In proprietary trading, the challengestage is treated as a test of trading skill. Increasingly, it is not. It is atest of whether a trader can perform inside conditions engineered to besurvivable, and those conditions rarely resemble what the same trader will faceonce funded.London's trading industry is coming home!The assumption behind most challengeenvironments is straightforward: a clean simulation is a fair simulation.Instant fills at the requested price. No queue at the touch. No widening spreadwhen volatility hits. That assumptionis the problem.Frictionless Fills Distort What a PassRate Actually MeasuresA challenge environment with no slippageand no order book friction is not neutral. It systematically overstates atrader's edge, because a meaningful share of what separates a profitablestrategy from a losing one lives in the milliseconds between order submissionand fill.Strip that friction out, and the firm isno longer evaluating a trading strategy. It is evaluating a trader's ability tooperate inside an idealised version of the market, then handing that trader afunded account sized on the assumption that performance will carry over. It usually does not.The rationale firms give internally isthat frictionless simulation improves onboarding economics: more passes, morefunded accounts, more perceived value for the challenge fee. That argumentholds only if the firm does not expect the trader to remain profitable oncereal execution conditions apply.The moment a funded strategy meets areal order book, whether through live-routed execution or a simulationcalibrated to actual market depth, the edge that looked stable in the challengebegins leaking on every fill. Firms that builtfunded-stage risk models on challenge-stage performance data areunderpricing the drawdown that follows.Depth of Market Is a Risk Control, Not aTrader's ConvenienceDOM visibility is usually framed as atool for traders: reading intent, spotting resting size, timing entries aroundliquidity. That framing undersells what depth of market actually does for thefirm running the book.A simulated environment thatreconstructs real order book depth, including the layers behind the best bidand offer, forces large or poorly timed orders to walk the book the way theywould against live liquidity. That is where realistic slippage originates.Without a rebuilt order book, a demo server has no mechanism to punish size. Afifty-lot market order fills the same way a one-lot order does, at the sameprice, and the trader does not learn what that order actually costs until realcapital, or a live-mirrored account, exposes it.This matters more for the firm than forthe trader. A firm that cannot model how funded orders interact with real depthcannot forecast its own payout liability with any precision. It is running abook that it cannot price.The pace of consolidation across theindustry over the past two years has not been evenly distributed. Firms thattreated execution realism as a cost centre to minimise are disproportionatelyrepresented among the exits. Firms that treated it as core infrastructure arenot.Slippage Exposure Is the Mechanism, Notthe EducationA common defence of frictionlesschallenge environments is that execution discipline can be taught separately,through risk management content layered on top of a clean simulation.That defence mistakes information forexposure. A trader can be told that fast markets widen spreads and that stoporders can fill several ticks from the trigger price, and still hold nofunctional intuition for it. Intuition is built through repeated exposure toconsequence, not through being told a fact once.A challenge environment that neverproduces a bad fill teaches nothing about bad fills. It teaches the opposite:that execution is reliable. That is the exact lesson that puts a funded traderin the most trouble during their first high-volatility session with realcapital, or real-routed capital, behind them.Firms building this correctly runchallenge and funded environments off the same execution model, so a trader'sevaluation results already reflect the slippage, spread widening, and partialfills they will face after funding. That is a harder and more expensivesimulation to build.It is also the only version where a passrate means anything. A firm that can show a trader passed under conditionsstatistically close to live execution has a defensible claim about thattrader's edge. A firm that cannot is selling a credential, not a riskassessment.Payout Models Inherit Whatever ErrorSits in the Execution ModelEverything downstream in prop firmeconomics, drawdown limits, scaling plans, payout splits, is calibrated againstan assumption about how a funded trader's orders behave in the market.If that assumption is built onfrictionless challenge data, every downstream number inherits the error. Risklimits get sized for conditions the trader will never actually encounter.Scaling plans get built on performance data that does not reproduce under livedepth. Payout ratios drift away from what the firm's actual exposure supports.The gap betweenchallenge fee revenue and funded-trader payout liability is whatdetermines whether a firm survives its own growth. That gap widens fastest atfirms where the execution model used to evaluate traders and the executionmodel used to fund them are not the same model.Slippage and depth of market are nottrading-education topics. They are the inputs a firm's entire risk architectureis built on, whether or not that firm has chosen to model them accurately.The question is not whether a firm'schallenge environment feels realistic to the trader taking it. It is whetherthe firm can demonstrate, with the same rigour it applies to payout ratios anddrawdown limits, that its evaluation execution and its funded execution aredrawn from the same distribution. Most cannot yet answer that question withdata. The firms that can are the ones whose growth will hold up under its ownweight.This article was written by Shervin Arian at www.financemagnates.com.