How to Review a Startup Term Sheet in India (With a Worked Cap Table)

A term sheet is the two to five page document that sets out the commercial terms of a funding round before anyone drafts the real contracts. Almost every line carries a label, "non-binding," and most founders read that as "none of this matters yet." That is the most expensive misreading in Indian startup fundraising. In nearly every Indian term sheet, a handful of clauses, usually no-shop or exclusivity, confidentiality, and cost allocation, are expressly carved out as binding, even while valuation, the board seat, and the liquidation preference stay non-binding. Sign the page, and those few clauses bind you immediately, while the rest is still just a plan. (Adira, which publishes this guide, makes contract review and CLM software; we have a commercial interest in you understanding term sheets well, but this explainer is written to be useful whether or not you ever use our product.)

This guide covers what actually binds you on signing, then walks each economic and control term through one worked cap table example, so the numbers are not abstract, and ends with a red-flags table, a bad-to-better rewrite, and a checklist.

What actually binds you, and what does not

Most Indian term sheets open with a line like "this term sheet is an expression of intent and, save for the clauses stated below, does not create binding obligations on either party." The clauses stated below are the trap for a first-time founder. They typically include no-shop or exclusivity (no parallel fundraise for a defined window), confidentiality, and costs (you may owe the investor's legal and diligence fees if the deal falls apart on your side). Governing law and dispute resolution are sometimes on the list too.

The clearest recent illustration of how seriously Indian courts take this split is Oravel Stays Private Limited v Zostel Hospitality Private Limited (2025:DHC:3661), decided by the Delhi High Court on 13 May 2025. OYO and Zostel had signed a 2015 acquisition term sheet naming exactly five binding clauses, including exclusivity. When the deal collapsed, an arbitrator read the parties' later conduct as converting the whole document into a binding contract. The Delhi High Court set that award aside, holding that conduct cannot convert a term sheet into a fully binding contract where it expressly named only certain clauses as binding. Courts hold you to the named binding clauses and will not stretch the rest, however far talks progressed. Full case history and the binding no-shop mechanics: No-Shop and Exclusivity Clauses in Term Sheets.

Run this test on your own document: list every clause the "non-binding" paragraph exempts. Anything not on that list is not a binding promise. Anything on it is enforceable the moment you sign, not a future intention.

The worked example: one Series A round

One running example is used throughout. A startup with 80,00,000 fully diluted shares, held entirely by its founders (no existing ESOP pool), is raising a Series A. The term sheet offers Rs 8,00,00,000 (Rs 8 crore) for CCPS representing 20% of the fully diluted capitalisation post-round, a new ESOP pool of 10% of the post-round capitalisation created before Closing, and a 1x non-participating liquidation preference.

Valuation and dilution, the maths

Two headline numbers appear on almost every term sheet, and the second follows from the deal, it is not separately negotiated:

• Post-money valuation = Investment / Investor's percentage = Rs 8,00,00,000 / 20% = Rs 40 crore.
• Pre-money valuation = Post-money minus the new investment = Rs 40 crore minus Rs 8 crore = Rs 32 crore.

At 1,00,00,000 (1 crore) fully diluted shares post-round, that is Rs 40 a share; the investor's 20,00,000 shares cost Rs 8 crore at that price, and the math checks out. On the surface, founders keep 80% and the investor gets 20%. The next clause changes that.

The option pool shuffle

The term sheet also requires a new 10% ESOP pool "to be created prior to Closing." That phrase decides who actually pays for it. A pool created before Closing sits inside the pre-money cap table, so the founders alone absorb its dilution; the investor's 20% is calculated after the pool already exists, so the investor gives up nothing to fund it.

Run the numbers. Post-round: Investor 20,00,000 shares (20%), pool 10,00,000 shares (10%), Founders 70,00,000 shares (70%), total 1,00,00,000. At Rs 40 a share, the founders' 70,00,000 shares are worth Rs 28,00,00,000, not the Rs 32 crore the headline pre-money figure implied. The missing Rs 4 crore is the pool's value, and under a pre-money structure it comes entirely from the founders' side.

A pool created post-money instead carves the same 10,00,000 shares out of everyone's post-closing stake, investor included, roughly proportional to what each side holds, so founders keep closer to the full Rs 32 crore. The pool's size is rarely the fight worth having; whether it is pre-money or post-money is. Ctrl+F the term sheet for "prior" next to "ESOP" or "option pool," that is usually the tell.

