Chapter 26 · How Protocols Make Money

A Real-World Problem

You are comparing two protocols to decide which is the better business.

The first ranks near the top of its sector by total value locked. Daily trading volume is large, tens of thousands of addresses interact with it every day, and its community is lively. Every public metric looks impressive.

The second has one tenth of the first protocol's TVL, four-digit daily activity, and hardly anyone discussing it.

Most people would choose the first, with good reasons: it is larger, more active, and more widely recognized.

Now add one more line of information. Look at both protocols' income and spending for the period:

Protocol A   Users pay substantial fees, mostly distributed to liquidity providers by rule;
             the protocol's share, minus token incentives issued during the period, is negative.
Protocol B   Fees are much smaller, but incentives ended long ago;
             the protocol's share stays in its treasury and is positive.

A is paying for scale. B is earning money. Both can reasonably describe themselves as growing.

One thing needs to be clear immediately: this chapter is not saying B is a better investment than A. This stage is about evaluating and building a project, not trading it. Burning money to gain scale can be entirely reasonable at certain stages—provided you know you are doing it, when it should end, and what will remain afterward.

The real problem is that you will not find that extra line on the front page of any public dashboard. The front page shows TVL, volume, and user counts, and none of those three is revenue.

Chapter 23 taught you how to collect data and define its scope. This chapter goes one layer deeper: once you have the numbers, how do you arrange them into a statement that shows whether the protocol is making money?

Thought Experiment

Replace the protocol with a marketplace.

You open a market on an empty lot. The rules are simple:

• Vendors put out their goods, customers shop, and leave after buying.
• You take 1% of each sale. This is your only income.
• To attract activity, you print market vouchers. Each time a vendor makes a sale, you give them a voucher they can spend like money in the market.

After three months, four numbers are in front of you:

Total value of goods on the stalls    80 million    ← The vendors' goods, not yours
Sales over three months              60 million    ← Turnover, not your money
Your 1% commission                  0.6 million    ← Money you received
Vouchers issued, at face value      0.9 million    ← Something you printed

Now answer: how much money did this market make in three months?

The first number is easiest to present as an achievement: 80 million worth of goods sounds like a large market. But none of it belongs to you. The vendors could remove everything tomorrow, and its value changes with market prices—if the goods rise in price, your market gets “bigger” without a single additional transaction.

The second number is turnover. Sales of 60 million sound much more impressive than 0.6 million, but you receive only 1%. Worse, turnover can be manufactured: two people could repeatedly trade the same goods with each other, generating unlimited volume while you subsidize your 1% commission with vouchers.

The third number is money you actually received: 0.6 million.

The fourth is what you gave out: vouchers worth 0.9 million. They differ from cash—you can print them at almost no cost, with no cash leaving during the period. But eventually someone will spend them, exchanging them for real goods held by someone else. Who pays for those vouchers? Everyone holding market vouchers, through dilution.

So:

Excluding vouchers:   +0.6 million; a profitable business
Including vouchers:   0.6 − 0.9 = −0.3 million; a net outflow over these three months

Both answers are internally consistent, and they lead to opposite decisions.

Ask one more question, the one many people miss: if you stop issuing vouchers, how much of that 60 million in sales remains?

The accounts cannot answer this. You have to try stopping. Once you realize you are afraid to stop, you already know something about the answer.

Your Decision

You want to know how a protocol makes money and whether it actually does. Which number do you start with?

What Happens

The four choices converge on three inequalities. They form this chapter's backbone and are worth writing down:

TVL ≠ Revenue, Volume ≠ Profit, User ≠ PMF.

Take them apart:

All three share a structure: the left side is easy to obtain, grow, and display; the right side requires your own calculations, grows slowly, and is rarely displayed willingly. This asymmetry explains why the industry's materials are full of numbers from the left.

Pair each metric on the left with one on the right, and you have a portable comparison table:

The last row needs its own explanation and introduces the second half of this chapter. If a protocol reports a $100 million treasury and $90 million is its own token, that $100 million is not really $100 million. Selling it when funds are needed pushes the price down, leaving far less cash than the book value.

