Method

A score you are meant to argue with

Most IPO scores are a black box with a verdict attached. This page is the opposite: the pillars, the weights, the exact curves, and every place the model refuses to guess.

Four rules the model follows

Everything is surfaced. Each pillar on an issue page opens to show its raw metric, the curve the metric was mapped through, the resulting sub-score and the weight applied. You can disagree with the weighting and still check the arithmetic.

Missing data lowers confidence, not the score. A pillar with no usable inputs is dropped and its weight is spread proportionally over the pillars that do have data. The issue page reports what share of the intended 100 points was actually backed by numbers — measured per input rather than per pillar, so a pillar scored from one of its three inputs counts as one third, not as fully backed. A company is never punished for a figure we failed to find.

Sentiment is quarantined. Grey-market premium and subscription demand live in one pillar worth 8 points, flagged as sentiment wherever it appears.

Every curve flattens at the top. Past a point, more stops helping: a 90% growth rate and a 200% growth rate score the same, so no single extraordinary metric can drag a weak business to a strong score. A pillar only reaches 100 if every one of its inputs is at or beyond its top stop, and a total of 100 would need that of all six at once.

The six pillars and their weights

Where the hundred points sit

92 of 100 points come from figures a company has disclosed. 8 come from sentiment.

Growth22
Profitability22
Valuation22
Balance sheet16
Issue structure10
Market demand (sentiment)8

The 8-point sliver on the right is the only part of the score that is not a disclosed figure. Hover a segment for what it covers.

Growth

22 pts

What it measures.: Revenue CAGR across every reported year (45%), the latest full year’s growth (35%), and profit CAGR (20%).
Why it is weighted this way.: One flattering year is easy to produce. Compounding across the whole disclosed window is not, so the CAGR carries the most weight and the single-year figure is a check on whether the trend is still intact.
Where it refuses to guess.: Three refusals. A CAGR out of a zero or negative base is arithmetically defined but financially meaningless. A profit CAGR is void if ANY year inside the window was a loss, not merely the endpoints. And a period shorter than about a year is never annualised, so a nine-month stub cannot be quoted as a compound annual rate.

Profitability

22 pts

What it measures.: Net margin (38%), EBITDA margin (32%) and return on net worth (30%).
Why it is weighted this way.: Margin says how much of each rupee of revenue survives; RoNW says what the business earns on the capital already in it. A company can be large and growing and still destroy value on both counts.
Where it refuses to guess.: Loss-making companies score low here rather than being excluded — the loss is the information.

Valuation

22 pts

What it measures.: P/E at the upper band (38%), premium to the peer-group median P/E (34%) and price-to-sales (28%).
Why it is weighted this way.: Good businesses at bad prices are bad investments. The peer comparison uses the listed companies the issuer’s own prospectus names, which makes it hard to argue the comparison set was cherry-picked.
Where it refuses to guess.: P/E is null for a loss-making company, never zero and never negative. The weight shifts onto price-to-sales instead.

Balance sheet

16 pts

What it measures.: Debt to equity (42%), return on capital employed (34%) and whether operating cash flow is positive (24%).
Why it is weighted this way.: Leverage decides what happens when demand softens. RoCE is harder to flatter than RoNW because it counts borrowed capital too, and operating cash flow is the check on whether reported profit is real.
Where it refuses to guess.: The cash-flow input is a sign test, not a magnitude — a large negative is not scored worse than a small one, because the question is simply whether the business funds itself. And a debt-to-equity below zero is pinned to the bottom of the curve rather than the top: a negative ratio means equity is negative, which is the worst balance sheet an issuer can file, not the absence of debt.

Issue structure

10 pts

What it measures.: Share of the raise that is fresh capital (44%), promoter holding after dilution (34%) and the quality of the anchor book (22%).
Why it is weighted this way.: A pure offer for sale funds the people leaving, not the business. Promoters who sell down to a small stake have less riding on what happens next. Institutions that commit a day early have read the same prospectus you have.
Where it refuses to guess.: A high offer-for-sale share lowers this pillar but is not treated as disqualifying — it is 10 of 100 points, and the caveat panel names it in words as well. An anchor book that has not been built yet counts as missing data, not as a weak result: anchor bidding happens one working day before an issue opens, so scoring the empty slot would penalise every upcoming issue for the calendar.

Market demand

8 pts

What it measures.: Grey-market premium (50%) and total subscription multiple (50%).
Why it is weighted this way.: Sentiment is real information about what a seat is currently worth. It is also the least durable thing on this page.
Where it refuses to guess.: Capped at 8 points by design. A hot grey market can tilt a score; it cannot carry one.

The actual curves

Each metric is mapped to a 0–100 sub-score by piecewise-linear interpolation between these stops, clamped at both ends. These tables are generated from the same constants the scoring code uses, so they cannot drift out of date.

