Why Your Stock Screener Backtest Results Are Wrong (And How to Fix Them)

You spent hours building the perfect screener. RSI below 30, weekly volume above 1 million, price above the 50 SMA. You run a backtest and the numbers look incredible: 80% win rate over six months. You start trading with real money, confident in your strategy.

Three weeks later, you are staring at losses. The strategy that looked bulletproof on paper is bleeding money in real markets.

What went wrong? The answer might shock you: the backtest was lying to you all along.

The Hidden Problem With Stock Screener Backtests

Most stock screener platforms in India have a critical flaw that very few traders know about. When you run a backtest, the platform recalculates weekly and monthly indicators using the complete week or month data, even for days in the middle of that period when the complete data did not exist yet.

This means the backtest shows you results based on information that was not available on those days. It is like predicting yesterday's cricket match score using today's scorecard. The numbers look right, but the timing is completely wrong.

Example 1: The Weekly Volume Trap

Let us walk through a real-world scenario to make this crystal clear.

Suppose your screener has one condition: Weekly Volume > 10,00,000 (10 lakh).

Here is how volume accumulates through the week for a particular stock:

What should the backtest show? The stock did NOT qualify on Monday, Tuesday, and Wednesday because the weekly volume had not crossed 10 lakh yet. It only qualified on Thursday and Friday. A correct backtest should show exactly this — the stock appearing only on Thursday and Friday.

What do most screeners show? Since the final weekly volume (15 lakh) exceeds 10 lakh, these platforms mark the stock as qualifying for ALL five days of the week. Monday, Tuesday, Wednesday, Thursday, Friday — all green. The stock appears to have been in your screener the entire week.

This is fundamentally wrong. If you were actually trading on Monday based on this screener, the stock would NOT have appeared in your list. But the backtest tells you it did. You think your strategy caught it on Monday. In reality, you would have missed it completely until Thursday.

Example 2: The Disappearing Stock Problem

Here is another scenario that is equally dangerous. Your condition: Weekly Close > 50.

What should the backtest show? The stock qualified from Monday to Thursday because the running weekly close was above 50 on each of those days. On Friday, it dropped below 50 and no longer qualifies.

What do most screeners show? Since Friday is the final day of the week and the weekly close is 45 (below 50), these platforms remove the stock from ALL days of the week — including Monday through Thursday when it genuinely qualified. In the backtest, the stock completely disappears from the entire week as if it never existed.

Think about what this means. You might have actually seen this stock in your screener on Monday. You might have traded it and made a profit by Wednesday. But the backtest says it was never there. Your strategy looks worse than it actually was.

The Real Cost to Traders

These are not edge cases. This happens with every single weekly or monthly indicator: volume, close price, moving averages, RSI calculated on weekly candles, and more. The impact compounds:

• False confidence: Stocks appear for more days than they should, inflating win rates and making strategies look better than they are.
• Hidden winners: Stocks get removed from days they genuinely qualified, making strategies look worse and causing you to abandon profitable approaches.
• Wrong entry timing: You think a stock appeared on Monday but it actually only appeared on Thursday. Your entire entry strategy is built on false data.
• Compounding errors: In a screener with multiple conditions mixing daily and weekly timeframes, the errors multiply. A 3-condition screener might have completely different results than what really happened.

How BacktestKit Fixes This Problem

BacktestKit takes a fundamentally different approach. Instead of recalculating weekly and monthly indicators retroactively, BacktestKit freezes indicator values as they were on each trading day.

On Monday, when the cumulative weekly volume was 1,50,000, BacktestKit records that exact value. It does not go back and overwrite it with Friday's final weekly volume. The same applies to every indicator on every timeframe.

When you run a backtest on BacktestKit, what you see is exactly what you would have seen if you had opened the screener on that day. No more, no less. The backtest becomes a true time machine — showing you the exact reality of each trading day.

This is not a small improvement. For many strategies, the difference between a manipulated backtest and an accurate one is the difference between profit and loss.

Frequently Asked Questions

Why do other screeners show wrong backtest results?

Most screeners calculate weekly and monthly indicators using the complete period data and apply that value to every day within the period. This means mid-week data gets overwritten with end-of-week values, creating a mismatch between what you would have seen in real-time and what the backtest shows.

Does this affect daily-only screeners?

If your screener uses only daily indicators (like daily RSI, daily close, daily volume), this specific problem does not apply. It only affects screeners that use weekly or monthly timeframe conditions. However, many effective strategies combine multiple timeframes, which is where the problem becomes serious.

How much difference does accurate backtesting make?

It depends on the strategy, but we have seen cases where a screener shows a 75% win rate on other platforms and only 55% on BacktestKit — because the manipulated backtest was falsely adding stocks on days they should not have appeared. The 55% is the real number, and building your strategy on reality is always better than building it on fiction.