How to read a DRHP and RHP: the Jio IPO, page by page

An IPO document is about 600 pages long, and it’s the one place where a company has to tell you the uncomfortable parts in writing. You just need to know where they’re kept.

This guide walks through how an analyst reads an offer document, using a real one. Jio Platforms went from draft to listing between June and October 2026, so we can see all three documents and what happened afterwards. Every Jio figure comes from those filings unless I say otherwise, and page numbers refer to the final Prospectus. I’ve used Jio because it’s a rich example, not because this is a view on the stock.

The three documents, in plain terms

Think of buying a flat in a new building. First you get a brochure that the authorities are still reviewing: floor plans, but no final price. Then a price list with a range, just before bookings open. Finally, the sale agreement, with your exact price written in. An IPO works much the same way.

Sources: Jio Platforms DRHP, RHP and Prospectus.

One detail surprises people. It’s often said the RHP “contains the price band”, but Jio’s RHP cover still shows ₹[●]. It says the band would be announced in newspaper advertisements at least two working days before bidding opened. Jio’s band of ₹1,065 to ₹1,119 reached investors that way, so read the RHP and the price announcement together.

“Red herring” sounds odd, but the idea is simple: the document is complete except for the price. So why read the draft at all? Because it comes months earlier and shows you what changed. Jio’s own IPO page warned investors to rely on the RHP and not the June draft. The draft is for early research; the RHP is for your decision.

Jio’s timeline, for reference:

SEBI’s letter isn’t a stamp of quality. The Prospectus cover says plainly that SEBI has not recommended or approved the shares, and doesn’t guarantee the accuracy of the document.

The analyst’s route: 11 stops

Don’t read front to back. The contents page is your map, and the order below is how I’d move through it, from “what’s the deal?” to “what could go wrong?”

Stop 1: The cover and “The Issue” (p. 1, 72)

The question: who is getting the money?

An IPO has two possible parts. In a fresh issue, the company prints new shares and keeps the cash. In an offer for sale (OFS), existing owners sell their shares and the money goes to them.

Daily life: Your friend’s café is raising money. If they sell you a share to pay for a new oven, that’s a fresh issue. If a partner sells you their share and leaves with the cash, that’s an OFS, and the café gets nothing.

Jio’s offer was 100% fresh: 27 crore new shares and no OFS. Meta, Google and Reliance sold nothing in the IPO. That’s unusual at a time when, per SEBI’s bulletin, 75% of September’s mainboard money came through OFS.

Also note what the offer doesn’t tell you directly: the proportion being sold. By my arithmetic, 27 crore of 920.9 crore post-issue shares is just 2.93% of the company. At ₹1,119, the whole company was priced at about ₹10.3 lakh crore. So the public bought a very small slice of a very large company.

Stop 2: Objects of the Issue (p. 119)

The question: could I hold them to this in a year?

This is the company’s promise on how it will spend your money. Good objects are specific and checkable. Be careful with a large “general corporate purposes” line, which is close to a blank cheque (the rules cap it, and Jio states a limit of 25% of gross proceeds).

Source: Prospectus p. 119, converted from ₹ million. Figures subject to the basis of allotment.

So about 93% of the net money repays debt. It won’t fund a new network. That’s a clear plan and easy to verify later, and it does lower interest costs. Now read the fine print on the next few pages, as an analyst would:

• Which loans? Foreign-currency loans in dollars and yen, due in 2028, at rates of 1.40% to 5.13%. (p. 121–124)
• Who gets repaid? The lenders include HDFC Bank and affiliates of BofA, BNP Paribas and SBI Capital Markets, all connected to the banks managing the IPO. This is disclosed, and that’s what the footnote is for. (p. 124)
• A question to ask: those rates look low next to the 7.2% to 8.65% Jio pays on spectrum dues (see Stop 5). Why repay these first? The answer may lie in hedging costs, loan terms or maturity, since the stated rate isn’t always the full cost of foreign debt. The document lists its selection factors, but it doesn’t answer this directly. Good analysis often ends in a good question.

Stop 3: Our Business and Industry Overview (p. 156, 201)

The question: what actually brings in the money?

These are the most enjoyable sections and the least neutral. They are written to sell. Two habits help:

Check who wrote the industry data. Jio’s market data comes from an Analysys Mason report which, as the document states, was commissioned and paid for by the company. That doesn’t make it wrong. Nearly all IPOs do this. But a phrase like “largest” or “first” deserves a second source.

Compare the story with the numbers. Jio describes itself as a “technology platform company”. Now look at the risk section (p. 34): one subsidiary, Reliance Jio Infocomm (RJIL), the telecom operator, produced 87.74% of revenue, 94.88% of EBITDA and 93.76% of profit in FY26. Prepaid connectivity alone was 77.08% of revenue, and the company says its KPIs “do not reflect the technology platform specific metrics”.