Once created, the pool is subject to a minimum one-year gap between an option's grant date and its first vesting date under Rule 12(6)(a) of the Companies (Share Capital and Debentures) Rules, 2014, for a private company (Regulation 18(1) of the SEBI SBEB Regulations, 2021 for a listed one). Full ESOP and founder reverse-vesting mechanics, including the standard four-year, one-year-cliff schedule and good-leaver/bad-leaver terms: Founder and Employee Vesting Clauses in India.

Liquidation preference: who gets paid first

The 1x non-participating preference decides who is paid first on a sale, and by how much, and it often matters more than the headline valuation. Run the example at two exit prices.

Sale at Rs 25 crore (below the Rs 40 crore post-money valuation): the investor compares the flat Rs 8 crore preference to their as-converted 20% share of Rs 25 crore, Rs 5 crore, and takes the larger. They take Rs 8 crore off the top; the remaining Rs 17 crore goes to founders and the pool.

Sale at Rs 60 crore (above post-money): the investor compares Rs 8 crore against 20% of Rs 60 crore, Rs 12 crore, and converts to take the larger. The remaining Rs 48 crore splits among common holders. Non-participating gives the investor a floor, not a floor plus a slice of the upside, which is why 1x non-participating is the anchor position. A 2x or participating structure shifts both outcomes sharply in the investor's favour, at every exit price. India has no standalone statute called "liquidation preference"; it is built from Section 43 of the Companies Act, 2013 (defining "preference share capital") read with a June 2015 MCA exemption that lets private companies contract around the statutory default in their Articles. Full math across three preference structures and why an actual IBC liquidation ranks preference holders very differently from a sale: Liquidation Preference Explained.

Anti-dilution: protection against the next round

Anti-dilution protects the investor if a future round prices lower, a "down round." Our investor's CCPS convert at Rs 40 a share. Suppose a Series B, eighteen months later, issues 20,00,000 new shares at Rs 25 each. Using the standard broad-based weighted-average formula, NCP = CP1 x (A+B) / (A+C), where A is the 1,00,00,000 shares outstanding before the down round, B is the shares the Series B money would have bought at Rs 40 (12,50,000), and C is the shares actually issued (20,00,000):

NCP = 40 x (1,00,00,000 + 12,50,000) / (1,00,00,000 + 20,00,000) = 40 x 1,12,50,000 / 1,20,00,000 = Rs 37.5

The conversion price drops from Rs 40 to Rs 37.5, a modest adjustment. A full-ratchet clause on the same facts would reset it straight to Rs 25, nearly tripling the adjustment regardless of how small the round was, which is why weighted average, not full ratchet, is the term worth holding the line on. Where the investor is a non-resident, the down round's own price is separately gated: under Rule 21 of the FEMA (Non-Debt Instruments) Rules, 2019, a company issuing equity instruments to a person resident outside India must price them per an internationally accepted valuation methodology certified by a Chartered Accountant, Merchant Banker, or Cost Accountant. Full formula and carve-outs: Anti-Dilution Clauses Explained.

Board composition and reserved matters

A term sheet fixes board size and who nominates whom, commonly two founder nominees, one investor nominee, and one independent director, with a promise to add a seat at the next round. What constrains founders day to day is the reserved matters list, decisions the investor must separately approve despite a founder board majority: a budget deviation beyond a stated threshold, new debt above a set amount, changes to the ESOP pool, related-party transactions, a change in the nature of the business, and any amendment to the Articles.

That last item matters more than it looks. Section 5(3) of the Companies Act, 2013 lets a company's Articles carry "entrenchment" provisions:

"The articles may contain provisions for entrenchment to the effect that specified provisions of the articles may be altered only if conditions or procedures as that are more restrictive than those applicable in the case of a special resolution, are met or complied with." Source: Section 5, Companies Act, 2013 (Indian Kanoon)

A reserved matter living only in the shareholders' agreement (SHA) is a personal promise between the signing parties. The same protection entrenched in the Articles binds the company itself, and survives a director or share transferee who never signed the SHA. Check whether the term sheet promises the reserved matters will reach the Articles, or only a future SHA nobody has drafted yet.

Drag-along, tag-along, and right of first refusal

Three linked rights decide who can force, join, or block a future sale. A drag-along lets a majority shareholder force minority holders, including founders, to sell on the same terms once a buyer is accepted. A tag-along lets a minority holder join a sale the majority is making, at the same per-share price. A right of first refusal (ROFR) requires a selling shareholder to offer shares to existing shareholders first, at the third-party price, before selling outside. Watch the thresholds: a drag-along triggered just above 50% is far riskier for founders than one needing 75% plus board consent, and an ROFR response window under 15 to 30 days functions as a near-automatic waiver. Full mechanics: Drag-Along, Tag-Along, and Right of First Refusal.

