IPO Oversubscription Explained

A headline saying an issue closed "subscribed 12.4 times" sounds like a verdict, and it is read like one. It is actually a fraction. Somebody divided the total number of shares that applicants bid for by the total number of shares the company offered, and 12.4 was the answer. Nothing in that arithmetic says the price is right, that the business is sound, or that anybody who receives shares will do well. What the fraction does describe precisely, and usefully, is how crowded the queue was, which is the single biggest determinant of whether an application will receive anything at all.

The fraction, spelled out

Subscription equals shares bid for divided by shares offered. If a pool contains 10 lakh shares and applicants have collectively bid for 45 lakh shares, that pool is subscribed 4.5 times. A value below one means the pool did not attract enough demand to be filled, which is undersubscription. A value of exactly one means demand matched supply. Anything above one is oversubscription, and the amount above one measures the excess demand that will have to be turned away.

Two details in that formula are worth pausing on. First, the numerator counts shares bid for, not applicants and not rupees, so a single very large application and thousands of small ones can produce the same number. Second, both figures refer to a specific pool, not to the issue as a whole. Getting the second point wrong is the most common reason a reader forms the wrong expectation about their own chances.

There is no single subscription number

An Indian public issue reserves separate pools of shares for qualified institutional buyers, non-institutional investors and retail individual investors, with additional reserved pools for employees or existing shareholders in some issues. Our companion article on QIB, NII and retail IPO categories explains that split in detail. Each pool runs its own subscription calculation, and each is allotted separately. The overall figure you see in a headline is simply total bids across all pools divided by total shares offered, which makes it a weighted blend of numbers that have nothing to do with one another.

Consider an illustrative issue of 1 crore shares, reserved as 50 lakh for institutions, 15 lakh for non-institutional applicants and 35 lakh for retail. Suppose institutions bid for 4 crore shares, non-institutional applicants for 3 crore, and retail applicants for 1.05 crore. Total demand is 8.05 crore shares against 1 crore offered, so the headline reads 8.05 times subscribed. But the institutional pool is 8 times subscribed, the non-institutional pool is 20 times subscribed, and the retail pool is only 3 times subscribed. A retail applicant reading 8.05 and bracing for near-impossible odds has misread their own position by a wide margin. The numbers here are invented to demonstrate the arithmetic.

“The overall subscription multiple describes an issue. The multiple for your own category describes your application. Only one of those two numbers is about you.”

From a multiple to a probability

This is the part that actually affects an applicant, and it works differently in each category. The mechanics of the calculation are set out in the basis of allotment for every issue, and our dedicated article on the basis of allotment goes through the process step by step.

Retail: a lottery for minimum lots

Retail applications are made in whole lots. When the retail pool is oversubscribed so heavily that the minimum lot cannot be given to every applicant, allotment is made in minimum lots to as many applicants as can be accommodated, and those applicants are selected by a computerised draw. Continuing the illustration above: 35 lakh retail shares with a lot size of 70 shares gives 50,000 lots. Retail demand of 1.05 crore shares is 1.5 lakh lots. If every application was for one lot, roughly one application in three succeeds, which matches the 3 times figure exactly.

The consequence that surprises people is what happens if you apply for more lots. In the heavily oversubscribed regime, one application is one entry in the draw regardless of how many lots it asked for, and the maximum you can receive is one lot. Applying for five lots in that situation blocks five lots' worth of money and does not multiply your probability of being drawn. There is a second regime, though: if the pool is oversubscribed only mildly, so that every applicant can be given at least the minimum lot, the remaining shares are distributed proportionately among those who asked for more. In that case a larger application does receive more. The pool's own multiple determines which regime applies, which is exactly why the category figure matters.

Non-institutional: proportion within a sub-bucket

The non-institutional pool is divided by application size into two sub-buckets, each with its own share of the reservation, so that very large bids do not crowd out moderate ones. Within each sub-bucket, allotment reflects the size of the application relative to total demand in that sub-bucket, subject to a minimum lot being given to as many applicants as possible and a draw where even that cannot be accommodated. Because much non-institutional demand is funded by short-term borrowing arranged for the bidding days, a high multiple here often reflects the cost and availability of that funding as much as any view about the company.

Institutional: proportionate, and anchors are already settled

The institutional pool is allotted proportionately, so a fund that bid for a tenth of the demand in that pool receives roughly a tenth of it. Note also that a portion of the institutional reservation may already have been placed with anchor investors before the issue opened, as explained in our article on anchor investors. Whether a published institutional multiple is calculated on the full institutional reservation or on the net portion after the anchor allocation differs between sources, so check what the figure in front of you actually refers to.

Why issue size distorts the headline

A subscription multiple is a ratio, and ratios say nothing about absolute magnitude. Take two illustrative issues that attract exactly the same ₹5,000 crore of applications. The first is raising ₹100 crore, so it reports 50 times subscribed. The second is raising ₹5,000 crore, so it reports 1 time subscribed. Identical rupee demand; wildly different headlines.

Small issues therefore produce dramatic multiples routinely, because it takes very little money to overwhelm a small pool, while a large issue can attract enormous absolute interest and still look unremarkable on the multiple. If you want to compare demand across issues meaningfully, compare the rupee value of bids, or at least look at the multiple alongside the issue size rather than in isolation. Our note on IPO issue size explains how that number is constructed.

