How do you handle drawdown rules across a portfolio of EAs (prop firm style)? - page 2

New comment
The first mistake is assuming that running multiple EAs automatically reduces your overall risk. It doesn't. Diversification only works when the strategies are genuinely independent. You can have five different EAs running on five different symbols, but if they're exposed to the same market conditions, you're essentially taking the same risk five times

Personally, I would approach this from the portfolio level rather than managing each EA individually

First, forget about fixed lot sizes. Each EA should have its own risk allocation based on its historical drawdown, volatility, recovery factor, and actual contribution to the portfolio. An EA generating higher returns doesn't necessarily deserve a larger allocation, especially if those returns come with excessive downside exposure

Second, correlation should be measured using synchronized equity or return data, including floating P/L rather than simply comparing winning and losing trades. Even then, historical correlation isn't enough. Strategies that appear completely independent can suddenly become highly correlated during extreme market conditions

And finally, I wouldn't wait until the portfolio reaches 5% or 8% drawdown before taking action, especially on a prop firm account. By then, you may already be dangerously close to violating the rules. I'd establish a portfolio wide risk budget, with daily and overall equity limits comfortably below the firm's actual thresholds

As for cutting position sizes versus disabling the weakest EA, I prefer reducing exposure before the situation becomes critical. But simply switching off the worst performing EA isn't necessarily the answer either. Today's weakest performer could be tomorrow's strongest. What matters is whether its underlying strategy is still behaving within its expected statistical parameters

One final point: your master risk controller should have the authority to override every EA, including preventing new entries and closing positions when necessary. Otherwise, you're not really controlling portfolio risk. You're just monitoring it

The objective isn't to squeeze the maximum possible profit out of multiple EAs. It's to make sure one bad market session doesn't wipe out months of performance or violate your account limits

That's how I'd approach it.

Alex Holloway #:
The first mistake is assuming that running multiple EAs automatically reduces your overall risk. It doesn't. Diversification only works when the strategies are genuinely independent. You can have five different EAs running on five different symbols, but if they're exposed to the same market conditions, you're essentially taking the same risk five times

Personally, I would approach this from the portfolio level rather than managing each EA individually

First, forget about fixed lot sizes. Each EA should have its own risk allocation based on its historical drawdown, volatility, recovery factor, and actual contribution to the portfolio. An EA generating higher returns doesn't necessarily deserve a larger allocation, especially if those returns come with excessive downside exposure

Second, correlation should be measured using synchronized equity or return data, including floating P/L rather than simply comparing winning and losing trades. Even then, historical correlation isn't enough. Strategies that appear completely independent can suddenly become highly correlated during extreme market conditions

And finally, I wouldn't wait until the portfolio reaches 5% or 8% drawdown before taking action, especially on a prop firm account. By then, you may already be dangerously close to violating the rules. I'd establish a portfolio wide risk budget, with daily and overall equity limits comfortably below the firm's actual thresholds

As for cutting position sizes versus disabling the weakest EA, I prefer reducing exposure before the situation becomes critical. But simply switching off the worst performing EA isn't necessarily the answer either. Today's weakest performer could be tomorrow's strongest. What matters is whether its underlying strategy is still behaving within its expected statistical parameters

One final point: your master risk controller should have the authority to override every EA, including preventing new entries and closing positions when necessary. Otherwise, you're not really controlling portfolio risk. You're just monitoring it

The objective isn't to squeeze the maximum possible profit out of multiple EAs. It's to make sure one bad market session doesn't wipe out months of performance or violate your account limits

That's how I'd approach it.

Good points…

…especially about correlation and having one controller at the portfolio level. 


I think this is where things get tricky in real trading, because an EA can look great on its own but behave very differently when combined with other strategies.

In my case, I’m interested in finding a prácticas balance between risk allocation and keeping the portfolio flexible. 

Disabling an EA just because it has a bad week can be a mistake, but letting it continue trading without checking if its behavior has changed is also risky.

One thing I’m still thinking about is how to define the portfolio risk budget in a way that adapts to changing market conditions, not just historical drawdown. For example, would you adjust exposure based on current equity and floating losses, or also use volatility and correlation changes as triggers?

I agree that a master controller is essential. The challenge is making it responsive enough to protect the account without constantly interfering with strategies that are still working as expected.


Cristian David Castillo Arrieta #:

Good points…

…especially about correlation and having one controller at the portfolio level. 


I think this is where things get tricky in real trading, because an EA can look great on its own but behave very differently when combined with other strategies.

