IPO grey-market premium: what GMP actually predicts (and what it doesn't)

Every time a hyped IPO opens, the same number gets passed around WhatsApp groups and Telegram channels like it is gospel: the grey-market premium, or GMP. "This one's running at ₹250 GMP" — meaning, supposedly, that the stock will list ₹250 above its issue price and you will bank an easy pop on day one. People apply because the GMP is high, and stay away when it is low. Whole application decisions get made on a single unofficial figure that no exchange publishes and no regulator stands behind.

This is the GMP-chasing trap, and it burns retail investors on a regular basis. The number feels like a forecast — a live, ticking prediction of your listing gain. It is not. It is a snapshot of sentiment in an unregulated side-market, set by a loose network of dealers, changing without notice, and impossible to independently verify. Sometimes it lines up with what happens on listing day. Increasingly, it does not.

The seduction is easy to understand. IPO applications are a lottery — you may not get allotted at all — and even if you do, you have to decide before listing whether the pop will be worth locking up your money and taking the risk. In that fog of uncertainty, a single hard-looking number that seems to say "this will list at ₹650" is enormously comforting. It converts a genuinely hard, multi-variable judgement into one glance at a WhatsApp forward. That comfort is exactly the problem: it substitutes a soft, gameable sentiment reading for the harder work of actually assessing the business, its price, and who is really buying. The people who lose money on IPOs rarely lose it because they did too much homework. They lose it because a big GMP told them the homework was unnecessary.

This piece walks through what GMP actually is, why it exists at all, and — the part that matters for your money — what two decades of Indian listing-day data and a run of 2025 main-board IPOs tell us about whether a high GMP really predicts a high pop. There is a genuine debate here, and I will give you both sides fairly, because the honest answer is more useful than a slogan. Then I will hand you a checklist for using GMP the way it should be used: as one weak signal among several, never as the deciding one.

What GMP actually is

The grey-market premium is the extra amount, over and above an IPO's issue price, that buyers are willing to pay for shares (or for the right to shares) in an unofficial, off-exchange market that operates in the window between an IPO closing and the stock listing on the exchange.

If an IPO is priced at ₹500 and the GMP is ₹150, it means people in this grey market are paying a ₹150 premium — betting the stock will list around ₹650. The premium is a live gauge of pre-listing sentiment and demand: how badly the crowd wants in before official trading begins.

That is the whole of it. GMP is a sentiment thermometer, not an order book. It does not measure how much stock institutions have actually committed to buy at what price. It measures how much a fragmented pool of grey-market participants feel like paying, right now, based on subscription buzz, brokerage chatter, and each other's quotes.

It helps to be precise about the difference between sentiment and demand, because the two get blurred and that blur is where the danger lives. When a qualified institutional buyer commits money at the issue price, that is demand backed by capital and a considered valuation — an actual order, at an actual price, that will actually be filled. When the grey market quotes a ₹150 premium, that is a mood: a reading of how much a small group of participants believe the next person will pay, which is itself a guess about how much the person after that will pay. Sentiment can be self-referential and self-reinforcing in a way that a real order book cannot. Everyone quoting a high premium partly because everyone else is quoting a high premium is a recipe for a number that looks confident and means very little. That is why GMP can be sky-high right up to the morning a stock lists flat or below issue — the mood was real, the demand behind it was not.

The jargon: kostak and subject-to-sauda

Two terms come up constantly in grey-market talk, and both describe deals struck before you even know whether you will get an allotment.

• Kostak is a fixed price paid to buy your IPO application outright, regardless of whether it is allotted shares or not. If you sell your application at a kostak of ₹800, you pocket ₹800 the moment the deal is done — even if you end up getting zero shares in the allotment. The buyer is paying for the chance your application gets shares.

• Subject-to-sauda (often just "subject to") is a deal that only settles if your application actually receives an allotment. If you get shares, the deal is honoured at the agreed rate; if you get nothing, the deal is void and no money changes hands. It removes the "did I even get allotted" gamble from the price.

Both are informal, verbal, trust-based arrangements. There is no contract you can enforce, no exchange clearing the trade, no regulator you can complain to if the other side walks away. Which brings us to the uncomfortable part.