Build a Model

A Five-Line Protocol Income Statement

This is the structure to take away. You can apply it to any protocol:

As a statement:

  Total fees paid by users           Fees
− Share paid to the supply side      LP / depositors / validators / market makers
= Share retained by the protocol     Protocol Revenue
− Token incentives issued           Incentive Cost
= Net revenue                        Net Revenue

Five lines, four numbers, three minus signs. Each line has a recurring trap:

The Other Side of Chapter 17

Chapter 17 examined yield sources: displayed APY equals fee APY plus interest APY plus subsidy APY. “Who pays this money?” was the first of its four questions.

That chapter took the recipient's side; this chapter takes the payer's side. The same money, viewed in two directions.

This connects the chapters and yields a useful inference: real yield on the user side equals the portion of protocol fees paid to the supply side. If the two calculations disagree, at least one has an incorrect definition—precisely the cross-check described in Chapter 23.

Effective Take Rate

Compress the protocol's business model into one number:

Effective take rate = Protocol share ÷ Total fees paid by users

The ratio varies enormously between sectors, and higher is not always better—it is inseparable from the intensity of competition.

The fee switch deserves close attention. Many protocols include a contract parameter determining how much collected fees belong to the protocol. It is often zero at launch and later enabled through governance. Research three things: whether the switch exists, its current setting, and who can change it. The first two can be read from the contract; the third is governance risk from Chapter 24.

Three Treasury Questions

Negative net revenue does not mean the protocol will die, provided it has funds to burn. The treasury is therefore the final part of the model:

The second question has a specific calculation worth writing down:

Cash runway in months = Spendable treasury assets excluding own tokens ÷ Average monthly net outflow

The numerator must exclude the protocol's own token because selling it depresses its price. Runway calculated from treasury book value is often several times the real figure.

Four Quadrants

Cross positive or negative net revenue with high or low incentives to get four kinds of protocol:

The bottom-right quadrant is the industry's most common and the one most in need of an honest description. Being there is not inherently wrong—using incentives to bootstrap activity can be reasonable. Chapters 28 and 30 explain when it works.

The problem is that it is often presented as the top-left quadrant. Distinguishing them only requires calculating the fourth line of the income statement.

What It Is Called

Hands-on

Revenue share source: Contract parameter … (record its name and observed value), not documentation claims
Incentive accounting: Record quantities issued and a separate series valued at daily average prices
Time window:          The most recent 90 complete calendar days, UTC

Total fees paid by users          ____________  Source:
− Supply-side share               ____________  Source:
= Protocol share                  ____________
− Incentive spending, in money    ____________  Source:
= Net revenue                     ____________

Effective take rate = Protocol share ÷ Total fees = ______
Incentive coverage = Protocol share ÷ Incentive spending = ______

Runway in months = Spendable assets excluding own tokens ÷ Average monthly net outflow

AI Lab

Document revenue for <protocol name> over the most recent 90 days.

Part one: do not give numbers yet. First produce a definitions document:
1. Total user fees: covered contracts, chains, and versions
2. Supply-side share: source contract parameter, its name, and current value
3. Protocol share: destination address and who controls it
4. Token incentives: distribution contract, counting method, and valuation prices
5. Time-window boundaries

Part two: report these five lines, citing a source for each:
  Total fees paid by users
− Supply-side share
= Protocol share
− Incentive spending during the period
= Net revenue

Part three: calculate two ratios: effective take rate, and protocol share divided by incentives.

Part four: separately list:
- What this analysis does not cover
- Which figures have explicit sources, and those sources
- Which figures you could not obtain

Rules:
- Cite every contract address and parameter name.
- Write “could not obtain” for unavailable figures; do not invent plausible values.
- Timestamp all collected figures.

Real-World Case

Change One Variable

Takeaway Questions

Chapter Check