Backtest the rule before you trade it.

Six rules, one stock at a time

Pick a stock and a rule — SMA crossover, EMA crossover, RSI mean reversion, MACD crossover, Bollinger Bands, or buy-and-hold as the benchmark — set its parameters, and Tradune replays it across that stock’s daily history. You see the equity curve against buy-and-hold, the drawdown chart, the price chart with every buy and sell marked, and the full trade log.

Read the drawdowns, not just the returns

The headline return is the least interesting number. The one that decides whether you’d have actually stuck with a strategy is the worst drawdown — the peak-to-trough stretch you’d have had to sit through. A backtest that hides that is selling you a fantasy.

Know what a backtest can’t tell you

A backtest is a hypothesis test, not a promise. Past behavior is not future behavior, and any rule can be curve-fit to look great in hindsight. Tradune shows you the historical record honestly — the interpretation, and the humility, are yours to bring.

How does the backtester simulate trades?

1. 1

   Daily bars for one stock

   The test runs on daily open, high, low and close prices for the symbol you pick — 500 trading days by default, up to about ten years, or a date range you set. It needs at least 50 days of data.

2. 2

   One of six rules

   SMA or EMA crossover (defaults 10/50 and 9/21 days) buys when the fast average crosses above the slow one and sells on the cross back. RSI mean reversion (14 days) buys below 30 and sells above 70. MACD uses 12/26 with a 9-day signal line. Bollinger Bands (20 days, 2 standard deviations) buy at the lower band and sell at the upper. Buy-and-hold buys on the first day.

3. 3

   All in, all out, at the close

   A buy signal invests all available cash at that day’s close; a sell signal sells the whole position at that day’s close. Fills happen on the same bar that produced the signal, which is slightly optimistic — in practice you would trade after seeing it. There is no partial sizing, shorting or leverage.

4. 4

   Costs on every fill

   Each buy and sell pays 0.3% of the amount traded unless you enter a flat fee per trade. The buy-and-hold comparison line is drawn without any cost, so the comparison does not flatter the rule.

5. 5

   Statistics that respect the sample

   You get final capital, total return, win rate, profit factor, average win and loss, total costs and maximum drawdown against the running peak. The return is annualized only when the run has at least 250 daily returns or spans about a year; Sharpe, Sortino and Calmar (at an assumed 4.2% risk-free rate by default) are withheld below that. A parameter sweep re-runs the rule across up to 50 values of one setting — useful for seeing how fragile a result is, and an easy way to curve-fit if you pick the best one.

What does one round trip cost?

A single buy-then-sell with the default 0.3% cost, on $10,000 of starting capital. Read what backtesting can (and can’t) tell you before trusting any result.

The trade log records this trade as +20% (sell price over buy price) and its dollar profit as $1,958.11 after the sell cost, so read both columns. A rule that trades often pays the cost every time; a result that only beats buy-and-hold before costs does not beat it.