Sizing Positions by Feel Instead of Rules: Why Your Gut Is Not a Sizing Algorithm

Position sizing by feel—making trades of different sizes based on confidence level, market conditions, or recent wins—is the silent killer of profitable trading. A trader with a sound strategy and 60% win rate can still blow the account by sizing inconsistently: oversizing winners (after wins), undersizing losers (to protect against the next loss), and scaling into losing positions (averaging down because the trade "feels like it will work"). This article shows how sizing by feel destroys math-based edge and how to rebuild position sizes using fixed rules.

Lede

Your strategy's profitability is not determined by your win rate or average win size—it's determined by position sizing discipline. A trader with a 40% win rate and consistent sizing can be profitable; a trader with a 70% win rate and inconsistent sizing will eventually blow the account. Position sizing by feel is insidious because it feels like control: you're "managing risk" by taking smaller positions when nervous and larger positions when confident. In reality, you're doing the opposite—amplifying losses by oversizing after wins and cutting positions after losses. The math is brutal: if you risk 10% on a trade and lose, you need 11% returns to break even, not 10%. Larger position sizes after draws are not recovery; they're exponential destruction.

Quick definition: Position sizing by feel is the practice of varying position size based on subjective factors (confidence, recent performance, emotional state) rather than objective rules (fixed risk amount, fixed percentage of capital, Kelly formula). It's the opposite of disciplined sizing.

Key takeaways

• Consistent position sizing is more important to long-term profitability than having a higher win rate or larger average winners.
• Sizing by feel after losses (averaging down, emotional undersizing) and after wins (overconfidence sizing) creates asymmetric risk: losses compound faster than gains.
• The Kelly formula and fixed-risk sizing models are deterministic; they remove emotion and scale sizing to your actual edge and bankroll.
• A trader risking 2% per trade can survive eight consecutive losses and remain at 85% of starting capital; a trader risking 5% cannot.
• Documenting your sizing rule and reviewing it monthly prevents drift (the gradual expansion of position sizes that occurs without conscious decision).

The Arithmetic of Position Sizing

Position sizing sounds simple but is actually the foundation of portfolio survival. Here's why: losses compound differently than gains.

If you have $100,000 and lose 20%, you're at $80,000. To return to $100,000, you need a 25% gain (not 20%). If you then lose 20% again, you're at $64,000. To return to $100,000 now requires a 56% gain. Three 20% losses in a row require a 97% gain to recover.

Extrapolate this: a trader risking 5% per trade faces these scenarios:

• After 3 losses: down 14.3%, needs 16.7% gain to recover.
• After 5 losses: down 22.6%, needs 29.2% gain to recover.
• After 8 losses: down 33.5%, needs 50.4% gain to recover.

A trader risking 2% per trade after the same 8 losses: down 15.7%, needs 18.6% gain to recover. Same strategy, same losses, vastly different damage. Position sizing is not about being conservative; it's about surviving long enough to let your edge play out.

The second math problem with feeling-based sizing: traders systematically oversize winners and undersize losers. After winning two trades, confidence rises and you size up the third trade. After losing, you size down or stop trading. This behavior does the opposite of risk management—it amplifies losses and limits gains.

Example: A trader with a 55% win rate (statistically positive edge) and an average winner of 2:1 risk-to-reward sizes the following way:

• Wins: sizes up to 5% risk
• Losses: sizes down to 1% risk
• Breakeven trades: sizes 2% risk

Over 100 trades (55 wins, 45 losses):

• 55 wins × 5% risk × 2 (reward ratio) = 550 points
• 45 losses × 1% risk × 1 = –45 points
• Net: +505 points

Compare to consistent 2% sizing:

• 55 wins × 2% risk × 2 = 220 points
• 45 losses × 2% risk × 1 = –90 points
• Net: +130 points

The feeling-based approach nets 505 points; the consistent approach nets 130 points. The difference is not the edge (both have the same 55% win rate); it's the sizing asymmetry. However, this advantage disappears in a run of losses (which statistically will occur). A trader sizing 5% on wins and 1% on losses faces devastating drawdowns when losses cluster—exactly when overconfidence has run them up to 5% per trade.

The Kelly Formula: A Rule-Based Alternative

The Kelly formula is a mathematical model for optimal position sizing given your edge and odds:

Kelly % = (Win% × Avg Win) – (Loss% × Avg Loss) / Avg Win

For a trader with a 55% win rate, 2:1 reward-to-risk ratio, and average winner of $2,000 and average loser of $1,000:

Kelly % = (0.55 × 2) – (0.45 × 1) / 2
Kelly % = 1.10 – 0.45 / 2 = 0.325 or 32.5%

The Kelly formula says you should risk 32.5% of your bankroll per trade to maximize long-term growth. However, this is the aggressive Kelly. In practice, traders use half-Kelly (16.25%) or quarter-Kelly (8.1%) to reduce drawdown volatility. The formula is not an instruction to risk that much; it's a ceiling and a guide.

