EA Money Management | Reverse-Engineering Recommended Margin from Max Drawdown, and Seeing How Averaging-Down Blows Up an Account in Numbers

In the world of EAs (Expert Advisors), a lot of time tends to go into searching for a winning logic. But most people who actually blow up their accounts don’t fail because of the logic — they fail because of money management. Even with the same EA, whether you survive or get wiped out in one shot depends on how much margin you use, what lot size you trade, and how much floating loss the setup is designed to withstand.

This page organizes, in language and numerical examples that even beginners can follow, the things you absolutely need to decide before running an EA: recommended margin, lot size, acceptable drawdown, margin level, and the stop-out threshold. In particular, using the settings of the nanpin-style (averaging-down) EA our lab runs in its verification account, “MAC v2.0” (1.2x multiplier x up to 15 steps, 30-pip spacing, no hard stop-loss), we look at real numbers to see how a nanpin position’s floating loss balloons. We covered the principles of discretionary money management in Stop-Losses and Money Management (the 2% Rule); here we apply that same thinking to EAs and automated trading.

Why Money Management Matters More Than Anything Else in EA Trading

To cut to the conclusion: money management matters more for survival than which EA you pick. The reason is simple — an EA won’t stop partway through the way a human would when “the market moves in a way nobody expected.” A discretionary trader can decide “something’s off today, let’s step back,” but an EA just keeps executing the rules it was written with, without emotion. That’s why you need to decide, in numbers, before you ever run it, whether your account can withstand the unexpected when it comes.

There’s another reason: the cruelty of compounding. If a single loss is 2% of the account, you still have plenty of room to recover even after a losing streak. But if the design is such that a single loss can be 10%, a handful of consecutive losses can grind the account down to nearly half, and getting back from there suddenly becomes very hard. Take a look at the quick-reference table below.

Loss per trade	Balance after 5 straight losses (starting from ¥100,000)	Gain % needed to get back to even
2%	Approx. ¥90,392	Approx. 10.6%
5%	Approx. ¥77,378	Approx. 29.2%
10%	Approx. ¥59,049	Approx. 69.4%
20%	Approx. ¥32,768	Approx. 205%

The bigger each individual loss is, the more the “cost of getting it all back” spikes once you’re down. Lose 20% at a time, and five straight losses shrink the account to a third of its size — meaning you’d then need to triple it just to get back to even. What you often see on an EA’s track record as max drawdown (max DD) is exactly a record of “how much the account temporarily got shaved down.” Think of money management as the work of designing your setup so that your account can survive that worst stretch.

How to Decide on Recommended Margin and Lot Size

The starting point for EA money management is the question of “how much do I need to deposit to run this safely” — in other words, the idea of recommended margin. If you decide this by feel, it will eventually break down somewhere.

Reverse-Engineer Recommended Margin as “Max DD x Safety Factor”

The basic formula for recommended margin is the same approach commonly used on EA sales sites.

Recommended margin = required margin + (max DD amount x safety factor)

Here, max DD is the “widest drop from a peak in account equity” recorded in that EA’s backtest or forward test. The safety factor is a margin of safety that accounts for the possibility that the worst case in the past could be exceeded in the future, and is generally taken to be around 2x. Since a past max DD is not necessarily the ceiling for the future, you add extra room on top of it.

Let’s look at a numerical example. Say an EA’s backtested max DD works out to ¥30,000. With a safety factor of 2, the funds you need to cover that DD are ¥30,000 x 2 = ¥60,000. Add the minimum required margin (say, ¥10,000) on top of that, and the estimated recommended margin comes to around ¥70,000. The point isn’t “max DD is ¥30,000, so ¥30,000 is enough” — you secure more than double that. The idea is to have enough of a cushion that your account survives even if it drops to the same depth as the past worst case, and to buy the EA time to recover.

This way of thinking is the same order of operations you learned in discretionary trading: “decide where you’ll cut the position before you decide the lot size.” With an EA too, you don’t start from “how much can I make” — you work backward from “how much could I lose in the worst case (max DD)” to decide your capital. Lot size and profit are outputs, not inputs.

Compounding vs. Fixed Lot, and Leverage, Margin Level, Stop-Out

There are broadly two approaches to deciding lot size. Fixed lot trading keeps the lot size constant regardless of account balance, running the same size steadily whether the balance rises or falls. Its behavior is easy to read, which suits a beginner’s verification account. Compounding increases lot size in line with the balance — gains accelerate as the balance grows, but because the lot size is also larger during periods of floating loss, the drawdown deepens proportionally as well. It’s important to understand that compounding amplifies both your “growth power” and your “loss power” at the same time.

Next, let’s go over three terms you absolutely need to know for EA trading.

• Leverage: the multiple of your margin you can trade with. Domestic brokers cap this at 25x, while overseas brokers (HFM, etc.) offer hundreds to thousands of times — a completely different order of magnitude. High leverage lets you hold a large lot size with the same margin, but it also means your margin level can plunge sharply on even a small adverse move.
• Margin level: an indicator of your account’s health, expressed as “equity ÷ required margin x 100(%)”. The more your floating loss grows, the more your equity shrinks, and the lower your margin level falls.
• Stop-out level: the mechanism by which, once your margin level falls below a certain value (20%-100% or so, depending on the broker), positions carrying a floating loss are forcibly closed. Once you hit this, the floating loss at that moment becomes a locked-in realized loss. The “one-shot wipeout” of a nanpin EA usually happens through this stop-out.

