Educational Simulation Archive
Costs and Assumptions in Simulation
REVIEW GUIDE · EDUCATIONAL JOURNALING

Costs and Assumptions in Simulation

Every simulator runs on a cost model — even when the model is 'everything is free and instant.' A serious journal writes that model down.

Strictly educational archive. Simulated analysis and hypothetical models only.
HYPOTHETICAL SIMULATEDReview Guide 09 / 10
Costs and Assumptions in Simulation

Every Simulator Has a Cost Model

Paper platforms differ in what they charge a simulated order. Some apply realistic commission schedules, some apply none, and many apply an implicit cost of zero by filling at a single midpoint price. Zero is still a cost model — just an unrealistic one.

Before journaling on any platform, find its fill and cost documentation. What price does it use for market orders? Does it model spread, commission, borrow fees, or delay? Whatever the answer, it belongs in the journal header for every session.

Commission, Spread, and Delay Assumptions

When the platform does not model a cost, enter your own assumption in the 'Assumed fees and spread' field. A flat synthetic commission per order, an observed or conservative spread, and a stated execution delay turn an abstract fill into a more defensible hypothetical.

  • Assumed commission per simulated order (flat or per unit).
  • Assumed or observed spread at order time.
  • Modeled or assumed execution delay.
  • Whether partial fills were possible in the model.

Writing Down What the Model Omits

The most dangerous field in any simulation is the one that does not exist: no market impact for size, no queue priority, no gap risk between sessions, no rejection of stale quotes. These omissions are not bugs — they are the simplifications that make a teaching platform usable.

Journals stay honest by listing the omissions as 'interpretation limits.' An entry that says 'model cannot show queue depth or impact' is more rigorous than a live-looking fill with no caveats.

Stress-Testing Your Own Assumptions

A useful review exercise: take one completed Review Sheet and rerun the arithmetic with worse assumptions — double the spread, add a delay, cut the fill to half quantity. If the lesson of the entry survives harsher assumptions, it was robust. If it depended on the friendly defaults, the entry was reading the model, not the market.

Log both runs. The spread between the friendly and the stressed version measures how much of your 'result' was borrowed from the simulator's generosity.

Cost & Assumption Ledger

Cost TypeSimulator DefaultAssumption to WriteField
CommissionUsually zeroPer-side fee you assumeFees assumed: 0.05%
SpreadOften none or fixedQuoted gap per entrySpread: 0.1% assumed
SlippageZero or ignoredDeviation from quoted priceSlip: 1 tick assumed
LatencyInstant fillsDelay between intent and fillLatency: 1 tick assumed
Margin/limitsUnlimited virtual capitalReal constraint you simulateConstraint: none modeled
Costs unrecorded are costs assumed to be zero — the most expensive default in any journal.

Journal Checklist

  • Are all cost assumptions written as fields, not defaults?
  • Is the simulator's cost model (none, fixed, modeled) named in the header?
  • Are assumed fees/spread applied consistently across entries?
  • Is 'cost not modeled' recorded where the sim provides nothing?
  • Does the entry separate gross result from net-of-assumptions result?

Review Note

Treat every simulator default as an assumption you chose, not a fact you observed. Chosen assumptions can be reviewed; invisible ones cannot.

Frequently Asked Questions

Because 'free' fills teach wrong economics. An assumed-cost field shows whether the idea survives real-world friction — even if the number is an assumption.

Whatever your platform omits silently — usually spread and slippage. Name the omission and write an assumption; that's more honest than the default zero.

Consistent beats precise. A documented 0.1% assumption used everywhere is reviewable; a 'roughly right' memory is not.

Continue Exploring