It is worth sitting with what "trust-based" really means here. In a normal market trade, the exchange and its clearing corporation stand between you and your counterparty — if the other side vanishes, the system still settles you. In the grey market there is no such backstop. A kostak or subject-to-sauda deal holds only because both parties, and the dealer who introduced them, expect to keep doing business with each other. Reputation is the settlement mechanism. That works, mostly, within a tight circle — and it collapses at exactly the moments you would most want protection, when a listing goes badly and someone is staring at a loss they would rather not honour. For a retail investor with no standing in that circle, the grey market is not somewhere you have rights. It is somewhere you have hopes. Keep that framing in mind every time the GMP number tempts you to treat it as solid.

Why GMP exists — and why it is unofficial and manipulable

GMP exists because there is real demand to trade IPO exposure before the stock lists. An IPO closes, allotments take a few days, listing takes a few more — and in that gap, people who want in (or want out) before day one create a market. Dealers match them. A premium emerges. It is a natural response to a timing gap.

But look at what this market is, structurally:

• It is not SEBI- or exchange-backed. No regulator oversees it, no exchange records it, no clearing house guarantees it. It sits entirely outside the formal market.
• The trades are legally unenforceable. These are informal, off-the-books arrangements. If a counterparty defaults, you have no legal recourse. The whole thing runs on reputation within a small circle.
• Prices are set by a loose, fragmented dealer network. This is not a formal "cartel" with a control room — it is a scattered, regional set of operators and intermediaries quoting to each other. There is no single official GMP; there are quotes, and websites that aggregate and report them.
• It is opaque and manipulable. Because there is no transparent order book and no audit trail, quotes can be talked up or down. A handful of participants posting aggressive numbers can move the reported GMP on a thinly-traded issue. The figure can change without notice, and nobody has to justify it.

So the number you see quoted as "the GMP" is really a reported, aggregated, unverifiable sentiment reading from an informal market. That is not a reason to sneer at it — sentiment is real information. It is a reason to treat it as exactly what it is: soft, gameable, and unofficial. It reflects mood, not the institutional order-book demand that actually drives a listing.

The deep question: does a high GMP predict a big pop?

Here is where people want a clean yes or no. The honest answer is messier, and the mess is the point.

The 2025 main-board evidence: a widening gap

Through 2025, several high-profile main-board IPOs showed the reported GMP and the actual listing outcome pulling apart — in both directions. The mainstream consensus coming out of these cases is that GMP chatter and real outcomes are diverging more visibly than before. A muted GMP preceded a strong listing in one case; an elevated GMP preceded a flat or negative one in others.

I want to be careful here, because this is directional, anecdotal evidence, not a measured statistical finding. A handful of cases does not prove GMP has "lost all predictive power." What they do show is that treating GMP as a reliable one-to-one forecast of your listing gain is a mistake — the recent record has too many misses for that.

All the GMP figures below are as reported in the unofficial grey market and cannot be independently verified. The listing outcomes are as reported by mainstream financial coverage.

IPO	Reported GMP signal	Actual listing outcome	Verdict
Groww	Muted — around ₹3 premium	Opened +12%, closed day one +31%	GMP badly under-predicted; strong pop despite weak grey-market signal
Tata Capital	Implied roughly 6–7% listing gain	Opened only +1.2%	GMP over-predicted; the pop mostly did not show up
Lenskart	Elevated GMP	Listed below issue price	GMP wrong on direction; a loss, not a pop
Studds	Elevated GMP	Listed below issue price	GMP wrong on direction; a loss, not a pop

Read down that table. In one case a near-zero GMP preceded a 31% first-day gain. In another, a 6–7% implied gain collapsed to barely over 1%. In two more, high GMPs preceded listings below the issue price — the opposite of what a premium is supposed to signal. If GMP were a dependable forecast, none of these should happen. They happened, in a single year, on the main board, to well-covered issues.

Look closely at the two failure modes on display, because they are different and both matter. Groww is a magnitude miss in the investor's favour — the grey market was near-silent, yet the stock ran hard, meaning anyone who skipped the IPO because the GMP looked uninspiring left a large gain on the table. Tata Capital is a magnitude miss in the other direction — the premium implied a respectable pop, and the actual open barely cleared water, meaning anyone sizing their expectation off the GMP was set up for disappointment. Lenskart and Studds are the most damning kind: direction misses. A premium is, by definition, a bet that the stock lists above issue. Both listed below it. The grey market did not merely get the size wrong; it pointed the wrong way entirely, turning an expected gain into a real loss for anyone who applied on the strength of the premium. A signal that can be wrong on magnitude in both directions and wrong on direction is not a signal you can lean your capital against.

