Turning Chaos into Clockwork

How Ethena makes money from thin air, why that money isn’t guaranteed, and how Boros lets you hedge rate exposure with a fraction of the capital you’d need for price exposure.

The Premise

There’s a stream of money flowing through crypto that most people don’t see. It’s not token emissions. It’s not MEV. It’s not airdrops.

It’s the price that degens pay for leverage.

Every eight hours, billions of dollars change hands between traders holding perpetual futures. Not from trading. Not from liquidations. Just from holding a position. This mechanism is called the funding rate, and it’s the foundation of a multi-billion dollar business you’ve probably heard of: Ethena.

This post will take you from zero to complete understanding of:

1. Why funding rates exist and why they’re structurally positive
2. How Ethena harvests them without taking price risk
3. Why this yield is volatile and dangerous
4. How Boros turns volatile yield into fixed yield
5. Why hedging rate exposure requires far less capital than hedging price exposure

Let’s start at the beginning.

Part 1: The Funding Rate

Why Perpetual Futures Need a Leash

A perpetual future is a contract that tracks an asset’s price but never expires. Unlike traditional futures that settle on a specific date, perps just… continue forever.

This creates a problem. Without expiry, what forces the perpetual price to stay close to spot price?

The answer is the funding rate: a periodic payment between longs and shorts that pushes the perp price toward spot.

If Perp Price > Spot Price:

• Longs pay Shorts
• This discourages longs, encourages shorts
• Perp price gets pushed down toward spot

If Perp Price < Spot Price:

• Shorts pay Longs
• This discourages shorts, encourages longs
• Perp price gets pushed up toward spot

Every 8 hours on Binance (every 1 hour on Hyperliquid), the exchange calculates the divergence between perp and spot, and transfers money accordingly.

The Math

Funding rate is expressed as a percentage of your position size, paid per period.

1
// Simplified funding calculation
2
const fundingRate = ((perpPrice - spotPrice) / spotPrice) * factor;
3

4
// What you pay or receive
5
const fundingPayment = positionSize * fundingRate;
6

7
// If you're LONG and funding is +0.01%:
8
// You PAY 0.01% of your position to shorts
9

10
// If you're SHORT and funding is +0.01%:
11
// You RECEIVE 0.01% of your position from longs

The actual formula is more complex. Most exchanges use something like:

Where the Premium Index measures the divergence between perp and spot (using order book impact prices, not just mid-prices), and the Interest Rate is a fixed baseline component, typically 0.01% per 8 hours on major exchanges like Binance and Bybit. This baseline represents the cost-of-carry difference between holding spot crypto versus USD.

The clamp function prevents extreme swings. But for intuition, the simplified version works: perp above spot means positive funding, longs pay shorts.

A 0.01% funding rate doesn’t sound like much. But it compounds:

Now we’re talking real money.

Why Funding is Structurally Positive

Here’s the insight that makes everything else possible: funding rates are positive more often than negative.

This isn’t random. It’s structural.

Think about who participates in crypto markets:

Who wants to short?

There’s a fundamental imbalance. More demand for leveraged long exposure than short exposure. Funding rate is the price of this imbalance.

Historical data confirms this. Even during bear markets, funding rarely stays deeply negative for long. When it does, arbitrageurs pile in (short spot, long perp, collect negative funding) and push it back up.

The floor is soft. The ceiling is high.

Part 2: Ethena’s Business Model

Before we go further: Ethena’s USDe is explicitly not a stablecoin in the traditional sense. Their docs are clear about this: “Ethena’s USDe is not the same as a fiat stablecoin like USDC or USDT. USDe is a synthetic dollar, backed with crypto assets and corresponding short futures positions.”

This distinction matters. USDe has different risks than fiat-backed stablecoins. It’s backed by crypto and derivatives, not dollars in a bank.

Delta Neutral: The Magic Words

Ethena’s strategy sounds almost too simple:

1. Hold spot crypto (ETH, BTC, liquid staking tokens like stETH, and stables like USDC)
2. Short equivalent perpetual and deliverable futures
3. Collect funding payments + staking yield

The reality is more nuanced than “buy ETH, short ETH perp.” Ethena holds a diversified backing portfolio:

• ~6% in staked ETH (earning ~3-4% staking yield)
• ~87% in BTC and ETH with corresponding short derivatives
• ~7% in liquid stables like USDC/USDT (earning treasury-rate yields, no hedge needed)

But the core logic is the same: hedge the delta, harvest the funding.

What if ETH crashes?

The Hedge

Let’s trace through what happens to Ethena’s portfolio when ETH moves:

ETH goes up 20%:

ETH goes down 30%:

The positions cancel perfectly. Ethena has zero exposure to ETH price. This is what “delta neutral” means: delta is the sensitivity to price, and theirs is zero.

