Section 8 - Position Sizing for Optimal Results

{fullwidth

Think and Trade Like a Champion — Mark Minervini

Macro Overview & Strategic Value

Section 8 converts the expectancy and risk-first principles from earlier sections into a concrete portfolio-construction system, answering Ed Seykota’s question — “once you know your expectancy, how much should you invest?” — with a mathematically bounded framework. Minervini’s core thesis is that position size sits on a spectrum between two ruinous extremes: over-concentration (risking catastrophic single-stock exposure) and over-diversification (“di-worsification,” which mathematically caps a portfolio’s ability to achieve superperformance). The optimal zone is a small number of concentrated positions (roughly 4–12 names) sized so that no single trade risks more than 1.25–2.5% of total equity.

This matters to a practitioner because position sizing is the mechanism that translates a positive-expectancy system into an actual compounding equity curve — a sound stop-loss and entry methodology can still produce ruin if position size is set arbitrarily rather than mathematically. Minervini explicitly links position size to stop-loss width (“backing into risk”): either the stop or the size must flex so that dollar risk stays within the equity-risk ceiling, meaning these two variables can never be set independently of each other.

Structurally, this section is the capital-allocation layer that sits above every individual trade decision covered in Sections 6–7 — it determines how big each Trend Template/VCP entry should be, how to actively reallocate capital between winners and laggards (the “two-for-one rule”), and how concentration should evolve as conviction in specific names strengthens or weakens.

Core Concepts & Mechanics

The equity-risk ceiling (1.25–2.5%) — total dollar risk on any single trade, not the position’s raw dollar size, must be capped at 1.25–2.5% of total portfolio equity; this is the master constraint governing all sizing decisions.
Backing into risk (position size ↔ stop width tradeoff) — position size and stop-loss percentage are inversely linked; widening one requires tightening the other to keep dollar-equity risk within the ceiling, and only one variable can be adjusted at a time to solve for the target risk level.
Concentration over diversification — Minervini deliberately holds only 4–12 stocks (up to 16–20 for larger professional portfolios), arguing that spreading capital across 15–20+ names (“di-worsification”) makes it mathematically impossible for even a huge individual winner to meaningfully move the portfolio.
Optimal position size derived from expectancy math — using Optimal F or the Kelly Criterion, a trader with a 2:1 reward/risk ratio and 50% batting average mathematically arrives at an optimal ~25% position size (roughly four equal positions), directly tying position sizing to the trader’s own measured statistics rather than arbitrary convention.
Scaling in based on proof of performance — Minervini often starts new positions small (5–10% of the portfolio) to limit risk while a stock “proves itself,” then increases size as it performs according to plan, aligning capital commitment with demonstrated (not hoped-for) strength.
The two-for-one rule (capital reallocation) — when weak performers are dragging on a portfolio, halving positions in the two worst performers (e.g., 20%→10% each) frees up capital to fund a full-size position in a stronger emerging candidate, without requiring full liquidation of the laggards.
Not selling leaders too quickly — especially at the start of a new bull market, retaining 25–50% of an original position in a proven leader after taking partial profits preserves exposure to potentially the market’s best continued performer, rather than fully rotating out for the sake of “locking in” gains.
Gap-risk exposure at extreme concentration — a stop-loss becomes worthless against an overnight gap-down (e.g., a 50% single-day drop), so extreme over-concentration (75–100% in one name) exposes the account to catastrophic, unhedgeable loss that no predefined stop can prevent.
The “garden” reallocation mindset — portfolio management is framed as continuous pruning and nurturing: stagnant or underperforming positions (“weeds”) should be trimmed to redeploy capital into stronger, still-developing candidates (“flowers”), even absent an outright stop-loss trigger.

Technical Terminology & Reference Table

Term	Operational Definition
Equity Risk per Trade	The percentage of total portfolio equity exposed to loss on a single trade (position size × stop-loss %); capped at 1.25–2.5%.
Backing into Risk	Solving for either stop-loss width or position size (holding the other fixed) so that dollar-equity risk stays within the target ceiling.
Optimal F	A position-sizing formula that calculates the fraction of capital to risk per trade to maximize long-run geometric portfolio growth.
Kelly Formula (Kelly Criterion)	A closely related optimal-betting formula using win rate and payoff ratio to determine ideal fractional position size.
Di-Worsification	Minervini’s term for excessive diversification that dilutes position sizes to the point of preventing superperformance contribution from any single winner.
Two-for-One Rule	Halving positions in two underperforming stocks to fund a full-size position in a stronger emerging candidate.
Risk of Ruin	The probability of catastrophic, irrecoverable account loss from oversized positions or repeated excessive risk-taking.
Gap-Down Risk	The risk that a stock opens sharply below its stop-loss level overnight, rendering the predefined stop unable to limit the loss as intended.

The Author’s Market Philosophy

Minervini assumes that concentrated, high-conviction positioning — not broad diversification — is the actual mechanism by which sustainable outperformance is achieved, since diversification’s smoothing benefit comes at the direct cost of diluting the impact of the trader’s best ideas. He treats participant behavior as prone to two opposite but equally damaging biases: gambling overconfidence (all-in sizing that risks ruin) and diversification-as-security-blanket (spreading capital so thin that no single position can matter), both of which he frames as failures to apply the actual math of expectancy to capital allocation. His mental model expects the reader to treat position sizing as a solvable equation derived from the trader’s own measured statistics (average gain, average loss, batting average) rather than a matter of comfort or convention, and to actively manage the portfolio as a living, continuously reallocated structure rather than a static set of buy-and-hold decisions.