The 20-year underpricing data: pops are real, but average out over a noisy distribution

Step back from GMP for a moment and look at the underlying phenomenon it is trying to call: IPO underpricing, the tendency for a newly listed stock to jump on day one because it was priced below what the market will bear.

The foundational dataset here is Jay Ritter's work at the University of Florida, the reference point for IPO underpricing research worldwide. For India, academic analysis of 235 IPOs from 1997 to 2008 found a mean first-day listing gain of about 14.45% — roughly 13.04% once you adjust for market movement over the same window. So yes: on average, Indian IPOs have historically popped on listing.

But an average is a dangerous thing to build a strategy on, because this one varies enormously over time and is driven by sentiment, not mechanics. The same research shows the underpricing was strongly time-varying — the hallmark of what economists call "hot issue markets." During the cold stretch of 1998–2002, IPOs were on average priced too high — they overpriced, meaning buyers who chased them lost on listing. Then in the hot market of 2007, underpricing peaked at around 22%. Same market, wildly different outcomes, depending purely on where sentiment sat in the cycle.

That is the crucial lesson the headline "+14% average pop" hides: the average is the middle of a very wide, mood-driven distribution. In hot markets you get big pops; in cold markets you get losses; and the average across a full cycle masks how many individual IPOs disappoint. GMP, being a pure sentiment gauge, runs hottest exactly when the market is most euphoric — which is precisely when underpricing is most likely to reverse. The signal is loudest when it is least trustworthy.

There is a subtler statistical point worth spelling out for anyone tempted to apply on the average. A ~14% mean first-day return, adjusted to ~13% for market movement, is a real and economically meaningful edge if you could capture it evenly across every IPO. But you cannot. You are allotted a lottery-limited slice of the hot issues everyone wants, while the cold ones you can get filled on are disproportionately the ones that disappoint. The mean sits above zero because a minority of blockbuster listings drag it up; the median experience of an ordinary applicant, spread across the full range of issues they can actually get into, is far less flattering. And the 1998–2002 stretch — when Indian IPOs on average overpriced, handing losses to buyers on listing — is the reminder that "IPOs go up on day one" is not a law of nature. It is a cycle-dependent tendency that has, for years at a time, run in reverse. GMP tells you nothing about where in that cycle you are standing; if anything, it whispers most persuasively at the top.

The counterpoint: an academic study calls GMP "viable, with caution"

To be fair to GMP, it is not worthless, and I would be misleading you if I implied the research universally dismisses it. At least one academic study of 270 main-board IPOs examined GMP as a short-term predictor of listing performance and concluded it is a "viable" predictor — but one that must be used with caution. In other words, there is a statistical relationship: on the whole, higher grey-market premiums have tended to associate with better short-term listing outcomes across a large sample. GMP carries genuine information about crowd demand, and crowd demand does influence the listing.

So we have two things that are both true:

• Across a large sample, GMP shows a real, if noisy, statistical association with listing performance — enough for a careful researcher to call it "viable with caution."
• On individual IPOs, and visibly through 2025, GMP misses often enough — in both magnitude and direction — that betting a single application on it is reckless.

There is no contradiction. A signal can be informative in aggregate and unreliable case by case. That is exactly the situation with GMP. It nudges the odds; it does not call the outcome. Anyone telling you GMP "always works" is selling something; anyone telling you it is "pure noise" is ignoring the large-sample evidence. The truth sits in the uncomfortable middle: useful as a weak input, dangerous as a decision rule.

The distinction between aggregate and case-by-case is the single most important idea in this whole piece, so let me make it concrete. Imagine a signal that is right 60% of the time across a thousand IPOs. In a spreadsheet, that is a real edge — a researcher would rightly call it "viable." But you are not applying to a thousand IPOs; you are deciding on this one, and on this one the signal is wrong four times in ten. Worse, when it is wrong on an IPO you applied to because the signal was strong, the loss is concentrated and personal, not averaged away across a portfolio of a thousand bets. Retail investors experience IPOs one lottery ticket at a time, which is precisely the setting in which an "aggregate-viable" signal is most likely to hurt. The academic verdict and the lived retail experience are both correct; they simply describe different things. The researcher is describing the sample. You are living a sample size of one.

