Venture Capital and Private Equity in India

Learning Outcomes

Modeling the downside cases here should leave the reader able to:

Compute exit proceeds under participating and non-participating preference, across a stack of several classes at different seniorities.

Model broad-based weighted average, narrow-based weighted average, and full-ratchet anti-dilution, and quantify what each costs the founders.

Explain the function of pay-to-play and identify the Indian mechanism that performs the same work.

Evaluate an economic term sheet for its behavior in a downside case instead of its headline valuation.

In the Room

Picture the call. The scene is a composite drawn from practitioner accounts. No company, person or transaction in it is real, and no figure in it is a market statistic.

The call was scheduled for four o’clock and the founders had been told it was good news. Their company had signed a sale to a strategic acquirer at a price that would be reported the following week as a successful exit for an Indian consumer brand. The two of them had spent 6 years on it.

The chief financial officer had the waterfall open on her screen and had already run it three times, because the first time she had assumed she had made an error. The company was selling for ₹28 crore. The preference stack sitting above the equity shares came to ₹45 crore.

She explained it in the order the document required. The Series C investors, who had put in ₹10 crore 2 years earlier at the bottom of a down round, were senior, and they would be paid in full. Next in the queue stood the Series B, ₹15 crore committed at the top of the market, also paid in full. That left ₹3 crore for the Series A investors, who had put in ₹20 crore and had been in the company longest.

One of the founders asked what was left after that. The answer was nothing, and the second question, which came more slowly, was about the option pool. Forty-one employees held vested options in a company that had just been sold, and every one of those options was attached to an equity share standing behind ₹45 crore of preference. The pool was worth nothing either.

Neither founder had negotiated the seniority provision. It had appeared in the Series C term sheet as a single line, in a round they had been relieved to close at all, and nobody in that negotiation had modeled what it did at an exit below the preference stack. The term that decided the outcome was not the valuation they had argued about for 3 weeks. It was a clause about payment order that took ninety seconds to agree.¹

This chapter is about the terms that decide who receives what, and about how they only become visible below a price nobody in the room expects to see.

18.1 Liquidation Preference: Multiple, Seniority, and Stacking

A liquidation preference is a contractual right to be paid a stated amount, out of the proceeds of a sale, before the equity shares receive anything. It is the single most consequential economic term in a venture financing, and it is routinely negotiated with less attention than the valuation, which is the more visible number and the less important one.

Three variables define it. The multiple states how much the class receives before anything flows to the equity, expressed as a multiple of the amount invested. The seniority states where the class sits in the payment order relative to other preference classes. Participation, taken up in the next section, states whether the class also shares in what remains after preferences are paid.

The multiple is the variable founders watch. A one-times preference returns the investment; a two-times preference returns twice the investment before anything else moves. Indian venture rounds are predominantly written at one times, and a multiple above one times normally signals something about the company’s circumstances and very little about the investor’s appetite.²

Seniority is the variable founders miss, and the scene above is the reason. Two arrangements dominate. Under a pari passu or pooled arrangement, all preference classes rank equally, and where the proceeds are insufficient each class receives the same proportion of its claim. Under a stacked or standard seniority arrangement, the classes are paid in reverse order of investment, so the most recent money is paid first and the earliest money is paid last.

The distinction is invisible at any exit above the total preference stack, where every class is paid in full and the ordering has no work to do. It becomes decisive the moment the proceeds fall short. Take Company B, the consumer business in the fund’s portfolio, at the ₹28 crore sale described above, with ₹45 crore of preference across three classes.

Under stacked seniority the Series C receives its ₹10 crore in full, the Series B receives its ₹15 crore in full, and the Series A receives the remaining ₹3 crore against a claim of ₹20 crore. The Series A investor recovers 15 percent of its money. Under a pari passu arrangement the same ₹28 crore is spread across the three classes in proportion to their claims, so each recovers 62.2 percent, and the Series A investor receives ₹12.44 crore instead of ₹3 crore.³

The aggregate is identical in both cases, and the distribution inside the stack is not. That matters enormously when the classes sit with different investors and not at all when they sit with the same one. In Company B a single fund holds all three preference classes, which is why the fund’s proceeds are ₹28 crore under either arrangement and why the seniority clause, in that particular company, changed nothing for the party that negotiated it.

Stacking also compounds across rounds. Each new round is normally granted seniority over everything before it, so a company that raises four times has built a payment queue in which the last investor stands at the front. The founder who signed the first such provision agreed to a principle whose full cost arrives two rounds later. That cost is entirely predictable at the time of signing.

One Indian complication belongs here and is developed in Chapter 19. Indian rounds are issued as compulsorily convertible preference shares under the Companies Act, and the preference entitlement lives in the articles of association as well as in the shareholders’ agreement.⁴ A liquidation preference expressed only in a contract, with no corresponding article, invites an argument at the moment when the parties have the least goodwill available for one.

