Skip to main content

​
What backtesting is

Backtesting simulates how a strategy would have behaved on historical market data. In Trinigence, backtesting answers:
  • Did the strategy take trades as expected?
  • What would the performance and risk look like historically?
  • How stable is behavior across time?
Backtesting is a simulation, not a guarantee of future performance.

​
What Trinigence backtesting simulates

A Trinigence backtest simulates, candle by candle:
  1. evaluating strategy logic on each candle close
  2. opening trades when entry conditions are met
  3. closing trades when exit conditions are met
  4. applying risk rules (stop loss, take profit, limits)
  5. tracking performance metrics and trade history
The same inputs always produce the same results.

​
What data is used

Backtests use historical candle data:
  • Open
  • High
  • Low
  • Close
  • Volume (when available and used)
Results are always dependent on:
  • the selected market (symbol)
  • the selected timeframe
  • the available data range

​
Candle-close evaluation

By default, strategy logic is evaluated on candle close. That means:
  • entries trigger only if conditions are true at close
  • exits trigger only if conditions are true at close
  • crossovers trigger only if the crossover is present at close
Intrabar behavior is not assumed unless explicitly supported.

​
What the backtest does not guarantee

Backtests cannot fully replicate live conditions such as:
  • slippage variations
  • real-time liquidity changes
  • exchange-specific execution quirks
  • sudden outages and operational constraints
The purpose is to understand logical behavior and historical distribution, not to predict the future.

​
Assumptions you should be aware of

Trinigence backtests operate under assumptions that keep behavior deterministic. Common assumptions:
  • strategy logic is evaluated deterministically per candle
  • the same data produces the same trades
  • no hidden discretionary decisions exist
When relevant, Trinigence surfaces assumptions used to complete a strategy.

What Trinigence fills automatically

See what the system infers to complete strategies.

​
Why results can look “wrong”

If a backtest doesn’t match your expectation, it usually comes from one of these:
Indicators and logic behave differently across timeframes. A 15m strategy is not a faster 1h strategy.
Filters can quietly eliminate most entries.
If you expected intrabar triggers, candle-close evaluation will look delayed.
Many strategies are controlled more by exits and risk rules than by entries.

​
How to validate a backtest

Best practice validation flow:
1

Start with a simple baseline

Use a minimal strategy with clear entries and exits.
2

Inspect trade history

Look at individual trades before trusting metrics.
3

Add complexity one layer at a time

Add filters, schedules, and extra conditions gradually.
4

Compare across time ranges

Validate behavior in different market regimes.

​
What to look at first

When reviewing results, start with:
  • trade count and frequency
  • win rate vs profit factor
  • max drawdown and drawdown duration
  • equity curve shape
  • biggest wins/losses

Backtesting metrics overview

Learn what each metric means and how to use it.

Metrics overview

Interpret performance and risk metrics.

Trade history & logs

Debug strategy behavior.

Common backtest pitfalls

Avoid misinterpretation.

Strategy structure

Understand what is being tested.
Backtesting answers: “What happened historically under clear rules?”
It does not answer: “What will happen tomorrow?”