AI Signal Track Record
What actually happened to every long-term research signal — not a backtest, not a simulation.
Track record
Every outcome published, checked against the Nifty 500
Each long-term research signal is tracked to its target, its stop-loss or the end of its horizon, at the recorded exit price, and set beside what the Nifty 500 did over the same dates. Nothing is removed and no period is hand-picked.
Avg per closed stock result
+1.2%
before brokerage & taxes
vs Nifty 500, same dates
+0.7 pts
+0.9% vs index +0.2% · each stock's own entry → exit dates · 66 of 72 could be measured
Win size vs loss size
1.4×
avg win +12.0% vs avg loss -8.4%
Hit target
47.2%
34 of 72 closed stock results
How to read this: winners have been about 1.4× the size of losers, so the average closed result is +1.2% even though about half reached their target. Measured over each stock's own holding dates, the 66 of 72 results that could be compared averaged +0.9%, against +0.2% for the Nifty 500. Every signal carries a stop-loss; the largest single loss on any published signal was -13.5%. Results swing from month to month, and past outcomes do not predict future ones.
Month by month
| Resolved in | Target / Stop | Hit rate | Avg result | Nifty 500, same dates |
|---|---|---|---|---|
| Oct 2026 | 0 / 4 | 0% | -9.4% | -3.5% |
| Sept 2026 | 1 / 5 | 17% | -5.4% | -1.5% |
| Aug 2026 | 8 / 4 | 67% | +5.9% | +2.3% |
| Jul 2026 | 8 / 6 | 57% | +5.1% | +1.5% |
| Jun 2026 | 7 / 5 | 58% | +3.4% | +0.3% |
| May 2026 | 4 / 12 | 25% | -4.4% | -1.2% |
| Apr 2026 | 6 / 2 | 75% | +5.6% | — |
Latest results
Reached target · 34
- CAPLIPOINT · 3 Sept 2026+8.0%
- JINDALSAW · 26 Aug 2026+14.5%
- YATHARTH · 26 Aug 2026+12.7%
- SMARTWORKS · 24 Aug 2026+9.3%
- WEWORK · 6 Aug 2026+11.7%
- ACE · 5 Aug 2026+14.3%
Long-term research signals (90–365 day horizon) · return = exit price ÷ entry price − 1 · costs not deducted · Nifty 500 = the NSE index level over each result's own entry → exit dates, averaged the same way, dividends excluded; results that cannot be measured (no recorded exit date, or no index data for those dates) are left out of that comparison · not investment advice · updated 19/11/2026
How these numbers are calculated
Each long-term research signal is followed until it reaches its target price, its stop-loss, or the end of its horizon. The result is the recorded exit price divided by the entry price, minus one. Only signals that have resolved are counted; still-open signals are shown as a count, not as a result. Because the scanner re-lists a stock each day it remains a setup, results can be counted per daily signal or once per stock per month — both are shown, with the method stated. Figures exclude brokerage, taxes and slippage. This is research data, not investment advice, and past outcomes do not predict future results.