Is IPO GMP reliable?

GMP is widely followed. How close does it actually get to the listing price?

What an older sample showed

Before launching, we looked at about 300 past Indian IPOs from one public GMP history, comparing the final GMP with the actual listing price:

• The GMP-implied gain was within ±10 percentage points of the actual gain about 7 times in 10, and within ±5 points about 4 times in 10.
• On average the GMP overstated the listing gain slightly, by around 2 to 3 points, and about 6 in 10 IPOs listed below the price the GMP implied.
• When the GMP was positive, the share listed above the issue price about 8 times in 10.
• When the GMP was zero, the share more often listed below the issue price than above it.

That was a one-off check on one set of historical numbers. It is useful background, not a promise about any future IPO.

Our own record

From September 2026 we publish our own estimate, freeze it before each listing, and score it against the actual listing price. The track record is updated after every listing and counts only IPOs we were actually forecasting while they were open.

Reasons GMP misses

• Market mood can change sharply between the IPO closing and listing.
• Grey-market volumes are small, so a handful of deals set the number.
• Many trackers repeat each other, so errors spread instead of cancelling out.
• SME IPOs are thinner and more volatile than mainboard IPOs.