Risk Management | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

BASKET TRADE Management: Using multiple trade management EA's on basket trades 11 replies

Alternative Trade Management/Money Management Methods 3 replies

Risk Management 65 replies

Money Management / Risk Management 24 replies

Successful Trade Management / Risk Management 0 replies

Attachments: Risk Management
Exit Attachments

Risk Management

  • #1
  • First Post: Dec 2, 2025 11:09am Dec 2, 2025 11:09am
The textbooks usually emphasize keeping a constant risk per trade to avoid ruin.
However, when I look at what experienced traders actually do, I often see them varying their risk size based on the setup.

My question is: Is there a mathematical relation or formula for this?
The logic seems to be:
If I have a rare strategy that offers a high probability (e.g., 80% win rate), I should theoretically risk more on it. Conversely, if I have a very common strategy with a lower win rate, I should risk smaller.
But if the textbooks say "keep it constant," how do you guys actually proceed with this in the real world?
I would really appreciate any keywords or reading suggestions on this topic, as I don't want to break the rules of risk management, but I also want to maximize the potential of high-probability setups.
Thanks!
  • #2
  • Dec 2, 2025 2:58pm Dec 2, 2025 2:58pm
Yeah, there is math for varying risk (mainly Kelly Criterion / fractional Kelly), and your logic about higher win-rate setups justifying more size is technically correct. The reason books push constant risk is because nobody actually knows their edge perfectly and oversizing “A+” trades is how most accounts die. In the real world, most traders keep a base % per trade and only scale a little for really clean setups, with hard caps. So: mostly fixed risk, tiny controlled variation, never big swings based on confidence.
  • #3
  • Dec 2, 2025 10:04pm Dec 2, 2025 10:04pm
A more scientific approach is to dynamically manage risk based on the characteristics of the trading instrument, its true volatility and amplitude, and the growth of your account size.
  • #4
  • Dec 3, 2025 5:08am Dec 3, 2025 5:08am
Quoting CipherMindAI
Yeah, there is math for varying risk (mainly Kelly Criterion / fractional Kelly), and your logic about higher win-rate setups justifying more size is technically correct. The reason books push constant risk is because nobody actually knows their edge perfectly and oversizing “A+” trades is how most accounts die. In the real world, most traders keep a base % per trade and only scale a little for really clean setups, with hard caps. So: mostly fixed risk, tiny controlled variation, never big swings based on confidence.
Thank you for the response. This system of having a basic risk and increase it a little bit for A+ trades seems to give me statistically edge on trades. I will go for it for now and still try to approach to numbers to put me on the winner side.
  • #5
  • Dec 3, 2025 5:11am Dec 3, 2025 5:11am
Quoting lynn-
A more scientific approach is to dynamically manage risk based on the characteristics of the trading instrument, its true volatility and amplitude, and the growth of your account size.
Thank you.
Do you know any good source, that really go deep (or at least deep enough) into mathematical aspects of it ?
  • #6
  • Dec 10, 2025 1:05am Dec 10, 2025 1:05am
I think it's safer to only risk around 1–2% of your account per trade. In the long run, being steady and protecting your capital feels much better than chasing big wins and loss later.
  • #7
  • Dec 13, 2025 11:26am Dec 13, 2025 11:26am
I think it all depends , recently prop firms are limiting us to risk 1% per trade, even tho you do not overrisk.. They limiting us obviously apparently they do not want to pay us.. But some of them will just limit you without reason
  • #8
  • Dec 13, 2025 11:29am Dec 13, 2025 11:29am
Good question. The Kelly Criterion is the mathematical answer, but it's dangerous in practice because:

  1. It assumes you know your exact win rate and R:R - you don't, you have estimates
  2. Full Kelly is extremely aggressive and the drawdowns are psychologically brutal
  3. Your "80% setup" might be 80% in backtests but 60% in live conditions

What works better: Fractional Kelly (usually 1/4 to 1/2 Kelly) gives you most of the growth with survivable drawdowns.
Practical approach I've seen work:

  1. Base risk: 1% per trade
  2. "A+ setup": maybe 1.5-2%, never more
  3. The difference between 1% and 2% feels small but compounded over 100 trades it's massive

The real edge isn't in sizing up winners - it's in sizing DOWN or skipping marginal setups. Most traders do the opposite.
Reading: Fortune's Formula (Poundstone) for Kelly history, and anything by Ralph Vince on position sizing.

  • #9
  • Dec 14, 2025 6:56am Dec 14, 2025 6:56am
Most books push 'constant risk' because it keeps beginners from blowing up. In reality, plenty of experienced traders do size up when the stats justify it.

