A row lands in the feed. Long, an asset you recognise, a timeframe, an entry price, a source. You go to place it at your broker, the order ticket asks you for a take profit and a stop loss, and there is nothing in the row to put in either box. Most people fill them in from habit at that point, usually with a percentage they use for everything, and that single reflex does more damage to signal following than bad entries ever do.
Worth being precise about what is actually in front of you first, because it shapes everything after. In the captures I am working from, the trade table carries Asset, Strategy, Direction, Status, Timeframe, Order, Time in UTC, Entry, Current, PnL, Run-up, Drawdown, Source and Duration. There is no take profit column and no stop loss column among them. So an empty exit is not a row that failed to populate. It is the normal shape of the data, and if you are waiting for a version of the feed that hands you both brackets, plan as though it will not arrive.
The row tells you where to get in, not where to get out
The feed describes itself as an aggregated queue across every engine, and lists what feeds it. Momentum, reversal and outperformer engines, plus whale, insider, political and prediction market trackers. Those things do not have the same relationship with an exit, and lumping them together is what produces the habit of stamping the same bracket on all of them.
Split them into two groups and the empty cells stop looking like an omission.
- Rows from a strategy that manages its own exit. A momentum or reversal strategy has a closing rule built into it. The exit is a future event computed off future bars, so there is no static price to display today. It has not been left blank, it has not been decided yet.
- Rows that are an event rather than a trade. An insider filing, a congressional disclosure, a whale wallet changing position. There is no exit rule anywhere in the world for these, because the row reports something that happened rather than a plan with two ends.
The practical difference is large. For the first group, your bracket is a safety net that should almost never be touched, and if it is being touched regularly it is set wrong. For the second group, your bracket is the entire exit plan, because nothing behind the row is going to tell you when to leave.
What the strategy's own risk numbers look like when they are published
Sometimes the closing side does get published, and when it does it is worth reading carefully because it tells you how much room these strategies actually need. One closed trade posted to the platform's trade broadcast was a long in spot silver on a four hour momentum strategy. It exited at 64.27 for a gain of 8.02 percent after holding for 28 days and 4 hours. The two numbers that matter for our purposes sat underneath. Run-up of 8.28 percent, drawdown of 8.00 percent, and a stop loss listed at 11.29 percent.
Sit with that. A trade that finished up 8 percent spent part of its life 8 percent underwater, and the strategy's own stop was set nearly a third wider again than the worst point the trade reached. The strategy performance block alongside it showed an average hold of 19 days and 1 hour across 34 trades.
Now imagine that row arriving in your feed with empty exit cells, and you stamping your usual 3 percent stop on it out of habit. You would have been taken out early, at a loss, from a trade that went on to make 8 percent, and you would have concluded the signal was bad. It was not. The bracket was.
Size off the strategy's stop instead of your comfort
The fix is an old one and it is the only one that works. When you cannot make the stop narrower without breaking the trade, you make the position smaller. The stop width is a property of the signal. The dollar risk is a property of your account. Those are two separate dials and most people jam them into one.
The arithmetic on a 5,000 dollar account, risking 1 percent, which is 50 dollars.
| Stop width you use | Position size that keeps risk at 50 dollars | What happens on the silver trade above |
|---|---|---|
| 3 percent | 1,667 dollars | Stopped out during an 8 percent drawdown, then watched it finish up 8 percent |
| 8 percent | 625 dollars | Survives by a hair, which is not a margin worth relying on |
| 11.29 percent, the strategy's own | 443 dollars | The trade is allowed to do the thing the strategy expected it to do |
Notice that the honest answer produces a much smaller position than the one your instinct wanted. That is not a downside of the method, it is the method telling you the truth about how much of this particular signal you can afford. A 443 dollar position that runs to completion is worth more to you than a 1,667 dollar position that gets shaken out of the same trade, and it is worth a great deal more than the version where you widen the stop but keep the big size and take a 188 dollar hit on a 1 percent risk budget.
If the position that comes out of that arithmetic is too small to be worth the commission, you have learned something useful. That signal is not tradeable at your account size, and no amount of bracket tuning changes it.
Brackets that do not argue with the logic
Once you are sized correctly, attaching the bracket at the broker is mechanical. Four rules cover it.
- Set the stop wider than the strategy's own tolerance, not narrower. If the strategy publishes a stop, use it or go slightly beyond it. If it does not, use the largest drawdown you have seen on closed rows from that same source as a floor, and accept that this estimate improves as you collect more of them.
- Do not set a fixed take profit on a trend following row. The engine stats at capture showed an average run-up of 230.4 percent per strategy against an average drawdown of 11.3 percent, which is the classic shape of a system that pays for a long tail of small losses with a handful of very large runs. A tidy two-to-one target amputates exactly the trades that fund the method.
- Add a time stop instead. The published average hold on that silver strategy was 19 days. A position from the same family sitting at nothing after three times its typical hold is not a winner being patient, it is capital doing nothing, and a calendar exit costs you far less than a price exit does.
- For event rows, write the exit before you enter, because nothing else will. Decide whether you are trading the reaction over days or the thesis over months, and put a date on it in the same ticket.
One caution about the headline numbers on these panels. The 96.6 percent average win rate on the counter strip is described alongside a count of 350 trades backtested across strategies. Backtested statistics are a description of a rule applied to history, not a forecast of your next fill, and the gap between the two is exactly the bracket you are about to attach. Treat those figures as a guide to how much room the logic needs, which is what this article uses them for, and not as an expectation about outcomes.
The row you should genuinely skip is the one where all of this collapses. If the source publishes no exit logic, no closed history you can inspect, and no stop, then you are not following a signal, you are taking a tip and inventing the risk management afterwards. Sizing cannot rescue that, and the empty cells in that case are telling you something true.