirmation of a thesis warrants more capital, not less. If the stock breaks out from a VCP on heavy volume, holds above the pivot for several days, then forms another tight consolidation within the emerging uptrend — that is the market saying “the thesis is valid.” The appropriate response is to increase size, not to take partial profits and reduce.

The mechanics of pyramiding correctly matter more than most traders appreciate:

• Add on strength, not weakness: The add-on price must be higher than the prior entry, and it must come from a position of strength (new tight consolidation or a brief pullback that holds above key support), not a panic-buy into a parabolic run.
• Each add is smaller than the prior tranche: The first position is the largest because it has the lowest average cost and the widest stop relative to the current price. Adds are progressively smaller to avoid inverting the cost basis dangerously high. A typical progression might be 50% initial, 30% first add, 20% final add — totaling 100% of intended full size.
• The stop migrates with the position: As you add, the stop for the entire position moves up to protect the profitability already accumulated. You should never be in a situation where a reasonable normal correction would turn a profitable multi-tranche position into a net loss.
• Never add to a losing position: This rule is absolute in Minervini’s framework. Adding to a loser is averaging down, which compounds exposure precisely when the thesis is being contradicted by price action.

The Mathematics of Drawdown and Recovery

The deepest reason the 7–8% hard stop exists is mathematical. Drawdown is not symmetric: the percentage gain required to recover a loss is always larger than the loss itself, and this asymmetry grows non-linearly as drawdowns deepen. A 10% drawdown requires an 11.1% gain to return to the prior equity high. That sounds manageable. But a 50% drawdown requires a 100% gain — doubling the remaining capital — just to get back to even. The table below makes this viscerally clear:

Portfolio Drawdown	Remaining Capital	Gain Required to Recover	Approximate Recovery Time (at 20%/yr)
5%	$95,000	5.3%	< 4 months
10%	$90,000	11.1%	∼ 7 months
15%	$85,000	17.6%	∼ 11 months
20%	$80,000	25.0%	∼ 15 months
25%	$75,000	33.3%	∼ 20 months
33%	$67,000	49.3%	∼ 30 months
50%	$50,000	100.0%	∼ 4+ years
75%	$25,000	300.0%	Potentially never

The recovery time column assumes a 20% annualized return — an aggressive but achievable target for a skilled active trader. Even at that pace, a 25% drawdown costs nearly two years of compounding just to get back to flat. A 50% drawdown effectively puts a trader behind by an entire market cycle. This is the mathematical case for never letting a single trade turn into a large loss: the compounding of time lost to recovery is often more damaging than the dollar loss itself.

For a deeper analysis of drawdown mechanics and the recovery curve, the free drawdown calculator lets you model any drawdown scenario against your own growth rate assumptions.

Worked Example: Sizing a SEPA Trade on a $50,000 Account

Theory crystallizes with numbers. Let’s walk through a complete position sizing example using Minervini’s framework applied to a $50,000 account.

Setup parameters:

• Account size: $50,000
• Risk per trade: 1% of account = $500
• Stock: trading at $82.00 at the pivot buy point
• Base low (structural stop level): $75.90 — the lowest point of the VCP consolidation
• Stop placement: just below the base low, at $75.50 (leaving a small buffer below the structural low)
• Stop distance: $82.00 − $75.50 = $6.50 per share
• Loss as percentage of entry: $6.50 / $82.00 = 7.9% — within the 7–8% maximum

Full position size calculation:

Shares = Risk Budget ÷ Stop Distance = $500 ÷ $6.50 = 76 shares (round down)

Total position value: 76 × $82.00 = $6,232 — representing 12.5% of the account in this single position. This is the notional exposure; the actual risk is $500 (1% of account), because that is the maximum loss if the stop fires.