If you want the math behind it, look into Kelly Criterion or Optimal F- they explain how expectancy can guide position size. Just be careful: full Kelly is way too aggressive for real trading.

Most pros keep it simple:

A+ setup = risk a bit more.
Normal setup =standard risk.
Low‑quality setup = small or skip

The key is having real data, not gut feeling. Without that, constant risk is still the safest path.
For Gold, NY session offers superior tradability compared to London.
  • #10
  • Dec 22, 2025 3:55am Dec 22, 2025 3:55am
Hi guys,
the main advice I can give is this: don’t rely on the Kelly Criterion as a practical risk model.
A much more robust approach is to base your risk decisions on a Monte Carlo analysis of your own real trading statistics.
From there, the mindset should shift completely:
stop reasoning about the quality of a single setup, and start reasoning about the current statistical position of your account.
If your account is in drawdown, risk should be reduced.
If your account is in a profitable phase, risk can be increased — always within predefined limits.
Put this way it may sound a bit generic, but this principle is the real foundation that leads to account survival first, and only then to consistent growth.
In my experience, this shift in perspective makes a much bigger difference than trying to optimize position size trade by trade.
  • #11
  • Dec 30, 2025 1:49am Dec 30, 2025 1:49am
To make this more concrete, I’m sharing a sample report where the same exact trade sequence
is evaluated under two different risk models.
Entries, exits, win rate, R:R and number of trades are identical.
The only variable is how risk adapts based on the statistical state of the account.
I’ve intentionally removed any logo or reference to avoid promotional issues —
this is shared purely to illustrate the concept discussed in this thread.
The key takeaway isn’t “more profit”, but how different risk distributions over time
affect drawdown behavior, recovery speed and equity smoothness.
This kind of analysis can be reproduced by anyone using Monte Carlo on their own trade history.
If the moderators are fine with it and people find it useful, I can explain the methodology step by step
without linking to tools or external resources.

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 429 KB
  • #12
  • Jan 10, 2026 9:27pm Jan 10, 2026 9:27pm
Constant risk is safe, but in practice traders tweak it based on the setup. The important part is knowing your stats and keeping discipline—otherwise even a high-probability trade can hurt your account.
  • #13
  • Jan 12, 2026 12:41am Jan 12, 2026 12:41am
Quoting Yomcollc
Constant risk is safe, but in practice traders tweak it based on the setup. The important part is knowing your stats and keeping discipline—otherwise even a high-probability trade can hurt your account.
I actually agree with you:
a fixed risk based on real statistics is a great starting point.
What I don’t agree with is changing risk based on how good or bad a single setup looks.
I don’t think in terms of the next trade — I think in terms of the account’s statistical condition:
drawdown, streaks, equity position.
Based on that state, risk should adapt before each trade, independently of the setup itself.
This is mathematically testable: with the same trades and the same drawdown, dynamic risk driven by account state can produce several times more profit.

If you’re curious, I can send you in private the research and a small simulator so you can test it yourself.
  • #14
  • Jan 13, 2026 2:49am Jan 13, 2026 2:49am
Quoting EmadM
The textbooks usually emphasize keeping a constant risk per trade to avoid ruin. However, when I look at what experienced traders actually do, I often see them varying their risk size based on the setup. My question is: Is there a mathematical relation or formula for this? The logic seems to be: If I have a rare strategy that offers a high probability (e.g., 80% win rate), I should theoretically risk more on it. Conversely, if I have a very common strategy with a lower win rate, I should risk smaller. But if the textbooks say "keep it constant,"...

You’re right, experienced traders often scale risk with setup quality. Using higher size on high‑probability trades and smaller size on lower‑probability ones is a practical approach many adopt despite textbooks stressing constancy.
  • #15
  • Jan 17, 2026 12:56pm Jan 17, 2026 12:56pm
Quoting EmadM
The textbooks usually emphasize keeping a constant risk per trade to avoid ruin. However, when I look at what experienced traders actually do, I often see them varying their risk size based on the setup. My question is: Is there a mathematical relation or formula for this? The logic seems to be: If I have a rare strategy that offers a high probability (e.g., 80% win rate), I should theoretically risk more on it. Conversely, if I have a very common strategy with a lower win rate, I should risk smaller. But if the textbooks say "keep it constant,"...
Myfxbook and others have calculators. What I do is I always risk the same money per trade, so the lots will vary based on stop-loss distance and the pair. If the stop-loss is too close I might risk less money than usual though since there is a high chance of hitting it.
  • #16
  • Jan 23, 2026 12:34am Jan 23, 2026 12:34am
Hi everyone,
a few people asked for something more concrete than theory, so I decided to share the full PDF behind the screenshot I posted earlier.
The document does not introduce a new strategy and does not change entries, exits, win rate or R:R.
It analyzes the same exact trade sequence under two different risk models:

  1. fixed percentage risk
  2. adaptive risk driven by the statistical state of the account (drawdown / equity position)

The goal is not “more profit”, but to observe how different risk distributions over time affect:

  1. drawdown depth
  2. recovery speed
  3. equity smoothness

There are no logos, tools or links — this is purely an educational Monte Carlo study that anyone can replicate on their own trade history.
I’m sharing it for discussion purposes. If something looks wrong or debatable, I’m happy to discuss the math behind it.