In other words, money management is also the work of making sure you have enough margin that, even carrying the maximum floating loss you can foresee, your margin level never falls all the way to the stop-out level. Reverse-engineering recommended margin from max DD exists precisely to protect that margin level.

What Makes Money Management Different for Nanpin (Averaging-Down) EAs

This is where most beginners get it most wrong. Nanpin (averaging-down) and martingale-style EAs look like they have a high win rate, but if you get the money management wrong, they’re about as close to a one-shot wipeout design as it gets. Using the actual settings of MAC v2.0 (a gold-only, SMC-based EA with nanpin), which our lab runs as a verification position, let’s open up how this mechanism works in numbers.

What the Nanpin Spacing, Max Number of Positions, and 1.2x Pyramiding Mean

Nanpin (averaging down) is a technique where, when the market moves against you, you add more buy (or sell) positions to shift your average entry price in a more favorable direction. Since even a small retracement can let you close the whole position in profit, it looks like a high-win-rate “EA that keeps winning.” Breaking down MAC v2.0’s settings gives the following.

• Nanpin spacing of 30 pips: adds one more step of entry for every 30 pips the market moves against the position.
• Up to 15 steps: keeps adding to the position up to 15 times if the adverse move continues.
• 1.2x pyramiding: increases the lot size by 1.2x with each additional step. Starting from 0.1 lot at step 1, it grows to 0.12 at step 2, 0.14 at step 3, and so on.
• 15-pip take-profit, no hard stop-loss: profit-taking is shallow at 15 pips, and stop-losses are left to the EA’s own management rather than a fixed hard stop-loss.

The combination of “shallow take-profit + adding to the position on adverse moves + no fixed stop-loss” racks up small, steady wins as long as things go smoothly. The problem comes when the adverse move doesn’t stop and the number of steps gets deep.

Why “Sizing the Lot Too Big” Leads to a One-Shot Wipeout (In Numbers)

When it stretches out to 15 steps at 1.2x, the lot size at each step stacks up as follows (starting from 0.1 lot at step 1).

Step	Lot at that step (approx.)	Cumulative lot (approx.)
Step 1	0.10	0.10
Step 5	0.21	0.74
Step 10	0.52	2.60
Step 15	1.28	Approx. 7.2

Even starting light at 0.1 lot, by the time it reaches step 15 the cumulative total balloons to about 7.2 lots. The final step alone is 1.28 lots, more than 12 times the first step. And because each step is spaced 30 pips apart, the floating loss stacks up from “small floating losses at the shallow steps” all the way to “enormous floating losses at the deep steps.” A rough estimate of the worst-case scenario where all 15 steps get filled on gold (XAUUSD) puts the floating loss on a scale that could reach anywhere from several hundred thousand to over ¥1,000,000 (this is an estimate that varies with the market, spread, and rate).

This is where the danger of sizing the lot too big really kicks in. If you start step 1 at 0.5 lot (5x) instead of 0.1 lot, both the cumulative total and the floating loss simply scale up 5x. With the exact same market and the exact same margin, just making the lot 5x bigger makes the floating loss you reach 5x bigger too, and the point at which your margin level breaches the stop-out level arrives far sooner. Many people who get wiped out on a nanpin EA aren’t failed by bad logic — they get stopped out before ever reaching the max step because they “sized the lot (initial position size) too big relative to their margin.” That’s why our lab runs MAC v2.0 on a thin setting of “+0.1 lot for every ¥10,000 of capital,” and publishes even the max floating loss on our track-record dashboard.

The Difference Between Domestic and Overseas Brokers (HFM as a Small, High-Risk Verification Position)

Even with the same EA, what money management means changes depending on which broker you run it with. Domestic brokers (JFX, OANDA, etc.) are regulated to a maximum leverage of 25x and have margin-call top-up rules (additional deposits required if a stop-out leaves a loss exceeding the margin), making it a relatively easy environment to apply “textbook” money management. Our lab also keeps its main operations domestic.

Overseas brokers such as HFM, on the other hand, offer leverage of hundreds to thousands of times. This means you can hold a large lot size with a small amount of margin — used well, this lets you verify strategies starting from a small amount, but it also means the margin level can collapse in an instant, and the risk spikes when combined with a nanpin EA. Our lab explicitly positions HFM not as a recommendation for everyone, but as a small-scale, high-risk “verification position.” In fact, we run MAC v2.0 in an HFM copy-trading slot, but this is “verification to observe behavior and floating losses with surplus funds” — not a place to put money you need for living expenses. Be sure to check the specific risks of overseas brokers’ high leverage on the HFM risk disclosure page.

