Subscription Patterns
0 issues · 0 with a listing outcome · updated 19 Sept 2026, 1:29 am IST
The headline subscription multiple is the least informative number on any IPO page, because it averages three groups with completely different information and completely different incentives.
| QIB | Institutions with analyst teams and access to management. |
| NII | Non-institutional, often applying with borrowed funds and frequently exiting on listing day. |
| Retail | Individuals, largely reading the same publicly reported GMP as everyone else. |
An issue subscribed 40× overall might be 55× QIB and 8× retail, or 4× QIB and 90× retail. Those are opposite situations, and every site prints the same "40x" for both. This page separates them and attaches what happened.
By pattern
| Pattern | Listings | Median gain | Mean gain | Positive | Range |
|---|---|---|---|---|---|
| Institution-led
QIB subscribed more than twice the retail multiple. The people with analyst teams wanted it more than the public did. |
0 | — | — | — | — |
| Retail-led
Retail subscribed more than twice the QIB multiple. Public enthusiasm ran ahead of institutional. |
0 | — | — | — | — |
| Retail frenzy
Retail subscribed more than ten times the QIB multiple. The gap is extreme. |
0 | — | — | — | — |
| Broad demand
No category dominated — all three within a factor of two of each other. |
0 | — | — | — | — |
| Thin
Total subscription under three times. Demand was limited across the board. |
0 | — | — | — | — |
Figures are shown only where at least 8 listings have accumulated. An average over four issues is not an average, and printing one with a quiet asterisk is how a number becomes a claim nobody decided to make. Median is given beside mean because one listing that doubled drags a mean across a small sample.
Issue by issue
| IPO | QIB | NII | Retail | Total | Pattern | Listing |
|---|---|---|---|---|---|---|
| No category-wise subscription recorded yet. | ||||||
How a pattern is assigned
By the ratio of retail to QIB demand, not by absolute multiples — so a small SME issue and a large mainboard one can carry the same label. The thresholds are round numbers chosen to be legible. They are deliberately not optimised: a boundary tuned to maximise the difference between buckets would be a fitted model pretending to be a category.
Related: GMP accuracy · listing performance · lock-in expiries