18.2 Participation, and the Caps That Soften It

A non-participating preference forces a choice. The holder takes its preference amount, or it converts to equity and takes its percentage of the whole, and it cannot do both. A participating preference removes the choice: the holder takes its preference amount and then also shares in the residue alongside the equity, on an as-converted basis.

Participating preferred is often called double-dipping, which is accurate and unhelpfully pejorative. The useful description is a device that changes the shape of the investor’s return curve. Under non-participating terms the payoff sits flat at the preference amount until the conversion indifference price and then rises with the company. Under participating terms it rises from the first rupee above the preference stack, improving the investor’s outcome at every exit value without exception.

A cap limits that improvement. A participating preference capped at two times means the holder stops participating once its aggregate receipts reach twice its money. Above the cap the holder must decide whether to abandon the preference and convert, which restores the choice that participation removed.

The behavior of these three structures is not intuitive, and Company B makes the point better than a description can. The company has ₹45 crore of preference across three classes, a fund holding 49.4 percent as converted, and the remaining 50.6 percent split between founders at 45.0 percent and the option pool at 5.6 percent. Run the same company through four preference structures at four exit values and the results are these.⁵

Founders’ proceeds, in ₹ crore:

The fund’s multiple on its ₹45 crore, under the same structures:

Those tables reward three readings.

The first is that the entire top half of both tables is identical. At ₹28 crore and at ₹45 crore every structure produces the same answer, because the proceeds never exceed the preference stack and nothing reaches the equity under any set of terms. A founder negotiating participation while the company heads for an exit below its own preference stack is negotiating something that will not matter. The multiple and the seniority decide a bad outcome; participation decides only how a good one is divided.

The second is that at ₹200 crore the capped participating structure leaves the founders better off than plain one-times non-participating terms, at ₹97.78 crore against ₹90.00 crore, and leaves the fund worse off, at 2.00 times against 2.19 times. That reverses the usual intuition that participating terms are always harsher. The cap stops the investor dead at twice its money, while non-participating terms let the investor abandon its preference, convert, and take 49.4 percent of everything, which at a large enough exit is worth considerably more than any cap.

The third is that the capped structure is unstable at the top of the range. At ₹200 crore the Series C holder, capped at ₹20 crore, does better by converting, and it will convert. Be careful how that figure is computed. Twenty-five percent of ₹200 crore is ₹50 crore, and that is the wrong answer, because Series A and B take ₹35 crore off the top before anything reaches the common pool. The right figure is about ₹43.6 crore, and the shortcut that produces ₹50 crore is the one Section 18.5 warns against. Every other class’s arithmetic then changes, because the residue and the share count both move. Resolving the position properly means solving all three classes at once, since each holder’s best action depends on what the other two decide. The companion model flags the decision and declines to fold it in. The tab says why.⁶

That circularity explains why capped participating preferred produces disputes at exit. Reasonable parties, working from the same document, reach different distributions depending on the order in which they assume the conversions happen.

The market has largely moved away from participation in the United States. The Cooley quarterly venture financing data put non-participating preferred at 96 percent of transactions in the fourth quarter of 2025, which makes participation a distress signal in that market.⁷ Indian practice follows the same convention in the ordinary case, and the Indian evidence base for that statement is a matter this chapter returns to in What Is Not Public.

18.3 Anti-Dilution in Its Three Forms, Modeled

Anti-dilution protects an investor against a subsequent round priced below the one it participated in. The protection operates by adjusting the conversion price at which the earlier investor’s preference shares convert into equity, which increases the number of equity shares the earlier investor receives, which dilutes everybody without anti-dilution protection. In practice that means the founders and the option pool.

Three formulations are in general use, and the distance between them is far larger than the distance between their names suggests.

Full ratchet resets the earlier investor’s conversion price to the price of the new round, without regard to how small the new round is. An investor who paid ₹150 per share, in a company that subsequently sells one share at ₹30, converts at ₹30.

Broad-based weighted average resets the conversion price to a weighted average of the old price and the new one, with the weights reflecting the size of the new issue against the company’s existing fully diluted capital, including the option pool and other convertible securities.

Narrow-based weighted average applies the same formula on a smaller denominator, excluding the option pool and typically other unissued securities, which produces a slightly lower adjusted price and slightly more shares for the protected investor.

Company B’s Series C is the natural test, because it was a steep down round. The Series B had been issued at ₹150.00 per share in 2021. The Series C was issued at ₹30.00 per share in 2023, a fall of 80 percent, raising ₹10 crore against a fully diluted base of 10,000,000 shares.

The adjusted conversion prices are these.⁸ Under broad-based weighted average the Series B price falls from ₹150.00 to ₹120.00, and the Series A price falls from ₹88.89 to ₹74.17. Under narrow-based weighted average the same prices fall to ₹118.21 and ₹73.29. Under full ratchet both classes reset to ₹30.00.

The consequences for ownership are these:

The teaching point sits in the bottom two rows and it is a matter of orders of magnitude. Weighted average anti-dilution, in either form, costs the founders slightly over 2 percentage points. Full ratchet costs them 17.41 percentage points, and moves the fund from a minority holding of 49.4 percent to a controlling 69.0 percent, on a ₹10 crore investment.