Daily life: A tea stall calling itself a “beverage and hospitality platform” may truly plan to become one, and might even succeed. But if 90% of today’s income is from cutting chai, value it first as a very good tea stall, then decide what the extra plan is worth to you.

The platform story might turn out right. The analyst’s job is to separate what the company earns today from what it hopes to earn, because you are paying for both.

Stop 4: Financials and KPIs (p. 74, 134)

The question: is the business getting better, and how do we know?

Start with the summary tables. “Restated” means the past years have been presented on a consistent basis, so you can compare like with like.

Source: Prospectus p. 134–135, converted from ₹ million. *One quarter, not annualised. Q1 FY27 = April to June 2026. ARPU and churn are for the exit quarter.

Here’s how to read this like an analyst:

• Revenue = customers × ARPU. That’s the whole engine. ARPU (average revenue per user) is roughly the monthly recharge. Jio’s rose from ₹181.7 to ₹215.6. Then compare: per the peer table, Airtel’s India mobile ARPU was ₹263.9 and Vodafone Idea’s was ₹177. (p. 139)
• Churn is the share of customers who leave each month, like the share of gym members cancelling. Jio’s 1.61% compares with 2.59% for Airtel India mobile and 3.8% for Vodafone Idea. But note the warning: definitions differ between companies, so treat these as a guide.
• Look at what’s going the wrong way. Revenue, margin and profit all rise. Return on capital employed falls, from 12.83% to 10.76%. Growth is easy to show, but it matters how much capital it takes.
• Check what’s inside EBITDA. Jio’s definition starts from profit before tax, so it includes other income, ₹2,874 crore in FY26. Without it, I calculate a margin of about 50%, not 51.91%. It’s a small difference here, but always read the definition of any KPI.

Stop 5: Cash flow and the balance sheet (p. 75, 77, 438)

The question: where did the cash really go, and what’s owed?

If you only do one thing beyond the summary, do this. Profit is an opinion shaped by accounting rules. Cash is what’s left in the bank.

Jio highlights a metric called “EBITDA less cash capex”, which grew from ₹1,449 crore to ₹42,071 crore. It’s a real improvement. But telecom has other large bills, so I built a stricter version from the cash flow statement:

Source: Prospectus p. 77, converted from ₹ million. The final line is my own measure, not a company figure.

Daily life: Your salary minus rent isn’t what you can save. You still have the home loan EMI, the car EMI and school fees. The bottom line above is closer to what’s really left at the end of the month.

The direction is striking: from about ₹17,700 crore short to about ₹20,300 crore surplus in two years, mostly because capex fell sharply as the 5G rollout matured. So the key question moves to the future: does capex stay low, or does it come back with the next spectrum auction? Risk factor 1 says new spectrum could require significant capital.

Now the debt, and the most important footnote in this guide. Jio reports net leverage of just 0.36x for FY26. Read the definition (p. 135): borrowings minus cash and marketable securities, divided by EBITDA. Now look at the balance sheet (p. 75). There’s a separate line called “deferred payment liabilities”, ₹1,04,514 crore, and the note shows it’s owed to the Department of Telecommunications for spectrum, in annual instalments, with interest of 7.20% to 8.65%.

Source: Prospectus p. 75, 134–135, and Note 17 on deferred payments. The second column is my calculation, not a company figure, and it excludes lease liabilities.

Daily life: Imagine a friend saying “I’ve almost no debt,” and that’s true of their bank loans, but they’re also paying off a flat to the builder over 15 years with interest. Their loan statement is accurate. It just isn’t the whole picture.

To be clear, this isn’t an accusation. The definition is fully disclosed, the liability sits in plain sight, and 1.73x isn’t alarming in itself. The lesson is that a KPI answers the question its definition asks. Rebuild the important ones yourself, and compare peers on the same basis.

Two more balance-sheet habits worth learning:

• Assets moving out of “under construction”. Capital work-in-progress and spectrum under development fell from about ₹2.1 lakh crore in FY24 to about ₹23,000 crore in FY26. While assets are being built, their interest can be added to their cost. Once they’re in use, it goes through profit, and depreciation starts. So finance costs rose from ₹4,905 crore to ₹8,653 crore in FY26, even as cash interest paid fell. It’s like interest on an under-construction flat: it begins to hurt your monthly budget once you move in.
• Tax that’s been deferred. In FY26, Jio’s profit before tax was ₹40,353 crore. Its tax expense was ₹10,300 crore, but only ₹118 crore was current tax, and the deferred tax liability reached ₹41,210 crore. So reported profit already deducts tax, though very little went out in cash. Check the notes to see when that may change, because it affects future cash flow.

