A backtest tests strategy rules against historical data. Practice trading rehearses signals and order mechanics without real money. Live trading adds actual fills, costs and emotional pressure.
Each stage answers a different question. Trouble starts when a trader asks one stage to prove everything.
What a backtest can establish
A careful backtest can estimate:
- how often the written setup appeared
- win rate and average win or loss under stated rules
- expectancy before and after modelled costs
- drawdown and losing-streak history
- sensitivity to different periods or market conditions
It can reject an idea that fails even under clean historical assumptions. It can also show whether one attractive chart was typical or an outlier.
A backtest cannot establish that the future will repeat the sample. It also cannot prove that you can identify and execute every signal in real time.
The backtest trader does not hesitate
Historical rules take every qualifying trade. They do not skip the next signal after two losses, cut a winner because open profit feels fragile or move a stop because the loss is uncomfortable.
That consistency is useful. It isolates the strategy from the trader. But it also means the test has not measured your execution.
Keep the distinction sharp:
Backtest result = strategy rules applied to historical assumptions
Live result = strategy rules plus actual execution, costs, data and behaviour
The gap between them can come from either side. A bad live result is not automatic proof of poor discipline. The historical test may contain look-ahead bias, survivorship bias, unrealistic fills or an overfit rule.
Practice trading has a real job
Paper or practice trading removes financial loss, so it cannot reproduce the full pressure of real money. That limitation does not make it useless.
Use practice trading to rehearse:
- recognising the setup without seeing the future candles
- entering orders in the right direction and quantity
- placing stops and targets correctly
- handling partial exits
- recording the plan before the outcome
- following the routine at the time of day it must occur
Practice trading is where you discover that a rule is ambiguous, an order type is unfamiliar or a strategy requires more screen time than your life permits. Those are valuable failures because they cost no capital.
Treat the fills as approximate. A simulated order may not reproduce queue position, spread changes, slippage or a gap through the stop.
A staged readiness framework
Move through four stages with written exit criteria.
Stage 1: define the strategy
Write the market, setup, entry, stop, target, size rule and conditions that cancel a trade. If two people would read the rule differently, it is not ready to test.
Stage 2: backtest the rules
Use a sample that includes different market conditions. Model costs and avoid changing parameters merely to improve the past curve. Record the number of trades, expectancy and drawdown.
Advance only if the rules show a plausible edge after costs and the result is not carried by a tiny number of unexplained outliers.
Stage 3: practice trading
Run the strategy forward without money at risk. Score whether signals were identified on time, orders were entered correctly and records were completed.
Advance only when the process is repeatable. A profitable practice result with frequent rule breaks is not readiness.
Stage 4: small live test
Use a position size where the full planned loss is financially tolerable. The goal is to compare live fills and behaviour with the earlier evidence, not to generate income.
Increase size only after enough trades show that costs, slippage and execution remain within the limits you set. There is no universal trade count that proves readiness.
Worked comparison: one strategy across three stages
Assume a hypothetical pullback strategy risks 1R per trade.
| Measure over 30 trades | Backtest | Practice trading | Small live test |
|---|---|---|---|
| Win rate | 47% | 43% | 40% |
| Average winner | +2.0R | +1.8R | +1.6R |
| Average loser | -1.0R | -1.0R | -1.1R |
| Average costs | -0.05R | estimated -0.05R | -0.12R |
| Rule adherence | assumed 100% | 90% | 80% |
The table does not prove why performance declined, but it directs the review.
The live average loss is worse than the planned -1R, so inspect gaps, slippage and stop changes. Average winners also fell, so check early exits. Costs more than doubled the model. Rule adherence dropped as real money was introduced.
Do not immediately optimise the entry. First reconcile the fills and decisions. Then rerun the historical result with the observed cost assumptions. If the edge disappears, the strategy needs work. If it survives but live adherence remains low, execution needs work.
Compare like with like
Use the same written rule version, market and measurement window where possible. Record every qualifying forward signal, including the ones you skip, so practice and live samples do not keep only convenient trades.
Compare:
- expected entry versus actual fill
- planned stop versus actual exit
- modelled versus realised costs
- eligible signals versus trades taken
- planned and realised R-multiples
- rule adherence by stage
This is more useful than asking whether the live curve looks similar after five trades.
Match the testing stage to the trading style
An intraday strategy needs practice under market-hour time pressure. A swing strategy needs realistic handling of overnight gaps and a routine that fits around other commitments.
Use the day trading versus swing trading comparison to decide which workload and risk pattern you are preparing to test. The strategy may look attractive while the routine remains impossible for you to execute.
Keep the plan and deviation together
Record the signal, intended order, actual fill, planned stop, actual exit and reason for any deviation. That creates a fair comparison between the strategy you tested and the trade you took.
Swingfolio keeps the trade plan and outcome in the same review. It cannot make a historical edge valid or remove real-money pressure, but it can make the execution gap visible.
Start the 30-day trial to compare planned rules, live fills and execution notes in Swingfolio before increasing size.
General information only. Not financial advice. Trading involves significant risk of loss. Past performance is not indicative of future results. Always do your own research and consider seeking advice from a licensed financial advisor.