A founder negotiating anti-dilution should therefore spend the available negotiating capital on one question, which is whether the provision is a ratchet or a weighted average. The broad-based against narrow-based distinction, which occupies a surprising amount of space in the practitioner literature, is worth 0.16 percentage points on these facts and does not merit an argument.

Two further observations belong here.

The first is that a single share issued at a lower price triggers a full ratchet, which gives it a mechanical severity out of proportion to the economics that provoked it. A ₹50 lakh bridge can reset a ₹15 crore position. Practitioners address this with carve-outs, and the carve-out list is where the real negotiation happens: option issuances, conversions of existing securities, shares issued in acquisitions or to lenders and strategic partners, and issues approved by the protected class itself.

The second is that anti-dilution and the down round interact with the option pool in a way that damages retention exactly when retention matters most. The ratchet dilutes the pool along with the founders, at a moment when the company has just repriced downward and needs to hold its people. A board that adjusts the pool back up after a ratchet has diluted it is issuing fresh options into a lower valuation, which is the correct action and which dilutes the founders a second time.

18.4 Pay-to-Play and Its Effect on Investor Behavior

A pay-to-play provision requires an existing investor to participate in a future round, in proportion to its holding, on pain of losing something it currently has. The penalty varies. In its severest form the non-participating investor’s preference shares convert into ordinary equity, which strips the preference, the anti-dilution protection, and normally the protective provisions in a single step. Milder forms convert the holder into a shadow class with reduced rights, or simply cancel the anti-dilution adjustment for that holder.

The provision exists because of a specific failure mode. In a down round, an investor holding a strong anti-dilution right has an incentive to decline to participate, since the ratchet increases its ownership without requiring new money, and the new capital arrives from someone else. If several existing investors reason that way at once, the round does not get funded, and a company that could have been saved is not. Pay-to-play removes the incentive by attaching a cost to standing aside.

Its effect on investor behavior runs in two directions. It disciplines the portfolio at the moment of a down round, forcing each holder to decide explicitly whether the company deserves more capital, and a holder unwilling to fund a company arguably should not retain a senior claim on that company’s eventual sale. It also imports a portfolio constraint that has nothing to do with the company, since a fund at the end of its investment period, or one holding no reserves against the position, may be unable to participate however strongly it believes in the business. The provision punishes an inability to pay identically to an unwillingness to pay, and a fund’s reserve policy therefore becomes a term in the negotiation of somebody else’s financing.

No Indian company has been publicly documented operating a contractual pay-to-play mechanic. No filed or reported instance has been located, which is a finding about the Indian public record and not a claim that the clause is absent from Indian term sheets.⁹

What India has instead is a company law mechanism that performs the same function through a different door. A rights issue under the Companies Act offers new shares to existing shareholders in proportion to their holdings, and a shareholder who declines is diluted by whoever takes up the unsubscribed portion. Where the rights issue is priced far below the last round, the dilution suffered by a non-participating shareholder is severe enough that the choice becomes formal.

Byju’s rights issue of approximately $200 million in early 2024 is the clearest Indian illustration, and this chapter’s India cases treat it with the distinctions it requires.¹⁰

The functional equivalence is real and the legal architecture differs, and a student who conflates the two will misread both. A contractual pay-to-play is a bargain among shareholders, enforced through the conversion mechanics of the preference shares. A rights issue is a statutory offer made by the company, subject to company law process, and open to challenge on grounds a contractual mechanic would never face.

18.5 Reading an Economic Term Sheet for Its Downside Behavior

The habit this chapter is designed to build is simple to state and rare to observe. An economic term sheet should be read at exit values below the last round’s post-money valuation, because that is the range in which its provisions do work, and because the range above it resolves trivially.

Four questions extract most of what matters.

What is the total preference stack, in rupees, against the amount the company would realistically sell for today? That ratio is the single most informative number in the document. A company with ₹45 crore of preference and a realistic sale value of ₹28 crore has a worthless common equity, and every incentive conversation with employees and founders should proceed from that fact instead of around it.

What is the payment order, and who holds each position in it? Seniority is invisible in good outcomes and decisive in bad ones, and a founder should be able to name which investor stands first in the queue.

At what exit value does each class stop taking its preference and convert? Those indifference prices are computable from the document alone and they mark where each investor’s interests change direction: below its indifference price an investor prefers a certain sale at a low number, and above it the same investor wants the company to hold out. A board whose members sit at different points on that curve will disagree about a sale offer for reasons having nothing to do with judgment.

What does the anti-dilution provision do to the founders’ position on a 50 percent down round? That is a five-minute calculation, and it turns an abstract clause into a number the founders can react to.

One caution about method belongs at the end of this section, because it affects every calculation above.

The conventional way to test whether a preference class converts is to compare its preference amount against its ownership percentage multiplied by the exit value. That test is a shortcut, and it is wrong whenever the preference stack is large relative to the exit, because it ignores that the other classes’ preferences come off the top before the converting class receives anything.