No-shop, the binding part, and conditions to closing

No-shop is covered in full in its own guide, linked above, because it is usually the one clause on this document that is actually enforceable the day you sign. Check the duration (30 to 60 days is standard at the early stage) and the verb: "soliciting" competing offers is narrow, "responding to" one is broad enough to bar even a reply to an offer you never sought.

Everything else, valuation, board seats, liquidation preference, sits behind conditions precedent: facts that must become true before the round closes, board and shareholder approvals, satisfactory diligence, no material adverse change, and, for a foreign investor, FEMA compliance. Under Sections 31 and 32 of the Indian Contract Act, 1872, an obligation gated this way "cannot be enforced by law unless and until that event has happened." A CP list with no named owner and no long-stop date can leave a round in limbo while you stay bound by exclusivity. Full mechanics, including the Supreme Court's Nathulal v Phoolchand ruling on a party blaming the other side for its own unmet condition: Conditions Precedent, Explained.

Red flags

Normal	Red flag	Why it matters
Non-binding clause names exactly which provisions bind (usually no-shop, confidentiality, costs)	"Non-binding" stated with no clear list of exceptions	OYO v Zostel shows this exact ambiguity produces years of litigation
ESOP pool timing stated explicitly, pre-money or post-money	Pool "to be created prior to Closing," no discussion of who bears the cost	A pre-money pool is funded entirely from the founders' side, quietly lowering their real valuation
Liquidation preference is 1x, non-participating	1.5x, 2x, or participating with no cap	Every rupee above 1x, and every point of participation, comes straight out of what founders and ESOP holders receive
Anti-dilution is broad-based weighted average with an ESOP carve-out	Full ratchet, or no ESOP carve-out	A small down round, or a routine option grant, can trigger a disproportionate adjustment
Reserved matters are specific, with thresholds, promised in the amended Articles	Vague "customary reserved matters," or promised only in a future SHA	Thresholds get negotiated after signing, once leverage has shifted; a right outside the Articles may not bind the company
No-shop runs for a fixed number of days to a stated outside date	Open-ended "until Closing," no outside date	The company stays locked out of other conversations indefinitely if the deal simply stalls
Conditions precedent name a responsible party and a long-stop date	Generic "customary closing conditions," no long-stop date	A deal can sit half-closed indefinitely while exclusivity and cost obligations keep binding you
Drag-along threshold is a supermajority (commonly 75%) with a floor price	Drag-along triggers just above 50%, no floor price	A bare majority can force a sale on terms a large minority never agreed to
Instrument is CCPS or CCD, fully and compulsorily convertible, stated upfront	Instrument unspecified, or optionally convertible for a foreign investor	A partially convertible instrument can be reclassified as debt (ECB) under FEMA, a stricter regime
Break fee tied to actual, documented, capped costs	Flat lump-sum break fee unrelated to real cost	Reads as a penalty under Section 74 of the Contract Act, cut down to reasonable compensation, but only after a dispute

Bad clause, better clause

Bad (a composite of language common in investor-drafted Indian term sheets): "This Term Sheet is non-binding and subject to definitive documentation. The Company shall create an ESOP pool of 10% of the fully diluted capital prior to Closing. The Investor shall be entitled to a 2x participating liquidation preference. Closing is subject to satisfaction of customary conditions precedent."

Better: "This Term Sheet is non-binding in its entirety, save for Clauses 9 (Exclusivity), 11 (Confidentiality), and 12 (Costs), which shall be binding on execution. The ESOP pool shall be sized at 10% of the fully diluted capitalisation calculated on a post-money basis, any shortfall funded pro rata by all shareholders (including the Investor) rather than solely by the Founders. The Investor shall be entitled to a 1x non-participating liquidation preference. Closing is subject to the following conditions precedent, each to be satisfied by [date] ('Long Stop Date'): (a) board and shareholder approval, within 15 business days of signing; (b) legal, financial, and business due diligence to the Investor's reasonable satisfaction; (c) no Material Adverse Change between signing and Closing. If any condition is not satisfied or waived by the Long Stop Date, either party may terminate on written notice."

What changed: the binding-clauses list is named, not implied; the pool moved from pre-money (founders-only cost) to post-money (shared cost); the preference dropped from 2x participating to 1x non-participating; and "customary conditions precedent" became three named conditions with owners and a hard long-stop date.

