Backtesting
Applying a set of trading rules to historical data to see how they would have performed.
How it is identified
Test: rules are fully specified in advance, applied without hindsight, and results are net of realistic transaction costs and slippage
Unit
qualitative
In depth
A backtest is only as good as its assumptions, and the three that most often invalidate one are survivorship bias, look-ahead bias and unrealistic fills. Survivorship bias comes from testing on today's index constituents, which excludes every company that failed; look-ahead bias comes from using information not available at the decision point, such as an index change applied before it was announced. Fills at the closing price are optimistic, since a real order pays spread and impact. A backtest that has been adjusted after seeing its results is no longer a test of anything.
Worked example
A strategy showing 22% annual returns before costs, trading 50 round trips a year at a realistic 0.35% each, gives up 50 x 0.35% = 17.5%. The net 4.5% is the number that would have been earned, and it is below the index.
Illustrative figures, chosen so the arithmetic is easy to follow. Not a live price and not a valuation of any company.
Educational reference only
This entry explains what “Backtesting” means. It is not investment advice and not a recommendation to buy or sell any security. Any numbers above are illustrative, not live prices, and nothing here predicts price direction or rates a stock. Consider your own circumstances and consult a SEBI-registered investment adviser before acting.