On Company B the shortcut gives the wrong answer twice out of sixteen tested cells, and it errs in the same direction both times. At a ₹45 crore exit under one-times non-participating terms, the shortcut says the Series C converts and the founders receive ₹5.95 crore. Testing the class against the residue actually available to it shows the Series C taking its preference and the founders receiving nothing. At a ₹100 crore exit under two-times terms, the shortcut hands the founders ₹17.85 crore where the correct figure is ₹8.89 crore.¹¹

Both errors overstate what the equity receives, which is a plausible reason the shortcut survives: nobody in the room has an incentive to object to it.

Framework: The Preference Stack

The Preference Stack resolves a set of preference classes into a payment order and a set of switching points, so that the distribution at any exit value can be computed and, more usefully, so that the exit values at which each party’s interests change can be named in advance.

Step one: list the classes with their three defining terms. For each class, record the amount invested, the multiple, the seniority position, and whether it participates. Record the as-converted ownership percentage separately, because it drives the conversion decision and has no bearing on the preference amount.

Step two: total the stack. Sum the preference amounts across all classes at their stated multiples. Below that figure the equity shares receive nothing, whatever structure the parties agreed. It is the most important single number in the term sheet and it appears nowhere in the term sheet.

Step three: compute each class’s conversion indifference price. For a non-participating class, divide the preference amount by the as-converted ownership percentage. On Company B this gives ₹118.52 crore for the Series A, ₹200.00 crore for the Series B and ₹40.00 crore for the Series C. The spread across those three numbers is the spread of interests on the board.

Step four: resolve the waterfall in seniority order, then re-test. Pay the non-converting classes in seniority order to the limit of the proceeds, distribute the residue across the equity and the converted classes, then re-test each conversion decision against the residue that class actually faced. Repeat until the decisions stop changing. Two passes are normally enough; three are always enough for a stack of three classes.

Step five: read the switching points as a governance map. Each indifference price is an exit value at which one investor’s preferred outcome flips. A sale offer that sits between two indifference prices has one investor arguing to accept and another arguing to refuse, and neither is behaving badly.

Where the Stack misleads. It assumes the conversion decisions can be resolved sequentially, which holds for non-participating classes and fails for capped participating ones, where each holder’s optimal choice depends simultaneously on every other holder’s. It also treats the preference amounts as fixed, when accruing dividends, ratchet adjustments, and unpaid coupons can move them between signing and exit. And it says nothing about whether the sale happens at all, which in a company below its preference stack is the live question, because the parties who control the decision are the parties who will be paid.

Framework: The Downside Test

The Downside Test stress-tests an economic term sheet at four exit values before it is signed. Its purpose is to replace a discussion about the headline valuation with a discussion about the distribution, which is the thing being negotiated whether or not anybody says so.

Value one: the total preference stack. Run the waterfall at an exit exactly equal to the sum of the preferences. Every rupee goes to the preference classes and the equity receives nothing. This value is the floor of the founders’ economic interest, and naming it out loud changes the tenor of a negotiation more than any other single number.

Value two: the realistic sale price today. What a strategic acquirer would plausibly pay for the business as it currently trades, which is neither the last round’s valuation nor the plan. If this value sits below value one, the common equity is worthless today and the option pool retains nobody.

Value three: the last round’s post-money valuation. At this value a one-times non-participating class created in that round is exactly indifferent between its preference and conversion. Above it the waterfall resolves toward simple proportional ownership; below it the document’s machinery engages. This is the boundary the term sheet was written for.

Value four: a good outcome, defined as three to five times the last post-money. At this value the structures separate most sharply, and the counterintuitive results appear. On Company B this is where capped participation turns out to favor the founders over plain non-participating terms.

How to use the four values. Build the founders’ proceeds and each investor’s multiple at all four, under each structure on the table. The output is a grid, and the grid is the negotiating document. A founder arguing about a valuation is arguing about one cell of it.

Where the Test misleads. It prices a single sale at a single moment, which misdescribes an exit that occurs in tranches or through a listing where preferences convert automatically on a public issue. It also assumes the parties behave as the arithmetic predicts, and a holder with a portfolio consideration, a fund life ending, or a reputational interest in a founder relationship may act against its own computed interest. The grid establishes what each party gains from each outcome. Predicting what they will actually do requires the material in Chapter 19.

India Cases

PharmEasy: The Down Round That Repriced the People Who Were Not in It

PharmEasy’s parent raised a rights issue in 2023 at a valuation reported as roughly 90 percent below its previous round. The transaction has been reported as triggering anti-dilution protection in existing investors’ agreements, with the consequence that investors from earlier and higher-priced rounds received additional shares.¹²

The case teaches the mechanism cleanly because the effect runs in a direction students find counterintuitive. The investors who suffered the largest paper loss on the down round were, through the operation of their own contractual protection, the parties whose share counts increased. The dilution landed on the holders without protection: the founders, the employee option pool, and any investor whose agreement carried a weaker formula.