How the number behaves during the bidding window

Subscription is published live during the bidding window and updated through the day, which tempts people to read it as a running poll. It is not one, because the categories do not bid on the same rhythm. Retail demand tends to build steadily from the first day. Institutional demand often arrives close to the deadline, because institutional and non-institutional bids cannot be withdrawn or lowered at any stage once placed, and committing early means being locked in while conditions can still change. Non-institutional demand also clusters late, since borrowed funds cost interest for every day they sit committed.

The practical result is that a figure seen on the first morning systematically understates where the book will end, and that a retail applicant watching the total climb has no reliable way to extrapolate it. Bids in some categories can also be revised or withdrawn before the close, so intra-window numbers are provisional in both directions. Treat only the final post-close figures as the ones worth reasoning from, and read them together with the detailed subscription data published by the exchanges.

The other end: undersubscription

An issue that fails to attract enough demand does not simply allot to whoever applied. Regulations impose a minimum subscription level for the issue as a whole, and a book-built issue additionally requires the institutional portion to reach a prescribed threshold. If either condition is not met, the issue does not proceed, no shares are allotted, and blocked application money is released back to applicants. Retail enthusiasm on its own cannot rescue an issue that professional investors have declined to fund.

Where one general category falls short while another has surplus demand, the offer document sets out how unsubscribed shares may be reallocated between them. That reallocation is one reason a category's final multiple can look different from what it showed while bidding was open, and it is another argument for reading the registrar's final basis of allotment rather than a mid-window screenshot.

A worked example from application to outcome

Assume, purely for illustration, a price band with an upper end of ₹200 and a lot size of 70 shares, so one lot costs ₹14,000 at the cap. An applicant bids at cut-off price for one lot; ₹14,000 is blocked in their bank account. The retail pool of 35 lakh shares translates to 50,000 lots, and by the close retail demand stands at 1.5 lakh lots, or 3 times subscribed.

Because the pool cannot give a minimum lot to all 1.5 lakh applicants, the registrar runs a draw for the 50,000 available lots. Our applicant has roughly a one-in-three chance. If drawn, 70 shares are credited to their demat account and ₹14,000 is debited; if not, the block is released and the full amount becomes available again. Had the same applicant bid for three lots and blocked ₹42,000, their chance of being drawn would have been the same one-in-three, and the most they could have received in this regime would still have been a single lot.

Now change one thing. Suppose retail demand had been only 60,000 lots against 50,000 available, a multiple of 1.2. Almost every applicant can be given the minimum lot, and the small surplus is distributed proportionately, so the three-lot applicant would plausibly receive more than the one-lot applicant. Same issue, same applicant behaviour, different regime, entirely different logic. That is why the retail multiple is worth reading before drawing conclusions about how many lots to apply for.

What the multiple cannot tell you

• It is not a valuation opinion. A heavily oversubscribed pool reflects how many shares people asked for at a price already fixed within a band, not whether that price is reasonable relative to earnings, assets or prospects.
• It does not forecast the listing price. Listing depends on conditions on that day, the size of the free float, sentiment and the behaviour of holders whose lock-ins have expired, none of which are captured in a pre-close ratio.
• It is not comparable across issues of different sizes, because a small reservation inflates the multiple mechanically.
• It does not distinguish conviction from leverage. Demand funded by short-term borrowing for a few days is counted identically to demand from an investor intending to hold for years.
• It says nothing about how long anyone intends to remain a shareholder after listing.
• It is not a substitute for the prospectus. Business model, financial history, use of proceeds, related-party arrangements and risk factors all sit in a document the subscription number never touches.

Misreadings worth avoiding

1. Quoting the overall multiple when the retail multiple is the one that governs your allotment.
2. Applying for extra lots on the assumption that it improves lottery odds in a heavily oversubscribed pool, when it mainly increases the money blocked.
3. Extrapolating a first-day figure, when the categories that move the number most typically bid at the end.
4. Comparing a small issue's multiple with a large issue's multiple as though they measure the same intensity of demand.
5. Reading an unsubscribed-looking figure early in the window as evidence that an issue has failed, before the institutional book has been filled.
6. Relying on forwarded screenshots rather than the exchange's own published figures, which are the record that the registrar's allotment will reconcile to.

Where the authoritative figures come from

Live and final category-wise subscription data is published by the stock exchanges for each open issue. The reservation for each category, the lot size and the price band come from the red herring prospectus. After the issue closes, the registrar publishes the basis of allotment, which is the definitive statement of how many applications were received in each category, what method was applied, and how shares were distributed. Where those documents and any secondary source disagree, the official filings govern. The underlying rules sit in SEBI's issue of capital and disclosure requirements regulations and can be amended, so confirm current thresholds and methods rather than relying on a remembered figure.

Used properly, an oversubscription number answers one question well: how many people wanted how many shares in the pool I am bidding into, and therefore how likely am I to receive anything. That is a genuinely useful thing to know before committing money for the days an application stays blocked. It is a poor answer to any question about value, and treating it as one is how a straightforward ratio turns into an expensive assumption.