IPO subscription tracked listing gains across 431 Indian IPOs, but it works as a sorter, not a direction call

By Strota Newsroom · 2026-10-20 · How Strota reports

Across 431 mainboard IPOs that listed on NSE between 2016 and 2026, the subscription numbers the exchanges publish lined up closely with how the shares opened on day one. The rank correlation between total subscription and the opening gain came out at 0.76. The more useful finding is narrower than 'subscription predicts pops', though. Most IPOs in this period opened higher regardless, so on the simple question of up or down, subscription added only a modest edge. Where it did real work was in separating the issues that almost always popped from the ones that were close to a coin toss.

That matters because of what Indian IPO watchers usually lean on instead: the grey market premium, or GMP. It is an informal dealer quote. No exchange or regulator stands behind it, and nobody keeps an official record of it. Subscription is the mirror image. The exchanges publish it, broken out by investor category, once bidding shuts. The question we asked was simple: how much of the listing-day outcome does that official figure actually explain?

The test covered every mainboard IPO on NSE listed from 8 February 2016 to 5 October 2026. Partly-paid issues and offers without subscription data were dropped, which left 431. For each one we took the final subscription by category, combined across NSE and BSE, and set the listing-day open and close against the final issue price.

First, the baseline. Of the 431, 301 opened above the issue price, 29 opened exactly on it and 101 opened below. The median opening gain was 9.7%. Every investor category's demand moved with the opening gain, but not equally: the rank correlation was 0.73 for the institutional (QIB) book, 0.70 for non-institutional bidders and 0.60 for retail.

Grouping the IPOs by total subscription makes the pattern hard to miss. Below 3x, 112 issues produced a median opening move of -0.7%, only 39 of them opened higher, and 57 of the 112 actually opened under their issue price. The 3-10x band managed a median of 3.5%, and 10-30x reached 6.9%. Then the curve steepens: in the 30-75x band, 91 of 99 opened higher with a median gain of 20.2%. All 78 issues in the 75-150x band opened above issue, at a median of 43.8%, and so did all 24 at 150x or more, at a median of 73.6%. Across those 102 heavily bid issues, the weakest opening was a gain of 1.89%.

Medians hide a lot of spread, and the top bands were not tidy. For the 150x-plus group, the middle half of outcomes ran from 38.2% to 110.8%. Heavy demand reliably meant an up open in this sample; it did not pin down how big that open would be.

A pattern found over the whole period could simply be fitted to it, so we also ran an out-of-sample check. The bands were built using only the 162 IPOs listed before 2023, then applied to the 269 that listed from 2023 onward, with a flat open counted as 'not up'. On direction, the band's historical median called the outcome correctly 78.8% of the time. That sounds strong until you notice that 71.4% of those 269 IPOs opened higher anyway, so a lazy rule of always predicting an up open would have scored 71.4%. Subscription bought roughly seven points over that base rate. Setting aside the 19 flat openings, the comparison is 80.8% against 76.8%. Real, but modest.

Sorting is where the bands earned their place. Of the 60 IPOs the bands tagged as poor bets to open higher, 40 duly failed to, compared with 28.6% of all 269 that failed to open higher. At the other end, every IPO subscribed 75 times or more opened up. On the size of the move, the typical gap between a band's median and the real opening gain was 9.9 percentage points, while one fixed guess for every issue (the pre-2023 median, 6.3%) missed by 11.6. The rank correlation between predicted and actual gains was 0.73. The levels themselves shifted between periods, though: the 75-150x median slid from 54.3% to 38.4%, so the band figures are a rough guide rather than fixed constants.

So how does the grey market quote compare? Outside trackers that line up past GMP readings with real listing outcomes put its directional hit rate at roughly 70%. A blanket bet on a pop would land near the same mark, since most issues open higher. Direction alone, then, is a weak yardstick for judging either signal. Subscription's value lies in telling the near-certain pops apart from the toss-ups, and it is official, checkable and free. GMP quotes also tend to climb as subscription climbs, which suggests much of what GMP signals is already sitting in the exchange's own numbers.

One more pattern showed up after the opening bell. On listing day, the median IPO closed 0.2 points lower than where it opened, and 51% of issues finished the day under their opening price. In this record, the listing pop happened mostly in the opening auction, not in the trading that followed.

The caveats are real. Subscription is only final when bidding closes, and figures seen mid-bid understate the end number, institutions especially, since they tend to bid on the last day. The data spans a single long and mostly kind decade for Indian IPOs, and a weak market on listing day can drag down even a heavily subscribed issue. Anchor investors sit outside the subscription figures, and SME IPOs were not part of the study. Above all, this describes how past listings behaved; it is not a forecast for any IPO.

More money stories

• New moon, full moon, same Nifty: 19 years of data show no reliable lunar effect
• RBI opens a dollar window for 3 oil firms as the rupee nears 97
• Banks pay 5.67 per cent on new deposits while your EMI rises
• Air India adds Rs 1,200 fuel surcharge as jet fuel hits Rs 137
• EPF ceiling rises to Rs 25,000. Employers must not cut pay
• Phone prices rose 16 per cent, so brands pay your EMI interest