Two features earn it a place over a cleaner foreign example. It is Indian and recent, and it was large enough that the transaction was reported in detail. And it demonstrates the interaction this chapter’s Section 18.3 closes on, since a company that has just repriced downward by 90 percent and simultaneously diluted its option pool through a ratchet is a company facing a retention problem created by its own financing document.

The limits of the public record here should be stated. The specific anti-dilution formulations in the individual agreements have not been published, the share counts before and after the adjustment are not disclosed, and the reporting does not distinguish which investors held ratchets from which held weighted average protection.

Byju’s: India’s Functional Equivalent of Pay-to-Play, and What It Is Not

Byju’s parent conducted a rights issue of approximately $200 million in early 2024 at a price reported to represent a discount of roughly 99 percent to its peak valuation of $22 billion. A shareholder who declined to participate faced dilution of a magnitude that made the offer effectively compulsory.10

The teaching value lies in the functional equivalence and in the legal difference, and both halves are necessary.

Functionally, the transaction did what a pay-to-play provision does. It confronted every existing holder with a decision to fund or be crushed, and it removed the option of standing aside while retaining a position.

Legally, it was something else entirely. A rights issue is a statutory offer by the company to its existing shareholders under the Companies Act, subject to the process requirements that attach to it, and open to challenge on grounds of oppression and mismanagement that a contractual mechanic among shareholders would not face. Investors did challenge it, and that challenge remains contested and under appeal.10

Two cautions govern the use of this case. The oppression proceedings and the subsequent orders are contested and should be taught as contested. And the company’s separate Term Loan B default, where the position is settled and was affirmed on appeal in the United States, is a different matter belonging to Chapter 20; students conflate the two, and the two should be kept apart.

Global Cases

In Re Trados: The Preference That Consumed the Exit

In re Trados Incorporated Shareholder Litigation, decided by the Delaware Court of Chancery in 2013, is the case the Indian record cannot supply. A company was sold for $60 million, $50 million of it in cash and $10 million in the buyer’s stock, against a preference of $57.9 million. On those two numbers the common stockholders should have seen something. They saw nothing, because a management incentive plan negotiated by the very directors the common stockholders would later sue took the first $7.8 million off the top, 13 percent of the proceeds, before the preference was reached at all. That left about $52.2 million against the preferred claim and nothing whatever below it. The preference ran above the money the investors had put in because it carried accumulated dividends, which is the detail that reconciles the arithmetic for a student who tries to rebuild it. The common stockholders then sued the directors, most of whom had been designated by the preferred investors.¹³

Note what actually did the damage. The proceeds covered the preference. What emptied the common stockholders’ share was an instrument ranking ahead of the preference, and no term sheet in this chapter has yet mentioned such a thing.

The court’s analysis is what raises the case above illustration. It held that directors owe their duties to the corporation and its common stockholders, and that a board dominated by preferred designees faces a conflict when approving a transaction paying the preferred and nothing to the common. It then found the transaction entirely fair notwithstanding, because the common stock had no economic value before the sale, so the common stockholders received exactly what their shares were worth.

That second holding is the one to dwell on. A board can approve a sale that pays the investors and leaves the founders and employees with nothing, and be correct to do so, if the common was already worthless. The remedy for the common stockholder in that position is not litigation after the sale. It is the term sheet, several years earlier.

The case pairs with this chapter’s opening scene on the same mechanism, and the pairing is the point of teaching them together: the Indian version happens, and it settles in confidence, so the reasoning is only available in a Delaware opinion.

Square: The Ratchet That Triggered in Public

Square, the payments company now named Block, Inc., carried in its Series E financing a provision guaranteeing its investors a return at the initial public offering, expressed as a minimum price against the approximately $15.46 per share they paid. The company listed in November 2015 at $9.00 per share. The guarantee triggered, additional shares were issued to the Series E holders, and every other holder was diluted accordingly. The mechanism and its consequences were disclosed in the registration statement.¹⁴

The case earns its place for a reason unrelated to its size. It is a ratchet whose operation is visible in a filed document, at a company whose share price is public, with a before and after that can be read off the record. Ratchets normally operate in private financings and their effects are never disclosed. This one operated in the transition to a public market and left a record.

What the case teaches beyond itself is who bears the cost. The ratchet protected the last investor before the listing. The dilution fell on the founders, the employees, and every earlier investor, and it fell at the precise moment those holders were being asked to accept a lower price than they had expected. Downside protection for one holder is downside amplification for every other holder in the same company, and the term sheet never states it in those words.

The Cooley Data: Participation as a Distress Term

Cooley publishes quarterly data on the venture financings it handles, including the proportion carrying participating preferred. The series shows non-participating preferred at approximately 96 percent of transactions in the fourth quarter of 2025.7

This is the book’s market-data case, and it does something no anecdote can. A single transaction with participating preferred establishes that the term exists; a quarterly series across a large deal population establishes what the term means when it appears, which is that the company was in a weak position, since 96 percent of its peers did not concede it. The series also supports a claim about direction that a snapshot cannot, since participation was common in the early 2000s and has become rare.

