20 Years, One Rule: What I Learned From the Accounts I Blew Up

Hello. My name is Jun, and I have been trading for about twenty years.

I should tell you upfront that most of those years were not good ones. I want to write about that, because I think the part nobody publishes is the part that actually teaches you something.

The accounts I lost

My first serious loss came from a system that looked perfect. Smooth equity curve, high win rate, small drawdown. I found it, I tested it, I believed it. It worked for about four months.

Then it did not, and it took most of the account with it in under two weeks.

What I did not understand at the time was that the smoothness had a price. The system was quietly recovering its losses by increasing position size — so every small loss looked like it disappeared, right until the losses stopped being small. The curve had not been smooth because the strategy was good. It had been smooth because the losses were being deferred.

I repeated a version of that mistake more than once. Different systems, same structure. It took me longer than I would like to admit to see the pattern.

What actually changed things

The turning point was not finding a better strategy. It was deciding what I would never do again, and then refusing to break that list even when breaking it would have made the numbers look better.

My list is short:

• No grid. Stacking positions against the market is borrowing from your future self at a terrible rate.
• No martingale. Under any name. Recovery mode, adaptive sizing, smart scaling — if lot size goes up after a loss, it is the same thing.
• No averaging into losses. A losing position does not become a better idea because it got cheaper.
• A stop loss on every single position. Not a mental one. Not a wide one I intend to move. A real one, set at entry.

Every one of those rules costs me money in the backtest. Every one of them makes my equity curve less pretty than the one that sold me on that first system. I keep them anyway, because the alternative is a curve that looks wonderful until the day it does not exist.

What I would tell someone starting out

If you are looking at automated systems right now, I am not going to tell you which one to buy. But there are three questions I wish someone had made me ask fifteen years ago.

How many trades is this result based on? A few hundred trades can produce almost any result by luck. If you cannot see the trade count, you cannot judge the number next to it.

What happens to the lot size after a losing trade? Ask directly. If the answer is vague, or if it involves a word you have to look up, assume the worst. An honest seller can answer this in one sentence.

What deposit was this tested on? Drawdown is a percentage, and percentages can be manufactured by changing the denominator. A 4% drawdown on $100,000 and a 40% drawdown on $10,000 can be the exact same trades.

None of these questions require you to be technical. They just require you to ask before you buy, rather than after.

I wrote about this in more detail here.

Why I still do this

I run a restaurant. Trading has never been the thing that pays my bills, and I think that is part of why I am still here — I was never forced to take a trade to make rent.

What keeps me building systems is narrower than getting rich. It is the problem itself: gold is fast, it is violent, it punishes hesitation and it punishes greed equally, and building something that survives it for twenty months without cheating is genuinely hard. When a test comes back with a profit factor under two and a drawdown north of twenty percent, and I know every one of those numbers is real, that is more satisfying to me than a screenshot showing four thousand percent would be.

I am not going to promise you that my system will make you money. I do not know that, and anyone who tells you they do is selling you something other than software.

What I can tell you is that it will not lie to you about what it is doing. The lot size will not grow because you lost. Every position will have a stop. And when I publish a number, it will be the number I actually got.

That is a smaller promise than most. I think it is the only one worth making.

I build XAUUSD systems under the name GoldDragon. If you found this useful, the numbers I mentioned are all published on the product page. Have a look here.

Thanks for reading. Trade carefully out there.