IPOs and Primary Markets

IPO GMP: The Grey Market Is Right 88% of the Time. That's the Trap

Final-day IPO GMP called 45 of 51 Indian mainboard listings in 2026. But the premium you see while applying is a different number, and it misses by five points.

By The Indian Markets Desk 13 min read

Two numbers that look identical and are not#

Searches for "IPO GMP" in India have averaged 2.24 million a month over the past year and peaked at 3.35 million in July 2026, according to Google Ads data. Searches for the phrase spelled out in full, "grey market premium", ran at a fourteenth of that. Read one way, the gap suggests the number is being consumed as a ticker rather than understood as a concept.

Defenders of it have a good statistic on their side. Run the 2026 Indian mainboard listings through a spreadsheet and the final grey market premium called the direction of the listing correctly 45 times out of 51. That is 88%, which sounds like a solved problem.

It isn't, because of a piece of timing that is rarely checked. The premium scored in those accuracy tables is the one quoted the night before listing, when the issue price is fixed, subscription figures are public and the anchor book is known. The premium you act on is quoted during the bidding window or before it opens, up to a week earlier, when none of that exists. They are different numbers, and the distance between them is what this piece is about.

What the grey market premium actually is#

Grey market premium, or GMP, is the price at which IPO applications and allotted shares change hands privately before a stock lists. It is quoted in rupees over the issue price. Add the two together and you have the market's informal guess at where the share will open.

It has no exchange behind it: no clearing corporation, no settlement guarantee, no SEBI oversight. Deals are verbal commitments between a small circle of dealers and their clients, squared up in cash after listing. A quote of ₹93 tells you nothing about whether it was struck on ten applications or ten thousand, because nobody publishes volumes.

The websites that display GMP are aggregators. They poll dealers, average the replies and publish. Which is why the same issue can show three different premiums on three trackers on the same afternoon, and why, in the last days before Fractal Analytics listed in February 2026, one tracker showed a premium of zero while another showed minus ₹28.

The market exists because of a gap in the calendar. Since 1 December 2023 SEBI has required listing within three working days of an issue closing. Applicants learn within a day or two whether they have shares but cannot sell for another two. Somebody will always offer to buy that risk.

One related term is worth pinning down. The kostak rate is a flat sum paid for an entire application, allotment or not. GMP is the premium on the share itself. Both are informal.

The scoreboard, read properly#

An independent tracker scored 178 IPOs listed through August 2026 and put directional accuracy at 79.67% overall and 88.24% on the mainboard. SME issues did worse, at 73.61%, on premiums averaging 21% against listing gains averaging 22.2%. Working from the 57 mainboard rows in that dataset, here is what the final GMP got right and what it got wrong.

Direction: 45 correct calls out of the 51 issues carrying a non-zero premium.

Size: an average miss of 5.0 percentage points, median 3.9. Only 16 of the 57 landed within two percentage points of the eventual listing price.

How far the final GMP missed the listing price
57 Indian mainboard IPOs, 2026, by size of the gap

Within 2 points   ████████████████        16 issues   28%
2 to 5 points     ███████████████████     19 issues   33%
5 to 10 points    ████████████████        16 issues   28%
Over 10 points    ██████                   6 issues   11%
                  └────┴────┴────┴────┘
                  0    5    10   15   20 issues

Author's calculation from the tracker's mainboard table. The gap is measured in percentage points between the GMP-implied gain and the actual listing gain. Tracker data is unaudited, and no official record of GMP exists anywhere.

The grey market, in short, is a decent compass and a poor tape measure. For someone deciding whether to apply, though, direction was rarely the hard question. Of those same 51 issues, 36 listed above their issue price regardless of what the grey market said. Ignore GMP entirely, assume every IPO lists at a premium, and you would have been right 71% of the time. The premium's 88% buys you about seventeen points over a coin that was already weighted in your favour.

The premium you saw is not the premium that was scored#

Fractal Analytics makes the timing problem concrete. It was a ₹900-a-share issue that bid from 9 to 11 February 2026. Five days before bidding opened, the grey market quoted ₹165, an implied gain of 18.3%. By the first morning of bidding it was ₹14. By the close, ₹5. The night before listing one tracker had it at minus ₹28. The stock opened at ₹876 on the NSE, down 2.67%.

Fractal Analytics: GMP-implied gain during the run-up
(% over the ₹900 issue price, by quote date, 2026)

 4 Feb  ████████████████████  +18.3   IPO Watch
 5 Feb  ███████████           +10.0   IPO Watch
 6 Feb  ████████████          +11.0   IPO Watch
 7 Feb  ██████                 +5.2   IPO Watch
 9 Feb  ██                     +1.6   bidding opens
11 Feb  █                      +0.6   bidding closes
13 Feb  (zero)                  0.0   IPO Watch
15 Feb  ███▌ below zero        -3.1   InvestorGain
16 Feb  ███  below zero        -2.7   actual NSE listing

Two trackers are shown because no single one covers the whole run. IPO Watch carried the premium up to 13 February and never printed a negative figure; InvestorGain carried it to minus ₹28 on 15 February. The disagreement is part of the point.