Systemic & Portfolio Integration

Position sizing is the direct capital-allocation expression of the expectancy formula from Section 3–4: risk-per-trade limits (1.25–2.5% of equity) and the stop-width/size tradeoff ensure that no single loss can meaningfully damage the compounding trajectory built by a positive-expectancy system. The two-for-one reallocation rule and “don’t sell leaders too quickly” principle extend systematic risk management to the portfolio level, continuously rebalancing exposure toward the strongest-performing, highest-conviction names while capping the drag from underperformers.

Important Formulas, Data, or Initial Examples

Position sizing example: $100, 000 p o r t f o l i o,$ 50,000 position (50%), 10% stop → $5,000 loss = 5% of equity (too much risk).
Backing-into-risk example: $25, 000 p o s i t i o n (25$ 2,500 risk = 2.5% of equity (high end of acceptable range); tightening the stop to 5% reduces risk to 1.25% ( $1, 250); a l t e r n a t i v e l y, c u t t i n g t h e p o s i t i o n t o$ 12,500 (12.5%) achieves the same 1.25% equity risk at a 10% stop.
Position Sizing Guidelines: 1.25–2.5% equity risk per trade; 10% maximum stop; average losses no more than 5–6%; never exceed 50% in one position; optimal 20–25% positions in top names; 10–12 stocks total (16–20 for larger professional portfolios).
Gap-risk example: an 80% position that gaps down 50% produces a 40% loss of total equity; a 25% position under the same 50% gap produces only a 12.5% loss — illustrating why the position-sizing ceiling protects against unhedgeable gap risk.
Optimal F/Kelly example: a 2:1 reward/risk trader with a 50% batting average and 5% average loss / 10% average gain mathematically derives an optimal ~25% position size (roughly four equally weighted positions).
Two-for-one example: five stocks each at 20% of the portfolio; halving positions in the two weakest (20%→10% each) frees 20% of total equity to fund a new full-size position.

Active Recall Evaluation

Explain the mathematical relationship between position size and stop-loss width described as “backing into risk,” and why a trader can’t set both variables independently.
Using the gap-down risk example, explain why a hard stop-loss order does not fully protect a trader against catastrophic loss at high levels of portfolio concentration.
Why does Minervini consider “di-worsification” just as dangerous to long-run performance as reckless over-concentration, even though the two errors seem to point in opposite directions?
Walk through how Optimal F or the Kelly Criterion, applied to a 2:1 reward/risk trader with a 50% batting average, arrives at an optimal position size of roughly 25%, and explain what this implies about portfolio structure.
Describe the mechanics and purpose of the “two-for-one rule,” and explain why Minervini prefers this reallocation method over either holding all positions unchanged or fully liquidating underperformers.

Answer Key (spoiler)

Dollar-equity risk on a trade equals position size (as % of equity) multiplied by stop-loss width (as % of the position); since the total must stay within the 1.25–2.5% equity-risk ceiling, increasing one variable mathematically requires decreasing the other to hold the product constant. A trader can’t independently choose an aggressive position size (e.g., 50%) and a wide stop (e.g., 10%) simultaneously, because the resulting dollar risk (5%) would breach the ceiling — one of the two must give way to keep risk in bounds.
A stop-loss order only executes at the specified price if the market trades through that level in a continuous fashion; if a stock gaps down overnight (e.g., 50% in one session), there is no trading activity between the prior close and the next open for the stop to trigger at the intended price, so the position is sold at whatever the next available (much lower) price is. This means the position-sizing discipline — not the stop-loss itself — is the only real defense against catastrophic gap risk, since a stop cannot protect against a price discontinuity.
Over-concentration risks catastrophic single-trade loss (risk of ruin), while over-diversification prevents any single winning position from being large enough to meaningfully move the portfolio’s overall return, effectively capping the account’s upside at an average of many small, diluted positions. Both errors, despite pointing in opposite directions, undermine the same goal — compounding meaningful, superperformance-level returns — because one exposes the account to unacceptable risk while the other guarantees mediocre, diluted results even from a correctly identified big winner.
Optimal F/Kelly-style formulas calculate the fraction of capital that maximizes long-run geometric growth given a trader’s actual win rate and payoff ratio; for a 2:1 payoff trader who wins half the time, the math resolves to roughly a 25% allocation per position, implying a portfolio of about four equally weighted positions. This tells the trader that mathematically optimal growth comes from concentrated exposure to a small number of high-conviction names — not from spreading capital across many positions — since the formula directly rewards concentration when the underlying edge (reward/risk ratio and win rate) is favorable.
The two-for-one rule reduces (rather than eliminates) exposure to the two weakest-performing positions by half, freeing that capital to fund a full-size position in a stronger, more promising candidate meeting the trader’s buy criteria. Minervini prefers this over holding unchanged (which lets underperforming capital sit idle or continue lagging) or full liquidation (which may prematurely abandon a name that hasn’t hit its stop and could still recover), since partial reduction preserves some upside optionality in the laggards while still reallocating the bulk of the capital toward higher-conviction opportunities.