In my case, I’m interested in finding a prácticas balance between risk allocation and keeping the portfolio flexible. 

Disabling an EA just because it has a bad week can be a mistake, but letting it continue trading without checking if its behavior has changed is also risky.

One thing I’m still thinking about is how to define the portfolio risk budget in a way that adapts to changing market conditions, not just historical drawdown. For example, would you adjust exposure based on current equity and floating losses, or also use volatility and correlation changes as triggers?

I agree that a master controller is essential. The challenge is making it responsive enough to protect the account without constantly interfering with strategies that are still working as expected.


I wouldn't rely on historical drawdown alone, nor would I use a single indicator to determine exposure. Instead I'd use a combination of all four: current equity, floating losses, volatility, and correlation. However, they shouldn't carry equal weight. Equity and floating drawdown should determine the portfolio remaining risk budget, while volatility and correlation should determine how that budget is distributed among the EAs

Here's how I'd structure it:
1. Equity and floating drawdown - Primary risk control

This should always take priority. Your master controller needs to calculate the remaining risk budget in real time, accounting for realized daily losses, floating losses, and the potential losses of all open positions if their stop losses are triggered

For example, suppose you're managing a $100K prop account with a 5% daily loss limit. I wouldn't allow the portfolio to consume the entire $5K. I'd establish an internal daily loss threshold of around 2-3%, depending on the strategies involved

Once the portfolio approaches that threshold, the controller should progressively restrict new exposure. If the hard internal limit is reached, it should stop new entries and, where necessary, close positions

2. Volatility - Dynamic position sizing

Volatility determines how much risk each position introduces under current market conditions

For instance, if XAUUSD volatility suddenly doubles, maintaining the same lot size could expose the portfolio to significantly larger price fluctuations

Rather than disabling the EA immediately, I'd adjust its position sizing according to volatility, provided the strategy still operates within its expected conditions 

However, I wouldn't continuously change lot sizes based on every small fluctuation. That would introduce unnecessary noise and potentially interfere with the strategy itself

3. Correlation - Portfolio exposure control

This is probably the most overlooked factor when running multiple EAs

Imagine three EAs operating independently, each risking 0.5% per trade. On paper, they appear diversified. But if all three become exposed to the same market movement, your effective portfolio risk could become highly concentrated

I'd monitor rolling correlations between their returns, alongside their actual directional exposure

If correlation rises significantly, the controller should reduce the combined risk allocation rather than blindly treating each EA as an independent strategy

4. Strategy performance - When to disable an EA

I wouldn't disable an EA simply because it experienced a losing streak. Losses are part of every statistical trading model

Instead, I'd monitor whether its current behavior deviates meaningfully from its established performance characteristics

For example, changes in average trade duration, loss distribution, execution costs, trade frequency, or drawdown patterns could indicate that the strategy is operating outside its expected conditions

That's when I'd consider reducing its allocation or temporarily suspending new trades

5. The master controller - My preferred approach

I'd structure it around three operating states:

• Normal: All EAs operate within their assigned risk budgets

• Defensive: Portfolio drawdown, volatility, or correlation exceeds predefined warning thresholds. New exposure is reduced, while existing positions remain subject to their normal risk controls

• Protection: The portfolio approaches its internal loss ceiling. New entries are blocked and existing positions are reduced or closed as necessary

The important distinction is that the controller shouldn't continuously interfere with trading decisions. It should manage exposure, not replace the underlying strategies

And remember, the controller must calculate limits according to the prop firm's actual rules, including whether daily drawdown is measured from starting balance, equity, or another reference point

One final consideration: I'd also include a safety buffer for slippage, commissions, and sudden market gaps. A perfectly calculated risk budget means very little if your execution assumptions fail during extreme volatility

My philosophy is simple: let each EA handle its trading strategy, but never allow an individual EA to dictate the risk of the entire portfolio

You don't need to predict which strategy will perform best tomorrow. You need a system that can survive when several of them perform badly at the same time

That's the difference between managing multiple EAs and managing an actual trading portfolio

Similar problem from a different angle: I run the same TradingView strategy across a funded account and a couple of challenge accounts. Same signal, but each account needs different risk: tighter daily limits on the funded one, more room on the challenges. Lot sizing per account by hand became a mess, and identical settings everywhere doesn't work either.


I'm moving to a setup where one webhook fans out to all accounts, and each account has its own risk profile (per-trade risk, daily loss buffer, reset time from the firm's server clock). +1 to Muhammad's point about restoring the daily lock after a VPS restart, and to Alex's point that the drawdown reference (balance vs equity) differs by firm. That's easy to miss when the accounts are with different firms.