The accuracy table records that as a hit. Minus 3.1 predicted, minus 2.7 delivered. Anyone who decided to apply on the strength of an 18.3% premium got a rather different experience.

Other issues tell the same story. Lenskart Solutions was quoted at ₹120 on 26 October 2025, a 29.9% premium on the ₹402 issue price. By the day before listing it was ₹15, and the stock opened at ₹395 on the NSE, below its issue price. Tata Capital was quoted at ₹50 in late September 2025 and ₹20 on 3 October, both before bidding opened, then listed at ₹330 against a ₹326 issue price, up 1.23%. SBI Funds Management showed ₹93 on day two of bidding, an implied 16.2% on its ₹574 issue price, settled at ₹65 by listing eve and opened 6.85% up. In each case the early quote was the loudest one, and it faded as real information arrived. All three listed before the 2026 mainboard sample used above, so treat them as illustrations of the mechanism rather than as additions to the count.

The error runs the other way too, which deserves saying plainly. LG Electronics India was quoted at ₹430 on listing eve, implying 37.7% on its ₹1,140 issue price, and opened at ₹1,710.10, up 50%. Groww's parent company showed 3% the night before and debuted 14% higher on the BSE. A low premium is not a warning any more than a high one is a promise.

What the research says, and where it disagrees#

Academic work on pre-listing markets is older than India's GMP trackers, and less flattering than the accuracy tables.

The best-known study, by Cornelli, Goldreich and Ljungqvist on the European when-issued markets, found an asymmetry that ought to sit on every tracker's home page: high grey market prices predict first-day prices very well, low ones do not, and long-run price reversal follows the high ones. Their reading is that grey market prices proxy for small-investor overoptimism. The premium works best, in other words, precisely when it is telling you sentiment has run ahead of value.

Indian evidence splits. A 2025 study of 270 IPOs reported a Spearman correlation of 0.886 between GMP and listing-day performance and called the premium a viable pre-listing predictor. A second paper published the same year ran regressions and found GMP was not a significant predictor of either the listing price or the closing price, with only a small link to subscription levels. Its conclusion was that GMP mostly reflects sentiment, and that institutional participation, foreign flows and market conditions do the real work.

Two studies, opposite findings. Both appeared in 2025 in modest journals and neither has yet been widely cited, so neither should be read as settling anything. The difference probably comes down to which GMP each used and over which stretch of the cycle, though that is interpretation rather than something either paper states. The ambiguity is itself a reason not to treat any single premium as a forecast.

A correct premium still may not pay you#

Suppose the number is accurate. Three things stand between it and your bank account.

You have to get shares. SBI Funds Management closed 41.66 times subscribed, with retail at 3.60 times, so roughly one retail application in four was allotted a lot. A 10% listing gain on nothing is nothing.

You have to sell, and most people do. SEBI's own study of investor behaviour found 54% of shares allotted to investors, excluding anchors, were sold within a week of listing. That describes behaviour rather than recommending it, but it does establish that most IPO applicants are trading, whatever they tell themselves. Gains on listed shares sold within twelve months are short-term, and since July 2024 they have been taxed at a flat 20% under Section 111A, not at slab rates.

And you are working without recourse. The grey market sits outside the regulated perimeter, so a counterparty who walks away is your problem alone. SEBI has been circling the question since early 2025. In January 2025 the then chair, Madhabi Puri Buch, said the regulator was working with the exchanges on a "when listed" facility so that allottees could sell their rights in an organised market. In August 2025 her successor, Tuhin Kanta Pandey, floated a pilot for a regulated venue for pre-IPO trading, subject to disclosures. Neither exists yet. Until one does, millions of people will keep refreshing a quote that no regulator stands behind.

Key takeaways#

  1. Final-day GMP called the direction of 45 of 51 Indian mainboard listings in 2026. Simply assuming every IPO lists at a premium would have been right on 36 of the same 51, or 71%.
  2. On size the average miss was five percentage points, and only 28% of issues landed within two points of the GMP-implied price.
  3. The premium quoted while you are deciding whether to apply is not the premium those accuracy tables score. Fractal Analytics went from an implied 18.3% five days before bidding to a listing 2.67% below its issue price.
  4. GMP undershoots as well as overshoots. LG Electronics India, listing in October 2025, opened 50% up against a listing-eve premium implying 37.7%.
  5. Cornelli, Goldreich and Ljungqvist found that in European when-issued markets high grey market prices predict the debut well and are followed by reversal, because they measure retail enthusiasm rather than value. The Indian evidence is split.