Stop 6: Basis for Issue Price (p. 129)

The question: why this price, compared with what?

This is where the company justifies its valuation. It’s the section most retail investors should read and few do.

Source: Prospectus p. 132–133, 139. Peer P/Es use closing prices on 23 Oct 2026. Airtel is consolidated, including Africa.

P/E tells you how many years of current profit you’re paying for. If a kirana shop earns ₹1 lakh a year and costs ₹33 lakh, its P/E is 33. Three things to notice here:

• The “industry average” can mislead. The document gives an average peer P/E of 23.09. But Vodafone Idea’s 3.54 comes from a profit that included ₹58,607 crore of exceptional items, as the footnote explains. In a two-company peer set, one distorted number moves the whole average. Always read the footnotes under the peer table.
• Peers are chosen. Jio compares itself with telecom operators while describing itself as a technology platform. A sharp reader asks what that choice implies for the valuation.
• Read ratios together. Jio’s P/E is below Airtel’s, but so is its return on net worth, 9.42% against 20.32%. Lower P/E doesn’t automatically mean cheaper; a business earning lower returns may deserve a lower multiple. You’d also compare growth, debt and scale before reaching a view.

Also in this section: the weighted average cost of acquisition, what insiders paid. Reliance’s average cost is ₹89.58 a share (p. 117). The capital history shows the 2020 private placements to Meta’s affiliate and Google were at ₹488.34 (p. 102–105). The public paid ₹1,119. A gap like this is normal for early investors, who took on more risk and have waited six years. But it does help you judge how much of the good news the price already reflects.

Stop 7: Risk Factors (p. 30)

The question: what could hurt only this company?

Jio lists its risks in order of materiality, as judged by the company, and that order is itself information. Skip the boilerplate (“the economy may slow”) and look for what’s specific. Jio’s first four:

1. Licences and spectrum. The telecom licence is due for renewal in October 2033, and most spectrum expires in 2041–42. The business depends on government permissions.
2. Dependence on the Reliance Group. Key agreements with the parent and sister companies. (More at Stop 8.)
3. A single distributor. Reliance Retail is the sole distributor of prepaid services, which are about 77% of revenue.
4. One subsidiary carries the group. The RJIL concentration we saw at Stop 3.

Notice how useful risk 3 is. Most people would never guess that almost all recharges flow through one related company. That’s what this section is for: finding the facts the marketing pages have no reason to mention.

Daily life: Read risk factors like the side-effects leaflet in a medicine box. “May cause headache” in every leaflet tells you little. “Don’t take with blood pressure tablets” tells you something specific to act on.

Stop 8: Related parties and the promoter (p. 31, 80, 318)

The question: does money move between insiders outside the shareholding?

A related party transaction is business with a connected company. It’s common and legal, particularly in Indian groups, but it needs a careful look, because the two sides aren’t strangers negotiating hard.

Daily life: A shop rents its premises from the owner’s brother. It might be a fair deal. But minority partners should know the rent, and whether it’s market rate.

Jio’s tables make this visible. In FY26, “revenue received in advance” attributed to the Reliance Retail arrangement was ₹1,12,935 crore. Jio bought ₹3,904 crore of equipment and intangibles from its parent, and that share of cash capex jumped to 46.46% in the June 2026 quarter (partly because total capex fell so much). Jio also paid the parent ₹2,625 crore in professional fees. The document says material transactions were at arm’s length and approved by the audit committee and RIL’s shareholders.

You can’t judge fairness from a table. But you can note the scale, follow the trend, and watch these figures in the annual reports after listing.

Stop 9: Capital structure and lock-ins (p. 102)

The question: who owns it, and when can they sell?

Reliance holds 66.43% before the issue and 64.48% after. Next are Meta (9.98%) and Google (7.73%), and the top ten hold 97.32%. A lock-in is a period when holders can’t sell, and the end of one can bring more shares to market.

Source: Prospectus p. 107–109.

Put those dates in your calendar. Here, with only 2.93% sold to the public and over 90% with long-term holders, the free float is small next to the shares that become eligible for sale after six months. That doesn’t mean they will sell, but you should know the supply exists.

Stop 10: Litigation and contingent liabilities (p. 79, 468)

The question: is anything large enough to matter?

Start with the materiality threshold: for Jio it’s ₹1,293 crore, so only civil cases above that need individual disclosure (other categories, like criminal and regulatory matters, are listed regardless). Then look at size relative to the company. Jio’s contingent liabilities were ₹1,469 crore, just 0.43% of net worth, and the Company itself shows “Nil” in each category. Most listed items at RJIL are ordinary for a telecom operator, such as disputes over tower radiation and customers’ blocked numbers.

