Blog
/
AI Trading Journal
AI Trading Journal
How AI Calculates Your Optimal Position Size for Prop Firm Accounts
July 2026
6 min read
Risk Management
Most prop firm traders use a fixed risk percentage — 1% per trade, every session, regardless of context. It's simple, disciplined, and wrong for a significant portion of your trading days.
Fixed sizing ignores the reality that your edge is not constant. Your win rate shifts by setup quality. Your drawdown floor changes your room for error. Your Discipline Score tells you whether you're currently executing your system or drifting from it. A position size that's appropriate on a high-consistency day with 8% drawdown headroom is too large on a low-consistency day with 2% headroom — even if the trade setup looks identical.
AI calculates this dynamically. Here's how.
Why Fixed Risk Percentages Fail Prop Firm Traders
The problem with fixed 1% risk isn't the number — it's the assumption behind it. Fixed sizing assumes your edge is the same every day. It isn't.
The hidden cost
A trader risking 1% per trade with a Discipline Score of 4.5 is not risking 1% of their edge. They're risking 1% on trades that may not meet their system's criteria — trades that historically lose at a higher rate. The correct size on a low-discipline session isn't 1%. It's something closer to 0.25–0.5%, or zero until the behavioral state improves.
On the other side: a trader running a 9-session Discipline Score streak of 8.5 with 7% of drawdown headroom remaining may be under-risking at 0.5%. Their edge is running at full expression and their account has substantial room. AI identifies both scenarios — oversizing in low-discipline states and undersizing in high-performance states.
The 4 Inputs AI Uses to Calculate Position Size
Input 01
Historical edge by setup type
AI segments your win rate and average RR by the setup tags you log. Your A-setup has a different expectancy than your B-setup. Position size should scale with setup quality — not be uniform across all entries.
Input 02
Current drawdown proximity
How far you are from your daily or max drawdown limit directly affects how much variance you can absorb. With 1% of daily drawdown remaining, a 1% risk trade puts you at zero headroom on a single loss. AI scales size down as the floor approaches.
Input 03
Recent Discipline Score
Your 5-session rolling Discipline Score tells AI whether you're currently executing your edge or drifting from it. Behavioral inconsistency means your live win rate is lower than your historical baseline — and size should reflect that reduced edge.
Input 04
Session context
Time of day, news events, and recent P&L in the current session all influence the optimal size. A trade entered 15 minutes after a losing trade in a volatile news window has a different risk profile than the same setup in a clean London open environment.
A Real Calculation Example
Setup type logged
A-setup — liquidity sweep + displacement FVG
Historical win rate for this setup
61% over 47 trades
Average RR for this setup
1.8:1
Calculated expectancy
+0.28R per trade
Current drawdown headroom (daily)
3.1% remaining of 5% limit
5-session Discipline Score average
8.1 — high consistency
Time since last loss
No losses today
AI recommended risk
0.9% — within normal range
In this scenario, the AI confirms the trader's standard 1% risk is slightly high given the daily drawdown proximity (3.1% remaining means a single 1% loss leaves only 2.1% buffer), and recommends 0.9%. If the Discipline Score had been 5.2 instead of 8.1, the recommendation would drop to 0.5% — because behavioral inconsistency reduces the effective expectancy of even a high-quality setup.
Dynamic Sizing by Account State
| Account state |
Discipline Score |
Daily headroom |
Recommended risk |
| Full edge, clean account |
8.0+ |
> 3.5% |
0.75–1% |
| Good edge, moderate buffer |
6.5–8.0 |
2–3.5% |
0.5–0.75% |
| Reduced edge or tight floor |
5.0–6.5 |
1–2% |
0.25–0.5% |
| Low discipline or near limit |
< 5.0 |
< 1% |
0–0.25% or stop session |
Why Pure Kelly Criterion Doesn't Work for Prop Firms
The Kelly criterion is a mathematically optimal bet-sizing formula that maximizes long-run growth. It's used extensively in quantitative trading. It's also unsuitable for prop firm environments as a standalone sizing method.
The reason: Kelly optimizes for long-run growth without a hard constraint on short-run variance. Prop firms impose daily drawdown limits — typically 4–5% — that Kelly will regularly breach on losing days, even with a positive-expectancy edge.
The practical solution is a fraction-Kelly approach adjusted for prop firm constraints:
- Calculate full Kelly from your historical edge:
Kelly% = WR − [(1−WR) / RR]
- Apply a 0.25–0.35 multiplier to reduce variance (quarter-Kelly or third-Kelly)
- Further reduce by your drawdown proximity factor — the closer to the daily limit, the smaller the multiplier
- Override with a floor reduction if Discipline Score is below 6.5
This is what AI journals implement automatically. You don't need to run the calculation — you need to log your trades accurately so the AI has the historical data to make the calculation correctly.
Let AI Manage Your Position Sizing
Logify's AI Coach tracks your historical edge by setup type, monitors your drawdown proximity and Discipline Score, and tells you the appropriate risk level before each session — automatically.
Start Free with Logify
Frequently Asked Questions
What is the optimal position size for prop firm traders?
There is no universal optimal position size — it depends on your strategy's expectancy, your current drawdown proximity, your recent Discipline Score, and the volatility of the instrument you're trading. For most prop firm traders, a base risk of 0.5–1% per trade is sensible, with dynamic adjustments based on drawdown floor distance and consistency state. AI calculates this dynamically from your journal data rather than applying a static percentage.
How does AI help with position sizing?
AI analyzes your historical trade data to identify your actual win rate, average RR, and maximum drawdown sequences for each setup type. It then calculates a risk-adjusted position size that reflects your real edge — not a theoretical one. When your Discipline Score drops or your drawdown floor narrows, AI recommends reducing size automatically. When your edge is running at full expression, it flags whether you're under-sizing relative to your historical performance.
Should you use fixed percentage or dynamic position sizing?
For prop firm accounts, a modified dynamic approach works best: start with a fixed base percentage (0.5–1%), then adjust based on drawdown floor proximity and Discipline Score. Pure fixed sizing ignores account state. Pure dynamic sizing (Kelly criterion) is too volatile for prop firm environments with strict daily drawdown limits. The middle path — fixed base with behavioral and account-state adjustments — is what AI journals like Logify implement.
Does position sizing affect the Discipline Score?
Yes. Position sizing variance — deviating from your planned risk percentage without a rules-based reason — is one of the highest-weighted components of the Discipline Score. A trader who sizes up because a trade "looks great" is committing a rule violation even if the trade wins. AI flags this as a consistency failure and reduces the session score accordingly.