Frequently asked questions#

Is the grey market regulated in India? No. It runs off-exchange, outside SEBI's framework, with no clearing corporation, no settlement guarantee and no grievance mechanism. Deals rest on the word of the dealer, so whether they could be enforced in a court is an open question you would rather not test.

What does GMP stand for? Grey market premium, the unofficial premium over the issue price at which IPO shares or applications change hands before listing.

Does a high GMP mean I should apply? It means dealers expected a positive debut on the day it was quoted. It says nothing about the company's earnings, valuation or risks, all of which are in the offer document filed with SEBI.

Why does GMP change so much during bidding? Because real information arrives. Category-wise subscription figures, the anchor book and the final issue price all land during the window, and the premium adjusts to them. An early quote is a guess made before any of it exists.

Which tracker should I believe? They disagree, sometimes even on the sign. If a site will not say how many dealers it polled and when, treat the quote as hearsay.

Can I trade in the grey market myself? Access runs through dealers, mostly in a few cities, on verbal terms. There is no regulated route and nothing to fall back on if the other side defaults.

Will SEBI regulate it? Two SEBI chairs have publicly discussed a regulated pre-listing venue, in January and August 2025. As of 11 September 2026 no such platform has launched.

Glossary#

Grey market premium (GMP). The unofficial price above the issue price at which IPO shares or applications trade privately before listing.

Kostak rate. A fixed amount paid to buy an entire IPO application, whether or not it receives an allotment.

Subject to sauda. A grey market deal that stands only if the application in question receives an allotment.

Listing gain. The difference between the issue price and the price at which a share opens on its first trading day, as a percentage.

Anchor investor. A qualified institutional buyer allotted shares one working day before the issue opens, locked in for 30 days on half the allotment and 90 days on the rest under Schedule XIII of the SEBI ICDR Regulations.

T+3 listing. SEBI's rule, in force since 1 December 2023, that shares must list within three working days of an issue closing.

When-issued market. A regulated forward market in shares that have been allotted but not yet listed. Several European markets have run one. India does not.

Directional accuracy. How often a signal gets the sign right, positive or negative, regardless of how badly it misjudges the size.

References#

  1. SEBI, study on investor behaviour in IPOs, press release 37/2024, 2 September 2024
  2. Business Standard, SEBI cuts the IPO listing timeline to T+3 from 1 December 2023, 9 August 2023
  3. Business Standard, SEBI to start pre-listing trading to curb grey market activity, Madhabi Puri Buch, 21 January 2025
  4. Business Standard, SEBI may introduce a regulated platform for pre-IPO firms, Tuhin Kanta Pandey, 21 August 2025
  5. Mr Money Frugal, GMP versus listing accuracy tracker, 178 IPOs to August 2026, updated 10 September 2026 (independent tracker, unaudited; the 57-row mainboard table is the basis for this article's own calculations)
  6. IPO Watch, Fractal Analytics IPO GMP history, and InvestorGain, Fractal Analytics GMP and listing
  7. PL Capital, Fractal Analytics shares debut at a 3% discount to the ₹900 issue price, 16 February 2026
  8. IPO Watch, Lenskart Solutions IPO GMP history, and Chittorgarh, Lenskart Solutions IPO listing data
  9. IPO Watch, Tata Capital IPO GMP history, and Chittorgarh, Tata Capital IPO listing data
  10. InvestorGain, LG Electronics India IPO GMP, and Chittorgarh, LG Electronics India IPO listing data
  11. IPO Watch, Groww IPO GMP history, and Business Standard, Billionbrains Garage Ventures market debut, 12 November 2025
  12. India Infoline, SBI Funds Management IPO, day-2 GMP of ₹93, 15 July 2026, and Groww, SBI Funds Management lists at a 6.85% premium, 21 July 2026
  13. Cornelli, Goldreich and Ljungqvist, Investor Sentiment and Pre-IPO Markets, Journal of Finance, volume 61, 2006
  14. Vigneshwar and others, relationship between grey market premium and listing-day performance, 270 IPOs, International Journal for Multidisciplinary Research, 2025
  15. Oraon, evaluating the influence of grey market premium on IPO subscription and listing performance, Indian Journal of Research in Capital Markets, 2025
  16. Income Tax Department, Section 111A, tax on short-term capital gains on listed securities
  17. Corporate Professionals, the IPO lock-in framework under the SEBI ICDR Regulations
  18. Google Ads search volumes for India, retrieved via DataForSEO, 11 September 2026

This article is journalism, not investment advice. It reports what was quoted, what was disclosed and what happened. Nothing here is a recommendation to buy, sell or hold any security. Consult a SEBI-registered investment adviser before acting.

Related coverage