Chris, when you pause accounts on the equity alert, do you pause all of them or only the ones nearest their limit? With mixed funded and challenge accounts I'm not sure treating them the same makes sense.

patrickcollinsemail #:


Chris, when you pause accounts on the equity alert, do you pause all of them or only the ones nearest their limit? With mixed funded and challenge accounts I'm not sure treating them the same makes sense.

Thanks Patrick,

in my case "affected" meant all accounts that ran the same main EA, not only the one closest to its limit. They were more or less all in the same trade. If I pause only the worst one, the same position keeps running on the others and i'm just waiting for the next alert.

With funded and challenge accounts mixed I'd split it in two steps. The alert level is per account, because the limits and the reference (balance or equity) are different. But the decision what to pause is about exposure. If the same signal is underwater on all of them, I'd pause it everywhere, also on the challenge accounts that still have room. Thats cheaper than losing the funded one.

Hope this helps :)
Chris
Cristian David Castillo Arrieta:

I've been running several EAs together on the same account instead of one single strategy, mainly to smooth the equity curve, but it's forcing me to rethink risk management completely. A single trailing stop or fixed lot size per EA doesn't work anymore, because the drawdown that actually matters is the combined one, not each EA's individual number.

Right now I'm doing something fairly simple: each EA reports its floating result to a shared variable/file, and a master risk block starts cutting position sizes proportionally once the aggregated equity drawdown crosses a threshold (testing 5% and 8% steps). It works, but it feels crude, especially when trying to respect daily loss limits like the ones prop firms impose.

For those running multi-strategy portfolios: how are you measuring correlation between EAs in real time, not just from backtest reports? And once combined drawdown gets close to the limit, do you cut size, pause the weakest EA, or something else? Curious what's actually working for people trading this way instead of a single EA.

combined drawdown and live correlation of several EAs under prop-firm limits; existing answers lack the measurement method.
MY ANSWER: Start by seeing each EA's contribution separately: a unique magic number per EA and a report built from deal history filtered by that number. Build daily P&L per EA, including floating result at day end, and compute rolling 30-60 day correlation. EAs that lose on the same days count as one position when sizing risk. At account level, track equity against the firm's exact rule (balance or equity, reset time zone, trailing or static) and set an internal limit well under it, e.g. 3.5% against a 5% rule. When hit, block new entries first, then close positions, and persist that state across terminal restarts.
These are the best methods I use:
First, checking the pip value of the asset being traded;
Adjusting the trade size to avoid overexposure;
Setting a stop-loss that represents less than 0.5% of the account balance;
And, most importantly, ensuring a timely exit from the trade.
Chris K. #:
Thanks Patrick,

in my case "affected" meant all accounts that ran the same main EA, not only the one closest to its limit. They were more or less all in the same trade. If I pause only the worst one, the same position keeps running on the others and i'm just waiting for the next alert.

With funded and challenge accounts mixed I'd split it in two steps. The alert level is per account, because the limits and the reference (balance or equity) are different. But the decision what to pause is about exposure. If the same signal is underwater on all of them, I'd pause it everywhere, also on the challenge accounts that still have room. Thats cheaper than losing the funded one.

Hope this helps :)
Chris

Thanks Chris, that makes sense. If it's the same trade on every account, pausing just the worst one doesn't really fix anything, the risk is still there on the others. "Alert per account, pause by signal" is a good way to put it.

when you pause a signal everywhere, what do you want to happen to the trades that are already open? Close them automatically, or leave them and decide yourself? I can see it going either way. Closing protects the limit, but it also locks in a loss that might have come back.

pat #:

Thanks Chris, that makes sense. If it's the same trade on every account, pausing just the worst one doesn't really fix anything, the risk is still there on the others. "Alert per account, pause by signal" is a good way to put it.

when you pause a signal everywhere, what do you want to happen to the trades that are already open? Close them automatically, or leave them and decide yourself? I can see it going either way. Closing protects the limit, but it also locks in a loss that might have come back.

Hi pat,

for me pause means first no new entries. The open trades I wouldnt close automatically on every account.

I'd look at how much room is left on each account if the open trade runs to its stop, or gaps past it. Where that would break the limit, close it. Thats mostly the funded one. Where there is still room, leave it with the stop and decide myself.

You're right that closing locks in the loss. But on a funded account a loss that might come back is not worth the account.

Best
Chris
12
New comment