Progressive exposure version (Minervini’s preferred approach):

Tranche	Trigger	Shares	Entry Price	Cost	Cumulative Shares	Avg Cost
Pilot (50%)	Pivot breakout on volume	38	$82.00	$3,116	38	$82.00
First add (30%)	Stock holds, clears $85.00 on volume	23	$85.20	$1,960	61	$83.24
Final add (20%)	New tight base clears $88.50	15	$88.70	$1,331	76	$84.59

At full size (76 shares, average cost $84.59), the stop at $75.50 represents a potential loss of $692 on the full position — slightly above the original $500 budget because the later tranches were bought higher. In practice, Minervini would have raised the stop as the position moved in his favor, so by the time the final tranche is added, the stop on the original 38 shares might already be at breakeven or better, limiting the total portfolio risk to a manageable figure even at full size.

Contrast this with the naive approach: buying all 76 shares at the initial breakout, being wrong on 50% of breakouts, and taking $500 losses at full size every time the trade fails immediately. Both methods have the same worst-case loss when the stock fails straight from the entry. But the progressive approach has a far lower average loss — because on the trades that fail in the first day or two (before any adds are made), the loss is on 38 shares, not 76.

You can replicate this exact arithmetic for any account size, risk percentage, and stop distance using the free risk calculator, which handles the shares-from-budget calculation automatically.

Why Minervini Is a Trend-Follower with a Fundamental Filter

A common misclassification is to call Minervini a “fundamental trader” because he screens for earnings growth, or a “technical trader” because he uses chart patterns. The more precise label is trend-follower with a fundamental filter — and this distinction matters for understanding the risk framework.

Pure value investors buy stocks that are cheap relative to intrinsic value; they are indifferent to price trend and often deliberately buy during downtrends (“it’s getting cheaper”). Pure technical traders may buy breakouts without regard for whether the underlying company has any earnings at all. Minervini does neither. He only buys stocks that are already in a confirmed Stage 2 uptrend — the trend must be established before he engages — and he requires the fundamental story to be accelerating, not just improving. This is trend-following logic (price must confirm direction) with a fundamental catalyst filter (the business must be genuinely exceptional).

The risk management consequences of this classification are significant. Because he follows the trend, Minervini is always long in the direction of the market’s dominant force. He does not fight the tape. When the tape turns against him, the hard stop and the market environment rules force him out of harm’s way quickly. He does not hold a position through a 30% drawdown because “the fundamentals are still intact” — that is value investor logic applied to a trend-following framework, a category error that destroys accounts. The 7–8% stop rule is not just a risk management rule; it is a statement about the information content of price. If price falls 8% below a base low, the trend has spoken, and the fundamental thesis no longer overrides it.

Putting It Together: The Risk Hierarchy in SEPA

The SEPA risk management system works because each layer supports the others. The fundamental filter raises the base rate of success, improving the win rate. The structural stop placement aligns the stop with the actual invalidation point rather than an arbitrary percentage. The 7–8% hard maximum enforces discipline before rationalization can set in. Progressive exposure means the largest size is deployed only when the trade has already proven itself. The market environment filter reduces exposure during phases when even valid setups fail at high rates. And the expectancy framework ties it all together — even a 40% win rate becomes a profitable long-run strategy when winners are systematically allowed to be 2–3 times the size of losers.

None of these rules is individually novel. The fixed-stop discipline appears in O’Neil’s CAN SLIM system. Progressive sizing appears in Livermore’s pyramid methodology. The market environment filter appears in Stan Weinstein’s Stage Analysis. What Minervini has done is combine them into an internally consistent system where each rule reinforces the others, so the failure of any single rule tends to be caught by the ones surrounding it.

The critical habit for applying this system is pre-trade planning. Before the market opens, the professional knows: what is the pivot level, where is the structural stop, what is the pilot size, at what price and conditions will adds be made, and what would constitute a market environment that argues for no position at all. None of these decisions should be made while the market is open and prices are moving. The rules exist precisely because real-time decision-making under uncertainty tends toward loss avoidance (holding losers) and gain-seeking (cutting winners) — the exact opposite of what produces positive expectancy.