India execution notes: instrument choice and FEMA

Indian rounds almost always use CCPS, Compulsorily Convertible Preference Shares, rather than plain equity, because CCPS lets the parties attach a liquidation preference, anti-dilution, and other bespoke rights that the Companies Act's default rules for ordinary equity do not permit. CCDs, Compulsorily Convertible Debentures, show up more in bridge rounds, where the instrument sits closer to debt on the balance sheet until conversion, or the investor wants a coupon before then. Under the FEMA (Non-Debt Instruments) Rules, 2019, only an instrument fully and mandatorily convertible into equity within a specified period qualifies as an "equity instrument" for a foreign investor; anything less, redeemable or partially convertible, is treated as debt under India's External Commercial Borrowing framework, a stricter regime on pricing and tenure. Check the conversion clause for "compulsorily and fully convertible" before assuming an instrument qualifies as equity.

Two further checkpoints matter on a foreign-led round. The price per share must meet the arm's-length valuation standard under Rule 21 above; a founder-friendly discount to a non-resident investor below that floor is not something the parties can simply agree around. And once shares are allotted, the company must report the issuance to the RBI, typically filing Form FC-GPR within 30 days, under the Foreign Exchange Management (Mode of Payment and Reporting of Non-Debt Instruments) Regulations, 2019. Missing this filing does not undo the round, but it complicates every later round and any exit until regularised. Source: RBI Master Direction on Reporting under FEMA.

The actual share issuance, once terms are agreed, happens through a preferential allotment under Section 42 of the Companies Act, 2013, which defines a private placement as an "offer of securities or invitation to subscribe securities to a select group of persons by a company (other than by way of public offer)" meeting the section's conditions. Source: Section 42, Companies Act, 2013 (Indian Kanoon). That filing, and the domestic rule requiring a registered valuer's report on price, run alongside the FEMA steps above wherever the round mixes resident and non-resident investors.

US and global contrast

The vocabulary, pre-money and post-money, liquidation preference, anti-dilution, drag and tag, is identical to Silicon Valley term sheets; Indian practice borrowed the structure directly. The US also has a widely standardised free template (the NVCA term sheet, and SAFEs or convertible notes for seed rounds) that most investors and lawyers already know clause for clause, shortening negotiation. India has no equivalent single standard, and the CCPS/CCD choice adds a layer of Companies Act and FEMA mechanics a US Delaware C-corp round, using plain preferred stock under one statute, never has to clear. The binding split is sharper here too: Indian practice, reinforced by OYO v Zostel, puts real weight on the term sheet naming its binding clauses precisely, not leaving the split to be inferred.

FAQ

Is a term sheet legally binding in India? Mostly no. Most Indian term sheets state that only certain named clauses, typically no-shop, confidentiality, and costs, are binding on signing, while valuation, board seats, and liquidation preference wait for definitive agreements. Read the "non-binding" paragraph to see exactly which clauses it exempts.

What is the option pool shuffle and why does it matter? A new ESOP pool created "pre-money" comes entirely out of the founders' existing shares, before the investor's percentage is calculated. Founders end up with less real value than the headline pre-money figure implies, even though the percentages look unchanged. Ask whether the pool is pre-money or post-money before agreeing to the valuation.

What liquidation preference is standard in an Indian Series A? 1x, non-participating, is the founder-friendly market standard. A multiple above 1x, or a participating structure, pays the investor significantly more at every exit price and is worth negotiating, not accepting as boilerplate.

Should our term sheet use CCPS or CCD? CCPS is the default for most priced equity rounds; CCDs are more common for bridge financing. Either instrument must be fully and compulsorily convertible to count as equity for a foreign investor, or it risks reclassification as debt under FEMA's External Commercial Borrowing rules.

How long should the exclusivity or no-shop period be? Typically 30 to 60 days at the early stage, tied to a fixed outside date or the signing of definitive agreements, not left open-ended, and named on the term sheet's short list of clauses that actually bind you.

Do reserved matters need to be in the Articles, or is the term sheet enough? Ideally both. A veto right existing only in a future SHA, never entrenched in the Articles under Section 5(3), may bind only the original signing parties personally, not the company or a future share transferee.

You can mark up a term sheet clause by clause, model the cap table math above against your own numbers, and flag which clauses are binding, for free, in Weave, before you send comments back to the other side.

This guide gets you to a working understanding of what a term sheet contains, what actually binds you on signing, and how the main economic and control terms interact using worked numbers. It does not tell you whether your specific valuation, preference multiple, or reserved matters list is a fair deal for your round, that depends on your negotiating leverage, comparable deals, and facts a lawyer needs to see. Talk to a lawyer before you sign a term sheet, and before you rely on any clause in it being binding or non-binding.