Measured Against the Realized Tape
Every convergence read is stamped at the close of the session it was published for, then measured against what the counter actually did over fixed forward windows — and against NIFTY 50 over the identical window, so a rising market does not read as skill.
The Record Starts Here
No read has completed its full forward window yet. Rather than fill this page with a backtest, we publish nothing until reads that were made in public have matured against real prices. A simulated track record is easy to produce and worth nothing; this one accumulates one session at a time and will show the misses alongside the hits.
- Reads are frozen at the session close they were published for. The forward window opens after that, so nothing is measured against information it already had.
- Horizons are fixed at 1, 5, 10, and 20 sessions and were set before any outcome was known — no window gets chosen after the fact for looking better.
- The full sample is published. There is no step where a read can be dropped for having gone the wrong way.
In the meantime, today's board is live on the scanner.
How This Is Measured
- Cohorts
- Counters are grouped by where they ranked on the board that session — top 5, top 10, top 25 of the NIFTY 50 universe. Rank is used rather than an absolute score band because the composite is scaled by the broad-market backdrop, so a fixed cutoff would drift with the market rather than with the counter.
- Entry and window
- Entry is the close of the session the read was published for. The window runs the next 1, 5, 10, or 20 trading sessions. A read enters the aggregate only once its window has fully closed.
- Excess return
- The counter's return minus NIFTY 50's return over the identical window. This is the headline number because raw returns in a rising market say more about the market than about the read.
- Hit rate
- The share of reads whose directional call matched the sign of the realized move. Reads carrying no directional view are measured on return only — they make no claim to score.
- What is not here
- No backtest and no simulation. Nothing on this page was computed over history the engine had already seen. Figures are gross of brokerage, taxes, and slippage, and a sample this size is a description of what happened, not evidence of an edge.
Past outcomes do not indicate future results. Drishtia is not a SEBI-registered Investment Adviser or Research Analyst, and nothing on this page is investment advice or a solicitation to transact. Figures are gross of brokerage, taxes, and slippage. Full disclosure →
Board last refreshed 27 Jul, 01:57 am IST