Last Updated: 2026-10-26
A profitable trade, a profitable account and a reliable income source are different claims. None follows from choosing a bot, keeping a journal or reaching a particular balance. Start with complete records, costs and a defined comparison period. If those are missing, the result is unknown, not an estimated monthly salary.
What a loss study can actually tell you
BIS Bulletin 69, published February 20, 2023, estimates Bitcoin losses using app-download timing and assumed purchases through December 2022. Its simulations are not complete realized trading histories. They support caution about that sample and period, not a universal failure rate for every strategy or a forecast of your return.
When someone cites “most traders lose,” ask for the instrument, population, dates, treatment of inactive accounts, open positions and costs. Evidence about equity futures or a Bitcoin purchase simulation should not silently become a statistic about current crypto-bot users. Conversely, a selection of successful accounts says little about everyone who started.
Reconcile account equity first
For a single reporting currency, including consistently valued open positions:
Period result = ending equity − starting equity − deposits + withdrawals
Synthetic example: start with $1,000, deposit $300, withdraw $100, and finish with $1,250. The result is $50, not the $250 increase in displayed balance. If fees and funding have already reduced account equity, do not subtract them again. Software paid outside the account still belongs in a separate net-cost calculation. Moving assets between your own included accounts is a transfer, not new profit.
Keep realized profit, unrealized change and missing valuations separately. A balance containing a token priced from a stale quote is not the same as cash available to withdraw. Cash-flow timing also matters: $50 divided by the initial $1,000 is not a properly cash-flow-adjusted return when money moved during the period.
Win rate needs payoff and cost context
Take 100 completed trades from a hypothetical rule: 55 wins averaging $8 and 45 losses averaging $10. The sample's gross result is:
55 × $8 − 45 × $10 = −$10
A 55% win rate did not make it profitable. If execution costs average $0.40 per completed trade, another $40 takes the result to −$50. These are teaching inputs, not expected win rates or reliable distributions.
With fixed average win W, average loss L and average cost C per trade, the arithmetic break-even win probability is (L + C) / (W + L). Here that is 10.4 / 18 ≈ 57.78%. Changing position sizes, losses in the tail, correlated trades and uncertain estimates make this a limited model. Do not size an account from the point estimate alone.
Use turnover, not trade count, for fees
If total filled notional is $20,000 at 0.1%, commission is $20. On a $1,000 account that is 2% of equity; on a $5,000 account it is 0.4% only if turnover stays $20,000. Scale turnover fivefold with the larger account and the proportional drag is again 2%. More capital does not remove a percentage fee.
A limit order can execute immediately against an existing quote; it is not automatically a maker fill. Use actual fills and the relevant product/tier. Keep spread and slippage accounting consistent: do not deduct a cost twice when it is already reflected in the recorded execution price.
Compare the result with the work and exposure
Record hours spent, maximum observed drawdown, unresolved positions and time unavailable for withdrawals. Compare with a baseline chosen before seeing the result and matched for dates, cash flows and costs. A passive asset holding is not a risk-free baseline; “do not trade” is also an available decision.
Twenty hours spent for $30 net is $1.50 per hour before tax, but a single period cannot forecast future hourly pay. A 40% loss needs a 66.67% gain on the remaining balance to recover; recovery is neither impossible nor guaranteed. Review out-of-sample observations and changed market conditions before attributing a result to skill.
The CFTC risk advisory explains why leverage and market conditions can amplify losses. A stop order is not guaranteed loss insurance. Use the capital worksheet to separate essential money from an optional experiment. There is no required deposit, tool subscription or calendar-based progression here.
Scope and disclosure: Sources checked October 26, 2026. Examples are synthetic and before tax unless stated; they are not investment advice, performance records or income expectations. AITR may earn affiliate commissions elsewhere. No tool or exchange is recommended as a route to profitability.