EODHD Academy
‹ Understanding DeFi Lesson 14 of 16
Contents Lesson 14 of 16

5 min read · professional

Why do the highest advertised APYs decay fastest?

Emission-funded yields do not merely fall over time. They fall because they were attractive — the mechanism that publicises the rate is the same mechanism that destroys it. Three variables, one equation, and two of them move against you automatically.

The equation

For a pool paid in a protocol's own token:

Reward APR = (tokens emitted per year × token price) ÷ total value deposited

Note the R. This is a simple annual emission rate; calling it APY would imply a compounding convention that nobody here has stated, and the number is quoted as APY on roughly every farm you will ever see.

Three inputs: emission rate E, token price P, deposits (TVL). Now watch what a high number does to each.

Worked: the decay, one variable at a time

A protocol emits 1,000,000 tokens a year to a pool. The token trades at $2.00. The pool holds $10 million.

APR = (1,000,000 × 2.00) ÷ 10,000,000 = 2,000,000 ÷ 10,000,000 = 20%

Step 1 — the yield attracts capital. Twenty percent is visible on every aggregator, and capital in DeFi is permissionless, so it arrives within days. TVL rises to $50 million:

APR = 2,000,000 ÷ 50,000,000 = 4%

Nothing about the protocol changed. The rate fell 80% because other people read the same dashboard. Your yield is diluted by everyone who agrees with you.

Step 2 — the recipients sell. Emission recipients typically have a zero cost basis, and the tokens have no claim on cash flows. Persistent supply meets whatever demand exists. Say the price halves to $1.00:

APR = (1,000,000 × 1.00) ÷ 50,000,000 = 2%

Step 3 — the schedule tapers. Most emission programmes decline by design, often halving on a schedule. Cut E to 500,000:

APR = (500,000 × 1.00) ÷ 50,000,000 = 1%

From 20% to 1% with no exploit, no depeg and no failure. Every step was the system working as intended.

The reflexive part

The three variables are not independent, which is what makes the decay so reliable:

  • A high APY raises TVL, cutting the APY.
  • Emissions create sell pressure proportional to the emission, cutting P, cutting the APY.
  • A falling P cuts the APY, which drives capital out, which raises the APY again — producing the characteristic sawtooth of mercenary capital rotating between programmes.

Sell pressure is worth its own number: at 1,000,000 tokens a year and $1.00, that is roughly $2,740 of new supply every day hitting the market, whether or not anyone wants it. A programme's emissions are a standing sell order with a schedule.

What an APY figure actually is

It is an annualised extrapolation of an instant. It takes whatever the pool earned in a recent short window and projects it over a year as if nothing changes — while the equation above guarantees that things change. Add compounding assumptions (APY versus APR) and the headline number can be inflated further without a single dollar of extra income.

Three habits follow:

  1. Read the base yield, not the total. The base — fees or borrower interest — is the part that survives if the emission programme ends tomorrow.
  2. Ask about the emission schedule. Fixed supply or perpetual? Halving when? Governance can change it, and governance is people.
  3. Ask what the position looks like at zero emissions. For a liquidity pool, subtract impermanent loss and you often find the honest answer is negative.

The honest framing

Emissions are not a scam by definition. They are a legitimate bootstrapping tool: pay users to provide something the protocol needs before organic demand exists, in the hope that the demand arrives before the budget runs out. The Anchor case in Unit 3 is the version where it did not.

What is not legitimate is presenting a dilution-funded transfer as though it were income. A yield paid in newly printed tokens is not a return until you have sold the tokens, and everyone receiving them is trying to do the same thing.

This lesson does not suggest chasing, avoiding, or timing any yield programme. It gives you the equation.

Try it now

  1. Take the equation and solve backwards: if a listing advertises 40% with $20 million of TVL, what dollar value of tokens must be emitted annually? Does that number look plausible against the protocol's market capitalisation?
  2. On a public analytics site, find a pool's APY history over six months. Plot it against the pool's TVL over the same period and observe the relationship.
  3. For any incentive token, compute the daily emission in dollars and compare it with the token's daily trading volume. If emissions are a large share of volume, you have measured the sell pressure directly.