SME IPO froth and the 2024 SEBI tightening

Nowhere does the GMP trap bite harder than in SME IPOs — the small and medium enterprise segment. These issues are smaller, less liquid, less scrutinised, and far more prone to grey-market froth: eye-watering reported GMPs attached to companies with thin fundamentals, driving frenzied retail applications.

The regulator noticed. On 18 December 2024, the SEBI board tightened the rules governing SME IPOs, aimed squarely at the froth. The key changes:

• General corporate purposes (GCP) capped. The amount an SME can raise for vague "general corporate purposes" is now limited to the lower of 15% of the issue size or ₹10 crore — closing a loophole where large chunks of proceeds had no defined use.
• No repaying promoter loans from proceeds. SME issuers can no longer use IPO proceeds to pay back loans taken from promoters, promoter groups, or related parties — a common way money round-tripped out of the company.
• A profitability test. An SME must show operating profit of at least ₹1 crore (EBITDA-type operating profit) in a qualifying period to be eligible — raising the bar on which companies can even come to market.

These are real, meaningful guardrails on SME issuers. But note carefully what they did not do: they did not bring GMP itself under regulation. The grey market remains exactly as unofficial, unregulated, and unenforceable after December 2024 as before. SEBI tightened who can list and how they use the money; it did not, and cannot easily, regulate an informal off-exchange sentiment market. So if you were hoping the 2024 reforms made GMP "safer" or more trustworthy — they did not touch it. The froth-prone dynamics of SME grey markets are still there.

The reason this distinction matters for your decisions is that regulatory action on IPOs is easy to misread as a blanket "the market is cleaner now, so the signals are safer." The froth SEBI moved against in December 2024 was precisely the environment in which inflated grey-market premiums had thrived — small, illiquid issues where a thin float and heavy retail hype produced spectacular reported GMPs on companies that could not support them. Tightening eligibility and use-of-proceeds makes the issuers sounder over time. It does nothing to make the premium quoted on them any more real, verifiable, or enforceable. A better class of SME coming to market is genuine progress; a GMP attached to that SME is exactly as soft as it ever was. Do not let the headline of a regulatory clean-up lull you into trusting the one thing the clean-up left entirely untouched.

How to actually use GMP: a checklist for retail investors

So should you use GMP to decide whether to apply? Use it — but demote it. GMP belongs near the bottom of your decision stack, as a mood reading, never at the top as a forecast. Here is the honest breakdown of what it can and cannot tell you.

GMP can tell you	GMP cannot tell you
Rough pre-listing crowd sentiment — is there hype or indifference?	What your actual listing gain will be
Relative buzz between two IPOs open at the same time	Whether the stock is fairly valued
A directional hint that hype is very high (a caution flag, not a green light)	The quality of the business or its fundamentals
That an informal market exists and is active for this issue	Anything SEBI-verified or officially recorded
—	Whether institutions actually want the stock at this price

Red flags to treat GMP as a warning, not an invitation

• A very high GMP on weak fundamentals. If the grey market is euphoric about a company with thin profits, high debt, or a stretched valuation, read the high GMP as a froth warning, not a buy signal. Hot-market euphoria reverses hardest.
• SME IPO froth. Elevated GMPs on small, illiquid SME issues are the most manipulable and least reliable of all. Extra caution, not extra enthusiasm.
• A GMP that spikes suddenly right before listing. Thin, opaque markets are easy to talk up. A late surge with no fundamental news is a manipulation risk, not a confirmation.

