How accurate has GMP actually been?

For every issue that lists, we take the grey market premium as bidding closed, convert it to an implied percentage gain on the issue price, and compare it with what the stock actually opened at. Both numbers are frozen at the time — the GMP is the consensus a reader would have seen before applying, and the listing price is the opening price from the exchange’s own daily bhavcopy.

Three measures are reported, because a single accuracy number would hide more than it shows:

• Directional hit rate — did GMP get the sign right? A premium that listed at a discount is a miss regardless of size, and this is the measure that matters most to someone deciding whether to apply.
• Within 5 and 10 percentage points — how often the magnitude was close, not just the direction.
• Mean absolute error — the average size of the miss, in percentage points, ignoring direction.

Why we publish the misses

Every page on this site carries a warning that grey market premium is unofficial and not recognised by SEBI. That warning is worth nothing unless it is testable, so this page keeps the record — including the issues where GMP pointed the wrong way entirely. Nothing is excluded for making the number look bad, and an issue with no GMP at the close is left out rather than scored as zero.

The sample here is small and grows one issue at a time. Read the individual rows rather than the headline percentages until it is large enough to mean something.