Revenue and profit growth

% a year	-15	-5	0	8	15	25	40	60	90
Score	0	8	18	34	48	64	80	92	100

Net profit margin

% of revenue	-25	-10	-2	0	3	6	10	16	28
Score	0	6	18	26	40	54	68	84	100

EBITDA margin

% of revenue	-10	0	4	8	13	18	26	40
Score	0	10	26	42	56	68	82	100

Return on net worth

%	-25	-5	0	6	11	16	22	32	45
Score	0	10	18	34	50	64	78	92	100

Return on capital employed

%	-10	0	7	12	18	25	34	48
Score	0	10	30	46	62	78	92	100

P/E at the upper band (lower is better)

×	4	10	16	22	30	42	60	90	140
Score	100	90	78	66	53	38	23	10	2

Premium to peer median P/E (lower is better)

%	-60	-35	-15	0	15	35	70	130
Score	100	88	74	62	48	34	18	5

Price to sales (lower is better)

×	0.3	0.8	1.5	2.5	4	6.5	10	18
Score	100	88	76	64	50	35	20	6

Debt to equity (lower is better)

×	0	0.15	0.35	0.6	1	1.5	2.5	4
Score	100	92	80	68	54	40	22	6

Fresh capital as a share of the raise

%	0	20	40	60	80	100
Score	18	36	55	72	88	97

Promoter holding after the issue

%	5	15	25	40	52	65	75
Score	18	34	48	65	78	90	97

Grey market premium (sentiment)

% over issue price	-15	-5	0	4	9	18	32	55	90
Score	2	12	26	40	52	66	80	92	100

Total subscription (sentiment)

×	0.3	0.8	1	2	5	12	30	80	200
Score	2	18	32	44	58	72	85	96	100

Where the data comes from

Curated dataset compiled from RHP/DRHP filings and exchange disclosures.

Figures for issue terms, financials, the peer table, offer objects and promoter holding come from each company's own draft and final prospectus filed with SEBI. Subscription figures come from the exchanges' published bid data. Listing-day prices come from the exchange price feed.

Where a figure could not be verified, the field is a dash. A dash means “not disclosed, or not yet confirmed” — it never means zero, and the charts render the two differently on purpose.

Each issue page says how its figures were checked, and links the issuer's own filing, the regulator, the exchange or the registrar where one is available. It does not publish a directory of the other compilations the figures were cross-checked against: they are an audit trail for us, not a reading list, and several of them are broker marketing rather than sources. The number of them is stated so the depth of the check is visible without turning the page into a link farm.

On ratios, the prospectus wins. Indian issuers define these on bases a reader's balance sheet does not contain: return on net worth is computed on average net worth — opening plus closing, halved — and debt-to-equity divides by total equity, which includes non-controlling interests, rather than the narrower net worth attributable to owners that the same document uses for RoNW and net asset value. Several issuers quote net debt rather than gross. Recomputing those from our own line items looked like a correction and was a category error, so where a prospectus states a ratio that is the figure shown, and hovering the label says whether you are looking at a disclosed figure or one of ours.

The one deliberate exception is price-to-earnings, which is always shown on post-issue earnings. Prospectuses are split between pre- and post-issue bases, and a comparison table that mixes them hands the “cheapest” label to whichever issuer published the more flattering convention.

Data on this site can be stale or wrong. The prospectus is the authoritative document; check anything that matters against it before acting.

Why grey-market premium is kept in a box

The grey market is an informal, unregulated, over-the-counter market in IPO applications and shares before listing. It has no clearing house, no regulator, no reporting obligation and no guarantee that any quoted trade will settle. Quotes are collected by intermediaries and differ between sources.

It is shown here because readers want it and because it is genuinely informative about current enthusiasm. It is labelled as unofficial on every surface it appears, capped at 8 of 100 points in the score, and never presented as a forecast of listing price.

That cap is a claim about how much the premium is worth, and a claim about data should be checked against the data. So here is the record across every issue in this dataset that had a quote before listing and has since listed.

Premium before listing, against the actual move

The last grey-market quote recorded BEFORE listing day, against what the stock actually did, across the 12 tracked issues that have both.

A tracked listing
Where premium equals outcome

Direction right
92%
of 12 listings
Average miss
9.8 pts
in magnitude
Bias
+6.9 pts
tended to undershoot
Worst single call
37.8 pts
Bharat Coking Coal

Why the grey market is kept in a box →

Hover any point for its name and figures; labels are shown standing only where they fit without overlapping. A point on the dashed line means the premium matched the outcome exactly. Above it the grey market undershot; below it, it overshot. Only quotes recorded strictly before listing day are counted, so a premium quoted once trading had begun cannot flatter the record. This is 12 issues from this dataset alone — far too small a sample to generalise from, and shown to put a number on how much weight the premium deserves rather than to validate it.

What this score cannot do

It cannot predict listing performance. Nothing can; listing-day price is set by demand on the day, not by the fundamentals the model reads.

It cannot read the qualitative risk factors in a prospectus — litigation, customer concentration, regulatory exposure, related-party transactions, the quality of the management team. Those sections are where the real surprises live, and there is no substitute for reading them.

It cannot tell you whether an issue suits your circumstances, horizon or risk tolerance. That is what a SEBI-registered adviser is for.

The IPOVista Score is a transparent summary of disclosed fundamentals and issue terms. It is research, not investment advice, and it cannot predict listing performance.

IPOVista publishes research and data compiled from public filings and exchange disclosures. Nothing here is investment advice, a recommendation, or an offer to buy or sell any security. Figures can contain errors and sources can be stale — verify against the prospectus before you act, and speak to a SEBI-registered adviser. Investing in securities carries risk, including loss of capital.

See the method applied to every tracked IPO →

% a year

-15

-5

Score

100

% of revenue

-25

-10

-2

Score

100

% of revenue

-10

Score

100

-25

-5

Score

100

-10

Score

100

140

Score

100

-60

-35

-15

130

Score

100

0.3

0.8

1.5

2.5

6.5

Score

100

0.15

0.35

0.6

1.5

2.5

Score

100

Score

% over issue price

-15

-5

Score

100

0.3

0.8

200

Score

100