Don’t count the number of cases. A large operator will always have many. Ask two things: is any single case large compared with net worth, and does anything involve fraud, the regulator or the core licence? Then look for penalties: the document mentions ₹123 crore in past non-performance penalties and liquidated damages tied to service interruptions under agreements with the government’s Universal Service Obligation Fund (p. 42).

Stop 11: Issue Structure and how you actually bid (p. 534)

The question: what are my real odds, and what am I signing up for?

Jio’s net issue was split among three groups, with separate quotas for employees and Reliance shareholders:

Source: Prospectus p. 72. Separately: Reliance shareholders ₹2,500 crore, employees ₹500 crore (at a ₹54 discount).

• Lot size: 12 shares, so ₹13,428 at the top of the band. You bid in multiples of 12.
• ASBA: your money stays blocked in your own bank account until allotment. Think of a hotel holding an amount on your card; it only gets charged if you get a room.
• Cut-off price: retail investors can tick this to agree to whatever final price is set.
• Allotment: when the retail book is oversubscribed, each investor’s allocation can’t be below the minimum lot, subject to availability. So applying for more doesn’t improve your odds much.
• Multiple quotas: eligible Reliance shareholders could bid in both the shareholder quota and the retail or non-institutional portion, without it counting as a multiple bid. Details like this are found only by reading this section.

What changed from draft to final

Comparing versions takes minutes and shows you what the company had to update.

Sources: DRHP p. 70, 106; RHP cover; Prospectus p. 79, 119.

Here the core plan stayed steady, which is itself useful. When a company changes the size of its offer, the sellers or its use of proceeds between draft and RHP, you’d want to find out why.

What happened to Jio

As reported by Groww, the issue was subscribed 22.19 times overall: QIBs 70.55x, non-institutional investors 14.00x, retail 2.72x, Reliance shareholders 7.60x and employees 2.93x. Shares listed at ₹1,358 on 28 October, a 21.36% premium to the issue price. (CNBC describes the debut as over 25%, so sources differ in how they measure it. Check exchange data for exact figures.)

That first day tells you what the market would pay then. It doesn’t tell you whether the analysis above comes out favourably. An investor who wrote down their questions in October can test them over the coming years: does the debt fall as promised, does capex stay low, does ARPU keep rising, and does non-telecom revenue become visible?

Institutions bid for about 26 times more of their quota than retail did. A good reason to read the document isn’t to copy them, but to know what they were looking at.

Your one-page summary

After your route, you should be able to complete this:

Then finish with one sentence: “This company is selling X to fund Y, at a price justified by Z, and the main thing that could go wrong is W.” If you can’t fill in those blanks, you aren’t ready to decide, and that’s a useful thing to know.

Six common mistakes

1. Treating grey-market premium as analysis. It’s unofficial, and it isn’t in any of these documents.
2. Treating a big brand as a good price. You can love your Jio connection and still need to ask what the share costs.
3. Reading only the summary. The net-debt definition and the peer-table footnote both sit in small print.
4. Taking KPIs at face value. They’re company-chosen and defined, and the document itself warns they may not be comparable.
5. Reading subscription numbers as proof. Oversubscription tells you about demand, not value.
6. Stopping at listing. Companies must keep reporting. Check the promises against the next annual report.

Glossary

ARPU

Average revenue per user, roughly the average monthly bill.

Anchor investor

A large institution allotted shares before public bidding opens.

ASBA

Your bid money stays blocked in your own bank account until allotment.

BRLM

Book running lead manager, the investment bank running the IPO.

Capex

Money spent on long-lasting assets, such as towers and spectrum.

Churn

The share of customers leaving in a given period.

Cut-off price

An option to accept whichever final price is decided.

EBITDA

Earnings before interest, tax, depreciation and amortisation.

Fresh issue / OFS

New shares where the company gets the cash / existing shares where sellers get it.

Lock-in

A period during which certain holders can’t sell.

NAV

Net asset value: net worth per share.

P/E

Share price divided by earnings per share.

QIB / NII / RII

Institutional buyers / non-institutional investors (above ₹2 lakh) / retail investors (up to ₹2 lakh).

RoNW / ROCE

Return on net worth / return on capital employed: profit earned on money invested.

WACA

Weighted average cost of acquisition: what holders paid per share on average.

IPO basicsFilingsJio

Maksud Ahmad Khan

Finance fresher and NISM Series XV Research Analyst, on the PGPM at ICFAI Business School, Mumbai. I write about IPOs, filings and Indian markets. Say hello.