Attached File(s)
File Type: pdf QUANTUM 2 eng.pdf   1.4 MB | 187 downloads
  • #17
  • Jan 24, 2026 9:56am Jan 24, 2026 9:56am
Quoting michelemonto
...It analyzes the same exact trade sequence under two different risk models: fixed percentage risk adaptive risk driven by the statistical state of the account (drawdown / equity position). {image}
Nella tabella del grafico la salita è impressionante; tuttavia nella tabella "Trade by results" sembra che il metodo standard sia migliore del metodo Quantum, se escludiamo le prime quattro operazioni, e comunque nel totale non molto dissimili. O sbalgio a leggere?
  • #18
  • Jan 26, 2026 6:16am Jan 26, 2026 6:16am
Quoting catafx
{quote} Nella tabella del grafico la salita è impressionante; tuttavia nella tabella "Trade by results" sembra che il metodo standard sia migliore del metodo Quantum, se escludiamo le prime quattro operazioni, e comunque nel totale non molto dissimili. O sbalgio a leggere?
Please note: that table shows only the first 17 trades, while the chart is built on more than 800 operations.
Comparing Quantum and Standard on just 17 trades is not representative: Quantum’s advantage emerges on the full sample, which is what you see in the equity curve.
  • #19
  • Mar 7, 2026 1:32am Mar 7, 2026 1:32am
Quoting Greg81
Good question. The Kelly Criterion is the mathematical answer, but it's dangerous in practice because: It assumes you know your exact win rate and R:R - you don't, you have estimates Full Kelly is extremely aggressive and the drawdowns are psychologically brutal Your "80% setup" might be 80% in backtests but 60% in live conditions What works better: Fractional Kelly (usually 1/4 to 1/2 Kelly) gives you most of the growth with survivable drawdowns. Practical approach I've seen work: Base risk: 1% per trade "A+ setup": maybe 1.5-2%, never more The...
First I want to thank you for the response. It opened in new way of thinking about the risk management for me. I havent heard anything about the kelly criterion and i will go for it (and also the fractional kelly).

I agree with the part that you've mentioned about how the real winners trade. Ive experienced it myself. I also found that in trading, where the emotions can kick very hard, its better you make strategies and models in a way that small percentages of choices be done by you so you can limit the effects of emotions, so finding these mathematical efficient ways to manage the risk can highly increase the chance of having a sustainable income.
  • #20
  • Mar 7, 2026 1:38pm Mar 7, 2026 1:38pm
Great thread. Lots of good points about Kelly and fractional Kelly already covered, so I'll add a different angle: portfolio-level risk.

Most of this discussion focuses on risk per individual trade, which makes sense. But if you're running multiple strategies or trading several pairs, there's a hidden risk most traders miss: correlation.

Say you risk 1% on EURUSD and 1% on GBPUSD at the same time. Your actual EUR exposure is closer to 1.5-1.8% because those pairs are heavily correlated. During a USD news event, both trades move against you simultaneously. Your "safe" 2% total risk is actually behaving like 3-4%.

What helped me practically:

1. I track a simple correlation matrix of the pairs I trade. If two pending trades have >0.7 correlation, the second one gets half size. Small rule, big impact on drawdown consistency.

2. I cap total portfolio exposure at 5% regardless of how many individual setups qualify. This forces prioritization - you naturally pick only the best setups when you have a budget.

3. I measure max drawdown at the portfolio level, not per-trade. Even if every individual trade "only" risks 1%, five correlated trades opening in the same session can create a 4-5% drawdown event that the per-trade math didn't predict.

The math: if you lose 30%, you need 43% to recover. Lose 50%, you need 100%. Protecting against deep drawdowns matters more than optimizing individual position sizes.

For anyone running EAs or multiple strategies simultaneously, I'd strongly recommend backtesting with portfolio-level metrics, not just per-strategy results. The interaction effects between strategies are where most blowups hide.
FXOptimize.com — Free EA Portfolio Optimizer for MT4/MT5
0 traders viewing now