Documentation habit for when execution goes wrong -- before you need it.
I do not enjoy disputes. I do enjoy having a folder ready so I am not reconstructing a bad fill from memory three weeks later.
Minimum pack I keep
Screenshot of order ticket / reject dialogue with platform time
Journal row: symbol, side, size, order type, intended price, fill, spread at click
Account statement export for that day
Short note: what I did next (resubmit, cancel, flat elsewhere)
Filename pattern with date and symbol. One folder per month. If I ever switch accounts, the same pack shows whether the problem was my timing, the session, or a pattern worth raising calmly with support.
This is not about picking fights. It is about not arguing anecdotes. Most "broker issues" in my history were me clicking into thin tape -- and the folder made that obvious.
What do you archive on a messy fill day so Sunday-you is not guessing?
I do not enjoy broker disputes. I do enjoy having a pack ready so I am not reconstructing a bad fill from memory. Minimum from this desk: ticket/reject screenshot with platform time, journal row (symbol, side, size, intended vs filled), and periodic deal-history exports.
Rule: after any weird fill or reject cluster, capture immediately — not after the session when the UI has moved on. That habit also informs account-switch decisions with evidence instead of anger.
Prop platforms deserve the same logging; eligibility arguments go better with timestamps.
I store the pack outside the trading VPS as well. When the machine is the problem, evidence trapped on the machine is not evidence you can use.
Monthly I verify I can actually open the archive quickly. A folder you cannot find during a dispute is decorative.
What is the one field you still forget to capture in the moment — server time, or the partial-fill residual?
For me it is the price on screen at the moment I clicked. The fill stays in the history forever, but what I saw a second earlier vanishes, and that is the half of the argument support always asks about.
A screen recorder with a rolling replay buffer fixed it. Mine keeps the last 90 seconds in memory and one hotkey saves it to the monthly folder, so a bad fill costs me a keypress instead of a reconstruction.
The MT5 terminal log is worth adding to the pack too. It records each request with the time to the millisecond and the execution delay, a line like a market buy done in 312 ms, and the file for that day sits in the logs folder as plain text.
One habit makes single fills easier to judge. I keep a running average of slippage per side over the last 200 trades, so a 1.5 pip slip on one ticket gets compared with a normal figure of 0.2 instead of with my mood.
Support also replies faster when the request names ticket numbers and server times rather than a description. Has anyone had a broker send over their own server log for a disputed ticket?
Partial fills are the field I forgot most often, to answer PropScalpDesk's question properly. On a 2 lot market order that filled 1.4 and left the rest, my journal showed one ticket while the statement showed two deals, so the residual looked like a separate trade.
Now the journal row has a filled size column next to the requested size. If the two differ, I note the deal number for each piece before doing anything else.
The other thing I added since my last post is a reference price from a second source. A demo terminal from a different broker records ticks for the pairs I trade, and in a dispute it shows whether the market itself traded at my fill price.
Twice that settled the argument in the broker's favour, which was useful in its own way. Both times the spike was real, and I stopped chasing a refund I had no case for.
Time limits are worth knowing too. Several brokers only review execution complaints raised within a few days, so I send the pack the same week rather than waiting for the monthly review.
Does anyone keep the reference feed on a separate machine? Mine shares the VPS, so an outage there would take out both records at once.
Every regulated broker is required to publish an order execution policy, yet most traders never open it. The document describes how orders are routed, which venues are used, how slippage is treated, and under what conditions orders may be rejected or delayed.
Several clauses deserve close attention. Language about last look, price improvement, and the broker's right to adjust prices in abnormal conditions reveals how the execution model works. Statements about whether the broker acts as principal or agent show who sits on the other side of trades.
For scalpers, the sections on minimum holding times, order frequency, and trading strategies are especially relevant. Some policies reserve the right to cancel trades that exploit pricing delays. Others explicitly allow scalping. Comparing the policies of two or three candidate brokers side by side often exposes meaningful differences that never appear in marketing materials. A few hours spent reading and annotating these documents can prevent months of frustration with an unsuitable venue.