Most traders adjust the Kelly output downward for safety:

• Full Kelly: highest growth, highest drawdown
• Half Kelly: balanced growth and drawdown
• Quarter Kelly: conservative, reduced volatility
• Fixed percentage: 2% risk, the simplest rule

For a trader with a 55% win rate and 2:1 ratio, quarter-Kelly would suggest risking roughly 8% of bankroll per trade. Half-Kelly would suggest 16%. Most experienced traders operate at 2–5% per trade, which is much more conservative. This conservatism is not weakness; it's recognition that real-world edge is smaller than theoretical edge, and drawdowns hurt less.

Fixed-Risk Sizing: The Simplest Rule

The easiest position sizing rule is fixed-risk: risk the same dollar amount or same percentage on every trade.

Fixed-dollar sizing example: Risk $500 per trade, every time. If the trade offers a 2% stop loss (worth $500), you buy position such that 2% equals $500. If the trade offers a 5% stop loss, you buy less to keep risk at $500. This scales position size to the stop-loss distance automatically and removes emotion.

Fixed-percentage sizing example: Risk 2% of your bankroll per trade. A $100,000 account risks $2,000 per trade. If you have a $80,000 account (after losses), you risk $1,600. Position size declines automatically after losses, limiting compounding damage. Position size increases automatically after gains, allowing compounding growth.

Fixed-percentage sizing is superior because it scales with your actual bankroll. A trader who risks a fixed $2,000 on a $100K account but then grows to $200K has not adjusted—they're now risking 1% instead of the intended 2%. Fixed-percentage captures this automatically.

Drift: The Silent Killer

Even traders who start with rules often drift toward feeling-based sizing. Drift occurs gradually: you make a few trades above your rule, wins arrive, and you justify continuing. You realize three months later that you've been averaging 3.5% risk instead of the intended 2%. By then, a drawdown strikes and you're in deeper than planned.

Prevent drift with a monthly review ritual:

1. Calculate the percentage of bankroll risked on each trade from the past month.
2. Compute the average (should be near your target: 2%, 2.5%, etc.).
3. Calculate the largest single trade (should be near your cap: no single trade above 5% or your rule).
4. If average or largest exceed target, write why (happened to add to winners, felt confident, etc.) and reset for next month.

This 10-minute monthly check prevents the slow creep that transforms a 2% rule into a 4% reality.

Position Size vs. Trade Entry

A subtle error: traders confuse position sizing with trade entry. Position sizing answers "How much?" Entry answers "When?" They are not the same. You might enter trades every day (frequent entry) with small position sizes (2% risk). Or you might enter once per week with the same sizing. Entry frequency doesn't change optimal sizing. A rule-based sizing system works regardless of entry frequency.

Similarly, sizing is independent of profitability. A trader with a 45% win rate needs smaller sizing than one with a 65% win rate (to account for more frequent losses), but both should use a consistent rule, not feel.

Real-World Position Sizing Disasters

Example 1: Averaging down. Michael is long gold at $1,800 with a 2% position size ($2,000 risk). Gold falls to $1,700 and his position is down $2,000. He feels the trade is still valid, so he "averages down"—adds another 2% position at $1,700. Now he has 4% risk instead of 2%. Gold falls to $1,600 and he averages down again, now holding 6% risk. Gold continues to $1,400, and he's down $6,000 on a position that began with $2,000 planned risk. His averaging-down strategy multiplied the loss 3x by feeling-based sizing increases.

Example 2: Overconfidence after wins. Sarah had a rule: risk 2% per trade. After winning five trades in a row, she felt confident and sized her sixth trade at 4% ("I'm hot"). The sixth trade hit her stop loss. She lost twice as much as planned because of the feeling-based increase. The confidence was not data; it was recency bias. Her win streak told her nothing about the sixth trade's probability.

Example 3: Undersizing after losses. Dev had a 2% rule but was emotionally bruised after a 3-trade losing streak. He dropped to 0.5% on the next two trades (undersizing due to fear) before returning to 2%. His undersizing meant he missed the recovery trades at smaller scale. He left money on the table by letting emotion override his rule.

Example 4: Disastrous scaling in. A trader with a rule to risk 2% per trade decides to "scale in" to a position based on feel. He enters 1%, then adds 2% more when the trade moves slightly in his favor, then adds another 2% when it seems to be working. He's now risking 5% on a single idea. The trade reverses and he loses more than planned because sizing became additive (feeling-based layers) instead of disciplined.

Sizing Decision

Real-World Examples

Example 1: Consistent sizing works. Jordan sets a rule: risk 2% per trade, maximum 4% in any single position. Over two years and 200 trades with a 52% win rate, Jordan nets $180,000 on a starting $200,000 account. Annualized return: 30% with a max drawdown of 18%. The rule prevented oversizing after wins and undersizing after losses. Consistency let the edge play out.