Two limitations belong with the case. The population is one firm’s deal flow, which skews toward institutionally financed United States companies, and no equivalent Indian series is published, so the same claim about Indian practice rests on practitioner convention and cannot be checked.

Across the Table

The role: a partner negotiating the economics of a term sheet against a competing offer from another fund.¹⁵

A competing term sheet changes what is being negotiated, because the founder now has a comparison and will make it on the headline valuation unless given a reason not to. The partner’s task in that situation is to move the conversation from the valuation to the distribution, which is where the two offers actually differ and where the higher-valuation offer is frequently the worse one.

The most effective move is arithmetic performed in front of the founder. A term sheet at a ₹120 crore pre-money with participating preferred and a full ratchet, set against one at ₹95 crore with one-times non-participating and broad-based weighted average, is a comparison no founder can make by inspection. Building the grid together, at four exit values, converts a debate about which number is larger into a discussion about which outcomes each party is underwriting.

A partner who wins on structure and loses on price has usually done the better deal, and should be willing to say so out loud. Conceding valuation to hold non-participating terms and a weighted average ratchet is a trade the fund’s own returns will vindicate at most exit values. It also builds the relationship, since the founder who later understands what the alternative structure would have cost remembers who proposed which.

The provision to hold hardest is the anti-dilution formula, and the provision to concede first is usually the option pool’s size. Full ratchet costs a founder 17 percentage points on the arithmetic in Section 18.3 and generates resentment that surfaces at the next round. Pool sizing is a negotiation about a number that will be revisited within 18 months in any case.

The question a partner should ask before signing is what the deal looks like at the fund’s median outcome. Term sheets get negotiated against the case in which the company succeeds, and portfolios are made by what happens in the middle. A structure that produces an acceptable answer only at a ten-times exit has not been stress-tested; it has been hoped over.

Applied Exercise

Objective. Model Company B’s exit under four preference structures, identify each class’s switching point, and quantify what anti-dilution costs the founders.

Companion site. The Term Sheet Econ model, where the four preference structures can be run against any exit value.

Steps.

Reproduce Company B’s capitalization at the point of sale: 13,333,333 shares fully diluted, founders at 45.0 percent, an option pool at 5.625 percent, and three preference classes totaling ₹45 crore of invested capital.

Compute each preference class’s conversion indifference price on a standalone basis, and explain why the Series C figure of ₹40.00 crore equals its round’s post-money valuation while the Series A and Series B figures do not equal theirs.

Run the waterfall at ₹28 crore under stacked seniority and again under a pari passu arrangement. Confirm that the fund’s aggregate proceeds are identical and state which party the seniority clause actually protects.

Build the founders’ proceeds grid at exit values of ₹28 crore, ₹45 crore, ₹100 crore, and ₹200 crore, across all four preference structures. Identify the cell in which capped participating terms leave the founders better off than one-times non-participating terms, and explain the mechanism.

Apply the market-convention conversion test to all sixteen cells and compare against the iterated test. Locate the two cells where the answers diverge, and state the direction of the error.

Model the Series C down round three ways under broad-based weighted average, narrow-based weighted average, and full ratchet. Record the founders’ ownership under each, and express the difference between the ratchet and the weighted average in percentage points.

Assume the board refreshes the option pool back to 5.625 percent after a full ratchet has diluted it. Compute the founders’ ownership after that refresh, and state who paid for it.

The question the exercise poses. A founder is offered two term sheets: ₹120 crore pre-money with participating preferred and a full ratchet, or ₹95 crore pre-money with one-times non-participating and broad-based weighted average. Model both on Company B’s financing history and write one paragraph identifying the exit value above which the founder prefers the lower-valuation offer.

What Is Not Public

No Indian participating-preference waterfall is available in the public record. Indian exits settle under confidentiality, the share purchase agreements are not filed, and even the practitioner literature that discusses Indian preference terms names no company and concedes that the law on several of the underlying questions is unsettled.¹⁶

No Indian pay-to-play mechanic has been publicly documented at all. The clause appears in Indian term sheets by practitioner account, and no filed or reported instance of one operating has been located.

The consequence for this chapter is structural. The participating-preference teaching load sits on a Delaware opinion from 2013, and the Indian version in Section 18.2 is a constructed hypothetical built on real Indian term-sheet conventions and clearly labeled as such. A reader should understand that the Indian arithmetic in this chapter is correct and that its Indian factual base is thinner than its foreign one.

What India does disclose, and what makes the chapter’s other half sound, is the preference structure itself. A company approaching a listing discloses its preference series, their conversion terms, and their entitlements in the offer document, which is more than a United States private company ever discloses before its own registration statement. India conceals the exit. It reveals the instrument.