Aspect	Domestic brokers (JFX/OANDA)	Overseas brokers (HFM = verification position)
Max leverage	25x	Hundreds to thousands of times
How easily margin level collapses	Relatively gradual	Prone to sudden collapse
Fit with nanpin EAs	Easier to structure sound money management	Easy to get wiped out from oversized lots
Our lab’s positioning	Main operating base	Small-scale, high-risk verification position

Applying the Discretionary 2% Rule to EAs

The 2% rule you learn in discretionary trading (keeping the loss on any single trade to 1-2% of the account) applies directly to EAs as well. For a fixed-lot, trend-following EA with a fixed stop-loss, all you need to do is set the initial lot size so that “loss width per trade x lot size” stays within 2% of the account. This is the same thinking as the discretionary lot-size formula (lot size = account balance x tolerated % ÷ (stop-loss width in pips x value of 1 pip)).

The tricky part is the nanpin type. Because a nanpin EA has no clear fixed stop-loss, you can’t simply apply “one trade = 2%.” So you flip the framing and manage based on “what percentage of the account the worst-case scenario (the floating loss when the max step is reached) represents.” First estimate what fraction of the account the projected floating loss would be if every step got filled, then adjust your margin and initial lot size so that stays safely short of the stop-out level — that’s how the 2% rule gets reframed for a nanpin EA. It’s safer to remember it as basing things not on “how much you lose per trade,” but on “how much you could end up carrying in total, worst case.”

How Money Management Learned Through Discretionary Trading Helps With EAs

The habit you build up in discretionary trading — deciding your lot size and worst case before you ever enter — sharpens your eye for picking EAs in exactly the same way. If you’ve experienced firsthand how frightening it is when your own discretionary position balloons beyond what you expected, then when you look at a nanpin EA’s track record, you’ll ask “how many lots does the max step end up at?” and “what’s the max floating loss?” before you ever get impressed by a high win rate. The full picture of discretionary money management principles is covered in our piece on stop-losses and money management, and the idea that an EA is simply that logic automated is covered in What Is an EA?. Someone who, in discretionary trading, “decides where to cut and the lot size first” can become, with an EA, someone who “decides max DD and initial lot size first.” It’s the same yardstick.

Let AI Simulate What Happens to Your Own Margin

The final piece of money management is picturing concretely “what’s the worst that could happen with this EA, given the amount I actually deposit.” This is where AI excels. Hand it the EA’s settings (initial lot size, nanpin multiplier, number of steps, spacing, instrument) and your own margin amount, and it can instantly work out — without emotion — the cumulative lot size if every step gets filled, the projected floating loss, and how far the market would need to move against you for your margin level to hit the stop-out threshold. When our lab talks about “translating and publishing difficult verification work,” this is exactly the kind of task we’re turning into numbers on beginners’ behalf.

There is a caveat, though. An AI’s estimate is still a calculation built on assumptions, and the future market, spread, and rate will change. Use the numbers not as an “absolute prediction,” but as material for judging whether your own capital can withstand it. You can find how to read each of the metrics that serve as your yardstick in How to Read an EA’s Track Record, and how to spot dangerous EAs from that same perspective in How to Spot a Dangerous EA.

Summary

The key to surviving long-term with an EA is money management, more than a winning logic. Work backward to set recommended margin as “max DD x safety factor (roughly 2x) + required margin,” and decide your lot size from the worst case, not from potential profit. A nanpin-type EA (MAC v2.0’s 1.2x across 15 steps, 30-pip spacing, no hard stop-loss) looks like it has a high win rate, but if the adverse move doesn’t stop, the cumulative lot balloons to about 7.2 lots, and the margin level approaches the stop-out level right along with the floating loss. The more you oversize your initial lot, the sooner the wipeout arrives. That’s why our lab positions HFM as a small-scale, high-risk verification position, keeps domestic brokers as the main base for operations, and publishes our track record down to the floating loss. Next, head to How to Spot a Dangerous EA to apply this money-management lens, How to Read an EA’s Track Record to go deeper on the numbers, and the HFM risk page to check the assumptions behind overseas brokers.

Frequently Asked Questions

• Q. How should I decide an EA’s recommended margin?
  A. The basic approach is to work backward using “required margin + (max DD amount x safety factor).” A safety factor of around 2x is a reasonable guideline. Since a past max DD isn’t necessarily the ceiling for the future, build in extra room. For more detail on how to read max DD, see How to Read an EA’s Track Record.
• Q. Is a nanpin EA safe as long as I have enough capital?
  A. More capital does make it easier to withstand reaching the max step, but it’s not “absolutely safe.” If the adverse move continues beyond what was assumed, you’ll hit the stop-out even with plenty of margin. Treat a nanpin EA as something you use only in a surplus-funds verification position, after understanding its mechanics and worst-case floating loss in numbers.
• Q. How small should I start my initial lot size?
  A. Rather than starting from an amount, first estimate “what percentage of the account the worst-case scenario would be,” and fit that within a range where the margin level doesn’t fall all the way to the stop-out level. Our lab starts thin during verification, at “+0.1 lot for every ¥10,000 of capital.” Confirm the behavior and floating loss with a small amount before making your decision.