If you're reading this while your Maya account is down and hasn't moved in the right direction for a few weeks, this article is for you specifically. Not for someone shopping around for a trading system. For someone already using one, watching a red number, and wondering whether to trust it or pull the plug.

Here's the honest answer to that question, and the data behind it.

The Discretionary Trader's Real Problem

A discretionary trader hits a bad stretch and has nothing to check it against. They understand the market. They know their own approach, sometimes better than anyone. But when the account is down double digits and hasn't recovered in weeks, there's no record anywhere that says what happens next, because their trades never repeat the same way twice. Every drawdown is a first-time event they're feeling for the first time, with nothing but a gut feeling standing between them and the exit button.

None of this is a knock on discretionary traders as people. Plenty of them genuinely understand the market and have real skill. The problem is structural, not personal. A discretionary approach can't be backtested cleanly because the decisions aren't fully mechanical. Judgment calls shift trade to trade depending on how the trader is feeling that morning, what happened in the last trade, what a headline said an hour ago. That means there's no fixed rule set to run against history, and no clean way to know how this exact approach performed the last five times it hit a drawdown like the current one.

So when the drawdown gets uncomfortable, the only reference point is memory and emotion, and both are unreliable exactly when clear thinking matters most. That's when good traders abandon good approaches, not because the approach stopped working, but because there was no data to check it against and the fear won. It's also when discretionary traders start second-guessing rules they'd normally follow, taking a trade they'd normally skip or skipping one they'd normally take, because the losing stretch has eroded their confidence in their own judgment. The system that felt reliable in a calm market suddenly feels like a coin flip, because in that trader's head, it partly is one.

What a Backtest Actually Is, in Plain Terms

A backtest takes a fixed set of rules, the exact criteria Maya uses to enter a trade, size a position, and exit it, and runs those rules against real historical stock and options price data, one trading day at a time, exactly as if the algorithm had been live the whole time. It's not a simulation of a vague idea. It's the same code, the same thresholds, the same math, applied mechanically to every single day of real market history available.

That matters because it means the backtest can't be flattered. The rules don't know in advance which days are about to go well. They fire the same way regardless, and the historical record shows you exactly what happened as a result, good stretches and bad ones both.

One number that shows up throughout this article is the Sharpe ratio, a standard measure of return relative to how much the account swings up and down to get there. A Sharpe ratio above 1.0 is generally considered solid. Above 2.0 is excellent. It's a way of asking not just "did the strategy make money" but "how smooth or rough was the ride to get there." Every year in Maya's track record below scores above 2.3 on this measure, except one.

Maya has been backtested across 3,308 trades from 2019 through 2025. Every entry rule, every exit rule, every position size is the exact same rule set applied to seven different years of market conditions, not seven different strategies cherry-picked for the years that worked. Every dollar figure and every drawdown percentage below is measured against the same reference account size, $30,000, so the numbers are directly comparable year to year. That's the full table, not a highlight reel:

Notice that four of the seven years have a double-digit drawdown built into them somewhere. That's not a footnote. That's the normal texture of a real multi-year track record, and it's exactly what gets lost when someone shows you only the annual return number without the path it took to get there.

Four Drawdowns, Four Different Years, Same Ending

It's worth walking through what each of these actually looked like while they were happening, because the point isn't the number in the table. It's what it felt like in real time, and what came after.

2019. An 11.6% drawdown over 26 days, concentrated in a single rough patch mid-year. Short and sharp rather than drawn out, the kind of stretch where the account moves fast in the wrong direction and then stabilizes. The full year still closed at $34,809, a 62.0% win rate, and a 2.33 Sharpe ratio.

2020. An 11.1% drawdown over 32 days, sitting right alongside the most volatile market conditions in a decade. The exit rules did what they're designed to do: cap the damage on losing positions instead of letting the account absorb the full force of that volatility. Full year: $17,407, a 56.4% win rate, Sharpe 2.56.

2023. The longest recovery stretch of any winning year. From August to October, the account fell 12.4% and stayed there for 80 days, nearly three full months without a new high. Anyone checking the balance daily through that stretch was watching a red number that simply wasn't moving. It resolved into one of the strongest years in the whole table: $32,983, a 61.0% win rate, Sharpe 2.67.

2025. The deepest drawdown of any winning year in the table. Between February and April, the account fell 14.0% peak to trough, 61 days underwater. Watched day by day instead of year by year, that stretch felt exactly like the start of a bad year. It turned into the best year in the entire seven-year record: $35,551, a 60.9% win rate, Sharpe 2.72.

Four separate years, four different market conditions, and in every one of them the drawdown ended and the strategy kept producing exactly the return profile the backtest said it would. That consistency, not any single year's number, is what the backtest is actually there to prove.

If You're in a Drawdown Right Now, Here's What the Data Actually Says

You don't need to guess whether this is temporary. You can check. Four times in seven years, this exact strategy has fallen 10% or more before recovering into a strong year. Twice, the recovery didn't just erase the loss, it produced one of the best years on record. That's not a promise about what happens this time specifically. It's the actual, documented behavior of the same rule set you're running right now, under conditions that have already repeated multiple times.

The instinct to close everything and step away when the number keeps going the wrong way is completely understandable. It's also, based on this exact track record, usually the moment furthest from the recovery, not closest to further losses. The years that felt worst in real time, 2023 and 2025 in particular, weren't the years that turned out badly. They were two of the best.

And you don't have to take our word for any of this. When you're in doubt, go look at the backtests yourself. Pull up the years above, look at exactly what a rough patch looked like week by week, and follow it through to how it actually came out on the other side. That data is the same data we're looking at when we decide whether a drawdown is normal or a sign something needs to change. There's no reason it should only live in an internal spreadsheet. Checking it yourself, in the moment you're most tempted to stop trusting the process, is a better cure for that doubt than anything we could write here.

A Few Honest Questions Worth Asking

"What if this time is actually different?" Every drawdown feels different while you're in it, because the specific headlines and specific stocks involved are always new. But the mechanism producing the drawdown, a cluster of capped losses in a row, is the same mechanism every time, and the exit rules respond to it the same way every time. The market conditions change. The rule set doesn't.

"Does past performance actually predict future results?" No, and we'd never claim it does. What seven years of backtested history across bull markets, a bear market, a pandemic crash, and multiple corrections does show is how a fixed rule set behaves across a wide range of conditions, which is a far more useful question than whether any single future outcome is guaranteed.

"How do I know the rules haven't quietly changed?" Every change to Maya's entry and exit logic goes through the same backtesting process before it's deployed live, tested against the same seven years of history to confirm it holds up or improves on what's already there. Nothing goes live on a hunch.

See the full seven-year backtested track record: tradewithmaya.com/backtests