What Carries Forward

Every term in this chapter is a rule for a range of exit values nobody in the room expects to see. That is why each is agreed in minutes and decides outcomes years later. The multiple and the seniority decide a bad exit, participation decides how a good one is divided, and anti-dilution decides who pays for a bad round. Downside protection for one holder is downside amplification for every other holder in the same company. The total preference stack is the number that decides a bad outcome, and it appears in no term sheet. It is the sum of the preference amounts at their stated multiples, and no structure pays the equity a rupee below it. Seniority is invisible above the stack and decisive below it, which makes it a term the founders can concede cheaply and the existing investors cannot. Participation changes the shape of the return curve. The American market has largely moved away from it, with non-participating preferred at 96 percent of transactions in the Cooley data. Anti-dilution is three formulations whose names are far closer together than their effects. A full ratchet and a broad-based weighted average produce different companies out of the same down round, and the ratchet dilutes the option pool alongside the founders at the moment the company can least afford to lose people. Pay-to-play exists because an investor with a strong ratchet may decline to fund a down round, since the ratchet pays it either way. The Indian functional equivalent differs in legal architecture. The habit the chapter installs is to read an economic term sheet at exit values below the last round’s post-money valuation, because that is the range where its provisions do work.

Chapter 19 turns from what the investor is paid to what the investor can stop. For a minority holder in India the veto layer is contractual almost in its entirety, and the chapter shows why the same clauses behave differently in Delaware, London and Mumbai.

Key Terms

Liquidation preference. A contractual right to be paid a stated amount out of sale proceeds before the equity shares receive anything. Defined by its multiple, its seniority and whether it participates.

Preference stack. The sum of all preference amounts at their stated multiples. The exit value below which the equity shares receive nothing under any structure.

Seniority. The order in which preference classes are paid where proceeds are insufficient. Stacked seniority pays the most recent money first; a pari passu arrangement pays every class the same proportion of its claim.

Participating preferred. A preference share that takes its preference amount and then also shares in the residue on an as-converted basis. Improves the holder’s outcome at every exit value.

Participation cap. A ceiling on aggregate receipts under a participating preference, expressed as a multiple of the amount invested. Above the cap the holder must choose between the capped amount and conversion.

Conversion indifference price. The exit value at which a non-participating class receives the same amount whether it takes its preference or converts. Computed as the preference amount divided by the as-converted ownership percentage.

Full ratchet. An anti-dilution formulation resetting the earlier investor’s conversion price to the price of the new round, without regard to the size of the new issue.

Weighted average anti-dilution. An anti-dilution formulation resetting the conversion price to a weighted average of the old and new prices, with weights reflecting the size of the new issue against the existing capital. Broad-based includes the option pool in that base; narrow-based excludes it.

Pay-to-play. A provision requiring an existing investor to participate in a future round on pain of losing its preference, its anti-dilution protection or its protective provisions.

Rights issue. A statutory offer of new shares to existing shareholders in proportion to their holdings. In India, the mechanism that performs the function a contractual pay-to-play performs elsewhere.

Downside case. The range of exit values at or below the last round’s post-money valuation, in which a term sheet’s economic provisions determine the distribution.

Discussion Questions

Company B’s seniority clause changed nothing for the fund, because the fund held all three preference classes. Identify the circumstances in which the same clause would have mattered enormously, and explain what that implies about who should negotiate it.

The Applied Exercise locates one exit value at which a capped participating structure serves the founders better than a plain one-times non-participating one. Having found it, explain the mechanism to a founder in three sentences, then state whether the result survives the Series C converting.

The market-convention conversion test errs in the founders’ favor in both cells where it errs. Assess whether that explains why the shortcut persists, and identify who in a negotiation has an incentive to correct it.

Full ratchet costs Company B’s founders 17.41 percentage points; weighted average costs them 2.23. Construct the strongest case an investor could make for a full ratchet, then assess it.

Pay-to-play is the one economic term in this chapter whose effect shows up as investor behavior in a bad round, where the others show up as arithmetic at exit. Assess whether a founder should want it in the term sheet, and identify the investor most likely to resist it and the reason it would give.

Further Reading

The standard practitioner account of term sheet economics is Brad Feld and Jason Mendelson’s Venture Deals, whose chapters on the liquidation preference and on anti-dilution are the clearest short treatment available. Its assumptions are United States assumptions throughout, and a reader should carry Chapter 16’s material alongside it when reading the sections on instruments. For the analytical treatment, Andrew Metrick and Ayako Yasuda’s Venture Capital and the Finance of Innovation, recommended at Chapter 14, works the preferred stock problem with the mathematics intact, including the option-pricing view of a preference stack that this chapter’s arithmetic approach deliberately avoids.

The Delaware opinions repay direct reading, and In re Trados is the one to start with. Court opinions are longer than a summary and they contain the facts a summary omits, which in this area are usually the facts that decide the case. The opinion is freely available from the Delaware courts.

For market data, the Cooley quarterly venture financing reports and the equivalent series published by other United States firms are free and current, and they are the only systematic evidence on which terms actually appear at what frequency. No Indian equivalent exists, and building one from offer documents is a research project a student could undertake.

References and notes

1. In the Room is an illustrative composite drawn from multiple real situations; no single company or individual is depicted. The arithmetic corresponds exactly to Company B in the companion model, tab Term Sheet Econ.

