Industry
Designing prop firm evaluations that hold up
An evaluation programme is a product with a payout obligation attached. Designed carefully it is a sustainable business; designed by copying whichever competitor is loudest, it becomes a slow transfer of money from the firm to a small number of traders who understood the rules better than the firm did.
The rules are the product
Traders do not choose a programme on its marketing. They choose on target, drawdown, time limit and payout terms, and they compare those numbers across firms with considerable care. Every one of those parameters is simultaneously a marketing decision and a risk decision.
Loosening a rule to look competitive is easy and immediately effective at attracting sign-ups. It is also the most common way a programme becomes unprofitable without anybody noticing for a quarter.
Drawdown definition does more work than the number
Whether drawdown is measured on balance or equity, and whether it trails the peak or stays static, changes the difficulty of a programme far more than shifting the headline percentage by a point or two.
It is also the parameter traders scrutinise most closely, and the one where ambiguous wording generates the most disputes. Whatever is chosen, it needs to be stated precisely enough that two people reading it reach the same conclusion.
Balance or equity — and whether open positions count
Trailing or static — and if trailing, what it trails from
Daily and overall limits, and how they interact
What happens exactly at the limit, not near it
Consistency rules are where disputes come from
Rules against single-day outsized gains, over-concentration or specific strategies exist for good reasons, but they are frequently written loosely and applied after the fact. That combination reliably produces a trader who passed on the published rules and failed on an unpublished interpretation.
If a rule cannot be evaluated automatically from the trade record, it is not yet written clearly enough to enforce.
Payout mechanics decide sustainability
Split, frequency and the conditions attached to the first payout determine whether the programme survives a good month for its traders. A generous split with frequent payouts and no scaling is attractive and difficult to sustain.
The honest version of this is to model the programme against a distribution of outcomes rather than an average one. Averages hide the case that actually threatens a programme: a small number of traders performing very well at the same time.
Publish it, then leave it alone
Changing rules mid-evaluation, even in the trader's favour, damages trust disproportionately. It signals that the parameters are negotiable, which invites every subsequent outcome to be contested.
Version the rules, apply the version in force when the evaluation started, and make it visible which version a given trader is on. It costs a little to build and removes an entire category of argument.
In short
What to take away
Drawdown definition matters more than the drawdown number
A rule that cannot be evaluated from the trade record cannot be enforced fairly
Model payouts against a distribution, not an average
Version the rules and hold each trader to the version they started on
Read next
Model the programme before you launch it
Rules, limits and payout mechanics can be configured and tested against real trade records rather than assumed.