Better signals to actually weigh

If you want to judge an IPO, these carry far more information than a grey-market quote — and every one of them is disclosed or observable, not whispered:

Signal	Why it beats GMP
Fundamentals (revenue growth, profitability, debt, cash flow)	Tells you whether the business is worth owning past day one — the only thing that matters if you hold
Valuation vs listed peers (P/E, P/B, P/S against comparable listed companies)	Reveals whether the issue price itself is cheap or stretched — the actual driver of a pop
QIB vs retail subscription	Heavy Qualified Institutional Buyer demand reflects informed, order-book money — far more meaningful than retail frenzy
Anchor investor quality	Credible institutional anchors committing ahead of the issue is a real vote of confidence, officially disclosed
Subscription pattern across the book	How the demand is distributed across investor categories, from official exchange data, not grey-market chatter

Notice the theme: the better signals are official, disclosed, and reflect institutional money putting itself on the line at a stated price. GMP reflects none of that. If a decision comes down to a strong fundamentals-and-valuation case versus a loud GMP, trust the fundamentals every time.

A useful way to hold all of this together is to ask, of any input you are weighing, a single question: is someone accountable for this number? An anchor investor's commitment is accountable — a named institution has put disclosed capital behind it at a disclosed price. A QIB subscription figure is accountable — it is official exchange data reflecting real orders. A valuation against listed peers is accountable — the comparables are public and you can check the arithmetic yourself. A GMP is accountable to no one. No name attaches to it, no capital is committed by publishing it, and no one is answerable if it is wrong. Once you sort your inputs by accountability, GMP naturally falls to the bottom — not because sentiment is worthless, but because a number nobody stands behind should never outweigh numbers that people and institutions have staked themselves on.

A word on currency and verifiability

Two caveats you must carry with you whenever you look at a GMP:

• The number is real-time, unofficial, and unverifiable. GMP changes continuously and without notice. Whatever figure you saw an hour ago may not hold now, and there is no authoritative source to confirm any given quote. The GMP figures in this article are cited as reported in the grey market precisely because they cannot be independently verified.
• It is not SEBI data. Nothing about GMP is officially recorded, regulated, or guaranteed. Do not treat any GMP number — including reputable aggregator sites' figures — as if it carried the authority of exchange or regulatory data. It does not.

And the standard, necessary line: this is general information, not investment advice. IPO investing carries real risk of loss, listings can and do go below issue price, and past underpricing averages do not predict any individual future IPO. Do your own diligence, or consult a registered adviser, before you commit money.

The bottom line

GMP is a sentiment thermometer dressed up as a price forecast. It exists for a real reason — a genuine market for pre-listing exposure — and it carries genuine, if noisy, information: across large samples, higher premiums do associate with better short-term listings, which is why serious research calls GMP "viable, with caution." But it is unofficial, unregulated, opaque, manipulable, and set by a fragmented dealer network. On individual IPOs — as 2025's main-board cases showed repeatedly, in both directions — it misses often enough that betting an application on it is reckless.

The 20-year record makes the deeper point: yes, Indian IPOs have popped ~14% on average on day one, but that average sits in the middle of a wide, mood-driven distribution that swings from overpricing in cold markets to ~22% underpricing in hot ones. GMP runs hottest exactly when the cycle is most likely to reverse. The signal is loudest when it lies most.

So use GMP the way you would use the crowd's mood at any auction: worth a glance, never the reason you bid. Weigh fundamentals, valuation against peers, institutional and QIB demand, and anchor quality first. Let GMP be the last, weak input — and treat a very high GMP on a shaky company as a warning, not an invitation. That is how you use the grey-market premium without letting it use you.

Sources and further reading

• Jay Ritter, University of Florida — the foundational IPO underpricing datasets and research, the global reference for first-day return analysis.
• Academic analysis of 235 Indian IPOs (1997–2008) — documenting a mean first-day listing gain of ~14.45% (~13.04% market-adjusted) and strongly time-varying, hot-issue-market underpricing (overpricing in 1998–2002, ~22% peak in 2007).
• Academic study of 270 main-board IPOs — evaluating GMP as a short-term listing predictor and concluding it is "viable, with caution."
• SEBI — board decisions of 18 December 2024 tightening SME IPO norms (GCP cap, ban on repaying promoter loans from proceeds, ₹1 crore operating-profit eligibility test).
• Mainstream financial coverage (e.g., Moneycontrol) and grey-market aggregators (e.g., Chittorgarh) — for reported 2025 main-board GMP figures and listing outcomes (Groww, Tata Capital, Lenskart, Studds). All GMP figures are unofficial and unverifiable.