Example 2: Feeling-based sizing destroys edge. Alex has the same strategy and win rate as Jordan but sizes by feel. After winning three trades, Alex adds 5% position sizes. After losing two, Alex cuts to 0.5%. Over the same two years, Alex nets $40,000 (still positive, but 78% worse than Jordan). The drawdown during the losing streak hit 35% because Alex oversized the early winners just before a drawdown, then undersized the recovery trades, missing them.

Example 3: Kelly formula baseline. Taylor calculates a 55% win rate, 2:1 reward ratio. The Kelly formula suggests 16% risk (half-Kelly). Taylor, being conservative, adopts 4% risk per trade (one-quarter Kelly). Over 150 trades, Taylor captures the edge at a safe drawdown level. The formula provided the framework; the conservative adjustment provided the safety.

Common Mistakes

1. Averaging down as risk management. Averaging down is the opposite of risk management. It increases exposure to a failing thesis. Use fixed sizing instead. If a position warrants a 2% risk, make that 2% assignment once, not in layers.

2. Oversizing after wins. A win tells you nothing about the next trade's probability. Overconfidence sizing after wins is not edge; it's bias. Stay disciplined and constant.

3. Undersizing after losses. A loss is not evidence to reduce sizing; it's randomness within your strategy's variance. Undersizing after losses causes you to miss recovery trades at full size.

4. Confusing position size with risk management. Position size is one of several risk tools. You also need stop losses, diversification, and portfolio-level limits. Sizing alone doesn't guarantee survival.

5. Not adjusting sizing downward for small accounts. A trader with a $50,000 account cannot risk 5% per trade ($2,500); the drawdown math becomes unsustainable. Smaller accounts require smaller position sizing (1–2% risk per trade).

6. Mixing fixed-dollar and fixed-percentage sizing. Pick one and stick with it. Mixing creates confusion and drift. Fixed-percentage is preferred because it automatically scales with bankroll.

FAQ

What's the correct percentage to risk per trade?

Most successful traders operate at 1–3% risk per trade. Conservative traders use 1–2%. Aggressive traders use 3–5%. If you're below 1%, you're not letting your edge compound. If you're above 5%, a normal losing streak will hurt badly. Start at 2% and adjust based on the drawdowns you experience.

Should I use the Kelly formula or fixed-percentage sizing?

Kelly formula gives you a theoretical optimal percentage. Most traders then apply a safety discount (half-Kelly or quarter-Kelly) and use that as their fixed percentage. So the process is: (1) calculate Kelly, (2) divide by 2 or 4, (3) use the result as fixed percentage. This combines the rigor of Kelly with the simplicity of fixed sizing.

What if my stop loss is very far away (8% of price) and risking 2% would mean a tiny position?

Move the stop loss closer (to 5%) or skip the trade. Do not expand your position size above your rule because a specific trade has a distant stop. Maintaining sizing discipline is more important than entry into any single trade.

Can I adjust sizing based on volatility?

Yes, but only within your framework. If you set a rule "Risk 2% on low-volatility trades, 1.5% on high-volatility trades," that's fine—it's still a rule, not feeling. However, most traders use this as cover to oversize (calling a trade "low volatility" to justify a 3% risk). If you adjust for volatility, document the criteria clearly.

What happens if I'm wrong about my win rate and it's actually 45% instead of 55%?

This is why Kelly formula is useful and why conservatism matters. If you calculated Kelly at 55% win rate and sized at 16%, but your real win rate is 45%, you're overexposed. Using half-Kelly (8%) or quarter-Kelly (4%) provides a safety margin for these miscalculations. It's not about being safe; it's about surviving your errors.

How do I track position sizing drift?

Monthly: calculate the percentage of bankroll risked on each trade. Average them. If the average is above your target, you've drifted. Example: If you track ten trades and average 3.2% when your rule is 2%, you drifted 60%. Identify which trades pushed above 2% and reset next month.

Related concepts

• Fixed Dollar Sizing Across Market Cycles
• Investment Policy Statements: Your Written Rules
• Chasing Performance Into High-Risk Assets
• Running a Personal Risk Mistake Audit

Summary

Position sizing by feel is the most common form of self-sabotage in trading. Your strategy can have a solid edge, but inconsistent sizing will destroy that edge through compounding losses and overconfidence moves. Fixed-risk (either fixed-dollar or fixed-percentage) sizing removes emotion and scales automatically to your bankroll. The Kelly formula provides a theoretical framework; most traders then apply a safety discount and use the result as a constant rule. The discipline is simple: calculate once, apply every trade, review monthly. This consistency is more powerful to long-term wealth than any prediction ability.

Next

→ Running a Personal Risk Mistake Audit