2. That Indian venture rounds are predominantly written at a one-times preference rests on practitioner convention. No published Indian series reports the distribution of preference multiples. Evidence grade: asserted.

3. Computed. The model is on the companion site, tab Term Sheet Econ. Company B is a composite illustration and not a real company; its financing history is constructed to be internally consistent, to be arithmetically exact, and to reconcile to the fund’s Portfolio tab, where the ₹28 crore of realized proceeds appears.

4. Companies Act, 2013: classification of share capital under section 43, the private placement mechanics under section 42, further issue under section 62(1)(c), Rule 13 of the Companies (Share Capital and Debentures) Rules, 2014 for the preferential allotment procedure, and Rule 9 of the same rules for the conditions on issuing preference shares. Developed in Chapter 16. Note that section 55 governs redeemable preference shares and prohibits irredeemable ones, and is not the provision under which compulsorily convertible preference shares are classified or issued.

5. Computed. The model is on the companion site, tab Term Sheet Econ, sections 4 through 7. All four structures reconcile to the exit value with zero variance.

6. The circularity in capped participating preferred is a property of the instrument and is derived in the companion model rather than attributed to a source. The workbook resolves participation with two rounds of cap redistribution and reports the conversion decision as a flag, with the reason stated on the tab.

7. Cooley LLP, quarterly venture financing report, reporting the proportion of transactions carrying non-participating preferred. Evidence grade: reported, and verified against the published report. The firm’s fourth-quarter 2025 report, published February 9, 2026, records “96% of deals having nonparticipating preferred stock,” which is the figure the text gives. The series is published quarterly and the proportion moves; take the then-current quarter before print, and keep the quarter named in the text, because a bare percentage from this series dates quickly.

8. Computed. The model is on the companion site, tab Term Sheet Econ, section 9. The weighted average formulation used is the standard one: the adjusted conversion price equals the old conversion price multiplied by the sum of the pre-issue share base and the shares the new consideration would have purchased at the old price, divided by the sum of the pre-issue share base and the shares actually issued. The broad base includes the option pool; the narrow base excludes it.

9. The finding that no Indian company has been publicly documented operating a contractual pay-to-play mechanic is a negative finding of the case-sourcing audit conducted for this book, and is stated as such.

10. Byju’s parent, rights issue of approximately $200 million in early 2024 at a reported discount of roughly 99 percent to a peak valuation of $22 billion, and the shareholder challenge to it. Evidence grade: reported; the oppression challenge and the subsequent orders are contested and under appeal, and should be taught as contested. The company’s separate Term Loan B default, whose position is settled, is treated in Chapter 20. A date stamp belongs on this entry, because the Indian proceedings concerning the parent were still running while this book was written, with orders, appeals, and applications continuing through 2026. Everything stated here is the position as at the date of drafting, no finding has been made on the matters in issue, and the status should be re-checked at proof stage rather than assumed from this page.

11. Computed. The model is on the companion site, tab Term Sheet Econ, comparing the market-convention conversion test against the iterated test across all sixteen structure and exit-value combinations tested.

12. PharmEasy’s parent, rights issue in 2023 at a reported valuation approximately 90 percent below the previous round, reported to have triggered anti-dilution provisions in existing investors’ agreements. Evidence grade: reported. The individual agreements are not public and the specific formulations, share counts, and per-investor effects are not disclosed.

13. In re Trados Incorporated Shareholder Litigation, Delaware Court of Chancery, 2013. Evidence grade: filed, and the figures are now verified against the opinion itself, retrieved from the court’s own site. It records merger consideration of $60 million, comprising $50 million in cash and $10 million in the acquirer’s stock; a management incentive plan payment of $7.8 million, described as 13 percent of the proceeds; an aggregate preferred liquidation preference of $57.9 million including accumulated dividends; approximately $52.2 million reaching the preferred after the incentive plan; and nothing for the common. The court held that the defendants carried their burden on entire fairness, on the reasoning that under the company’s business plan the common stock had no economic value before the merger, so its holders received the substantial equivalent of what they had before. The date of the decision and the case’s parallel citations should still be supplied in the form the house style requires.

14. Square, Inc., Form S-1 and related registration materials, disclosing the Series E ratchet and its operation against an initial public offering priced at $9.00 per share. The guarantee was of a 20 percent minimum return on the approximately $15.46 per share the Series E paid, so the conversion operated as though the offering had priced at approximately $18.56, and the shortfall to the actual $9.00 was made up in shares. Evidence grade: filed as to the registration statement. The 20 percent formulation and the $18.56 threshold come from contemporaneous reporting of the filing and were not read on the filing itself; confirm those two figures and the resulting share issuance against the registration statement before print. The text gives only the price paid and the offering price, both of which hold, and a reader who wants to compute the ratchet needs the threshold, which is why it is set out here.

15. The Across the Table narrative is an illustrative composite drawn from multiple real situations; no single institution is depicted.

16. The finding that no Indian participating-preference waterfall is available in the public record is a negative finding of the case-sourcing audit conducted for this book, and is stated as such.