Mistakes I made with daily loss limits

[LondonScalper]

I've had a daily loss limit for most of my career. I've also managed to get it wrong in more ways than I'd have thought possible.

Setting it as a fixed amount, not a percentage. In 2008 my limit was £500. When the account grew, £500 became too small and I started ignoring it. When the account shrank, it was too large to protect anything.

Measuring it on closed trades only. I'd hit the limit, then have an open position that was also losing, and tell myself it didn't count yet. Open losses count.

Not including costs. My limit was based on P&L before commission. On a busy day, commission added another 15 to 20% to the actual loss.

Treating it as a target. Strange, but true. On some days, being down 0.8% with a limit of 1% made me feel I had 0.2% to "use". I'd take trades I wouldn't otherwise have taken.

Having no weekly limit alongside the daily one. Four days at the daily limit is a terrible week, and I had nothing to stop it.

Currently: 1% daily, including open positions and costs, enforced by a script. 2.5% weekly, enforced by me, which is the weak point.

[Shadow Trader]

The "limit as a target" one is the mistake I see most, and I made it too. What fixed it for me was adding a second line well before the limit. At 0.5% down I stop taking new setups, and can only manage what's open. The 1% is then something I almost never reach.

On the weekly limit being the weak point, I'd say there's a simple test for that. Look back at the last few months and count how often you'd have hit 2.5% for the week. If it's rarely, the script isn't worth the effort. If it's every few weeks, it's worth automating like the daily one.

Another one I'd add to your list: resetting the daily limit at midnight server time rather than at the start of your trading day. If you trade Asia and London, the reset can split one bad session into two "days", and neither looks like a breach.

Including costs is something I'd push everyone to do. Commission on a busy scalping day can easily be a fifth of the loss, as you say.

Does your script close open positions at the limit, or just block new orders?

[Shadow Trader]

One more mistake I'd add to the list is measuring the daily limit only from the start of day balance. If I'm up 0.8% by 10:00 and then give it all back plus another 0.9%, the limit says I'm fine, but I've had a 1.7% swing in one session.

So I run a giveback rule alongside it. Once the day is up more than 0.5%, losing half of the peak gain ends the day, whatever the limit says.

Prop traders get a version of this forced on them through trailing drawdown, but on a personal account nobody makes you do it. Of all the changes I've made to my limits, this one cut my worst days the most.

The other point is testing the script itself. Mine silently stopped working once after the broker added a suffix to the symbol names, and I only found out by reading the log a week later.

Now I run a forced test on demo on the first Monday of each month, setting the limit to 0.05% and checking that it triggers. Has yours ever failed without you noticing?