Ask a group of Indian retail investors why they applied to an IPO and a large share of the answers will reduce to one number: the GMP. It travels through messaging groups and aggregator sites with the authority of a market price, it is quoted to the rupee, and it updates through the day. It looks exactly like data.
It is not data in any sense a careful investor should accept. The grey market premium is an unofficial quote from an unregulated, off-exchange market with no clearing, no settlement guarantee, no disclosure obligations, no audit trail and no recourse. This guide explains what it genuinely is, why the number is so easy to move, and why it is one of the weakest inputs you could put at the centre of a decision — which is precisely where most people put it.
What GMP actually is
Between the day an IPO is priced and the day it lists, there is a window in which people want exposure to the outcome and cannot get it on an exchange. An informal dealer market grew up to fill that window. In it, participants quote a premium over the issue price at which they will buy or sell an entitlement to the shares before listing. That quoted premium is the grey market premium.
There is no venue. There is no order book you can inspect, no volume figure you can verify, no counterparty guarantee, no regulator overseeing conduct, and no contract you could enforce if the other side simply did not settle. The whole structure runs on relationships between a small number of operators, mostly concentrated in a few trading hubs, and the numbers that reach you have been relayed through several hands before they appear on a website next to a chart.
The three things quoted in the grey market
It helps to know that the grey market is not one thing but three overlapping trades, because the distinction explains why the quotes move for reasons that have nothing to do with the company.
- Grey market premium — a price for allotted-but-unlisted shares, quoted as a premium over the issue price. This is the number that circulates most widely.
- Kostak — a fixed price paid for an entire IPO application, regardless of whether it receives an allotment. The buyer is effectively purchasing your lottery ticket and the seller locks in a small certain sum instead of an uncertain one.
- Subject to sauda — a price agreed for an application that pays out only if the application is actually allotted. It sits between the other two, and it is the trade most sensitive to expected allotment odds rather than to expected value.
Notice what all three are about. They are trades in the distribution of allotment and listing-day price, priced off expected demand. None of them involve anybody forming a view on the business, its cash flows, or its valuation. The grey market is not a valuation mechanism that happens to be informal — it is a betting market on a short-dated event.
Why it is unofficial, and what that means for you
This is the part that gets waved away. The grey market operates entirely outside the regulated securities market. That is not a technicality; it has four concrete consequences.
- No settlement guarantee. On an exchange, a clearing corporation stands between you and your counterparty. Here there is nothing. If the other side does not pay, you have an informal understanding and no mechanism.
- No price integrity. Exchange prices come from a matched, audited, surveilled order book. A grey market quote comes from whoever is quoting it, and nobody has an obligation to trade at it.
- No disclosure. There is no published volume, no participant register, no record of whether the quote reflects ten lakh rupees of interest or ten crore. The number arrives with no denominator.
- No recourse. Because the activity sits outside the regulated market, the investor-protection machinery you would normally rely on — grievance redressal, arbitration, regulatory action — does not extend to it.
Why the number is so easy to move
Set aside the regulatory argument for a moment and consider only the market microstructure, because that alone should end the matter. The grey market for a given issue is tiny relative to the issue itself. An offering raising thousands of crores can have a grey market in which the total genuine two-way interest is a rounding error by comparison.
In a market that thin, the quoted price is not the output of broad price discovery. It is the output of a handful of dealers, and it can be moved by a very small amount of activity — or by no activity at all, since a quote nobody trades on is still a quote that gets published and relayed.
Then add the incentives. Everybody with an economic interest in an issue subscribing well — and there are several such parties around every offering — benefits from a high circulating premium, because it drives retail subscription, which drives the appearance of demand, which supports the listing. Nobody with that interest benefits from a low one. A market this thin, with incentives this one-sided and no surveillance at all, is close to the textbook definition of a number you should not trust.
Judge a position on your own numbers
AIVITTA reads your live holdings read-only and gives you the risk, concentration and return picture that actually applies to your portfolio — no sentiment, no unofficial quotes.
The reflexivity problem
Here is the loop that makes GMP feel predictive without being informative. A premium is quoted. Retail investors see it and apply in larger numbers. Subscription figures rise. Rising subscription figures are reported as evidence of strong demand. Dealers, seeing the subscription figures, quote a higher premium. More retail investors apply.
At no point in that circuit does anyone learn anything about the company. The premium and the subscription number are each other's cause. When GMP appears to have predicted a strong listing, part of what happened is that GMP helped produce the demand that produced the strong listing — a self-fulfilling mechanism that works right up until the moment it does not, and then reverses just as quickly, because nothing underneath it was ever load-bearing.
This is also why GMP collapses so fast on bad news or a weak broader market. A number produced by sentiment has no anchor to fall back on. A valuation, however rough, at least gives you something to disagree with.
How often it misleads
Be fair to it: GMP is not random. It is a reasonable thermometer of the sentiment that will be in the room on listing day, and there is a loose relationship between a high premium and a positive listing pop. If your only question is whether listing day will be excited, GMP carries some information — the way a crowd outside a shop tells you something about the queue, and nothing about the product.
Two well-documented patterns should temper any comfort you take from that. First, IPOs across markets tend on average to list above their issue price — so a lot of what GMP appears to predict is simply the base rate of listing-day underpricing that would have shown up anyway. Second, and far more important, the average long-run performance of newly listed companies after the first few months has historically been weaker than the broader market. A high premium says nothing about that second phase at all, which is the phase that determines whether you actually made money.
The failure mode is specific and common: a heavily hyped issue lists at a large premium, the premium is treated as validation of the business, the investor holds rather than sells, and the stock spends the following quarters drifting below the issue price as the first real results arrive. GMP was directionally right about listing day and catastrophically misleading about the investment.
| Question | Does GMP help? |
|---|---|
| Will listing day be excited? | Loosely, as a sentiment thermometer |
| Is the issue price reasonable? | No — it contains no valuation content |
| Is the business any good? | No |
| Where will the stock be in a year? | No |
| Am I likely to be allotted? | No — that is the subscription figure, by category |
What to look at instead
Everything GMP is standing in for is available to you in documents nobody reads. The offer document contains the audited financial history, the objects of the issue, how much of the offering is fresh capital versus existing shareholders selling out, the risk-factor section, related-party transactions, and a peer-comparison table with valuation multiples for listed companies in the same business. That last table is the single most useful page in the document and takes two minutes to read.
If you want a structured way to work through it, should you invest in IPOs sets out the framework and the red flags. If you want the vocabulary first, the IPO terms glossary covers it. And the subscription figures — the real, exchange-published ones, read by category — are the correct input for the only question GMP is often used to answer badly, which is whether you are likely to be allotted at all. How allotment works covers that.
The uncomfortable summary is that GMP is popular because it is easy. It is a single number, it updates constantly, it requires no reading and no judgement, and it feels like being informed. Doing the actual work is slower and produces less confident conclusions — which is usually a sign that the conclusions are honest. AIVITTA is an analytics and education platform, not an advisor; we do not publish premium data, and we do not recommend applying to any issue.