What they ARE exposed to is the funding rate. As a short, they receive funding when it’s positive.

The Income Statement

Let’s model a year of Ethena’s operation:

Not bad. But look at Q3. They lost $55M in three months because funding went negative.

This is the problem with Ethena’s business: the yield is real, but it’s volatile.

(A note on this stress test: historically, only one quarter in the last three years had negative average funding, and that was during the ETH PoW arbitrage period around the Merge, a one-off event. The longest streak of consecutive negative funding days was 13 days. So Q3 here is a worst-case scenario, not the norm. But worst-case scenarios are what you plan for.)

sUSDe holders expect relatively consistent yield. In 2024, sUSDe averaged around 19% APY, far better than this model suggests. But that’s the point: yield is variable. It could be 19%. It could be 5%. It could briefly go negative. Ethena maintains a Reserve Fund to cover negative periods so users never see negative yield on sUSDe, but the underlying income stream remains volatile.

How do you build predictable products on unpredictable income?

Visualizing the Volatility

Unless you can convert volatile into fixed.

Part 3: Interest Rate Swaps

Before we get to Boros, we need to understand the oldest trick in finance: the swap.

Fixed vs Floating

There are two kinds of interest payments in the world:

Fixed: “I will pay you exactly 10% per year, no matter what.”

Floating: “I will pay you whatever the market rate is. Today it’s 8%, tomorrow it might be 15%, who knows.”

Fixed is predictable. Floating captures upside but carries risk.

The Swap Mechanic

A swap is an agreement between two parties to exchange cash flows:

Party A: “I’ll pay you 10% fixed”

Party B: “I’ll pay you whatever the floating rate is”

Every period, they compare rates and settle the difference.

No principal changes hands. Just the rate differential, applied to an agreed notional amount.

Example:

On $1M notional, for one month:

If floating had been 5% instead, A would pay B $4,167.

The swap lets you transform your exposure:

• If you’re naturally receiving floating and want certainty → receive fixed, pay floating
• If you want upside on rates → pay fixed, receive floating

Who’s On Each Side?

Every swap needs two parties with opposite needs:

Natural fixed-rate seekers:

• Companies with floating-rate debt who want predictability
• Protocols earning variable yield who need stable income
• Anyone who values certainty over potential upside

Natural floating-rate seekers:

• Speculators who think rates will rise
• Traders who want leveraged exposure to rate movements
• Market makers willing to take the other side for a spread

The swap market exists because these groups find each other.

Part 4: Boros

Now we can understand what Pendle built.

The Product

Boros is an interest rate swap platform for crypto funding rates. It lets you:

1. Lock in fixed funding income if you’re earning floating (like Ethena)
2. Speculate on funding rates if you think they’ll move
3. Provide liquidity by taking the other side of trades

The core primitive is the Yield Unit (YU):

Long vs Short YU

Long YU:

• You pay: Fixed rate (Implied APR, set when you open)
• You receive: Floating rate (actual Binance funding)
• You profit when: Actual funding > Fixed rate you’re paying
• You lose when: Actual funding < Fixed rate you’re paying

Short YU:

• You receive: Fixed rate (Implied APR)
• You pay: Floating rate (actual Binance funding)
• You profit when: Actual funding < Fixed rate you’re receiving
• You lose when: Actual funding > Fixed rate you’re receiving

Ethena’s Hedge

Let’s trace through how Ethena could use Boros. Pendle’s docs explicitly use Ethena-style delta-neutral farming as the canonical use case; it’s the obvious fit. Whether Ethena has actually integrated isn’t public, but the mechanics work the same regardless.

Before Boros:

Ethena’s only position:

• Short ETH perp on Binance
• Receives: Floating funding rate (whatever it happens to be)

After adding Boros hedge:

Position 1 (Binance perp short):

• Receives: Floating funding

Position 2 (Boros YU short at 10% Implied APR):

• Receives: 10% fixed
• Pays: Floating funding

Now let’s see what happens under different funding scenarios:

In every scenario, Ethena nets exactly 10%.

The floating components cancel out. What remains is pure fixed income.

The Transformation

This is the magic of swaps. Variable becomes fixed. Chaos becomes clockwork.

Where Does the Variance Go?

It doesn’t disappear. It transfers.

For every short YU (Ethena), there’s a long YU (speculator). The speculator is betting funding will exceed the fixed rate. If they’re right, they profit from Ethena. If they’re wrong, they pay Ethena.

If Floating > 10%: Speculator profits. If Floating < 10%: Ethena profits (beyond their fixed 10%).

But Ethena always gets 10%. The variance is the speculator’s problem now.

Who Would Take the Other Side?

Why would anyone long YU against Ethena?

1. Speculators who think funding stays high: If you believe bull market continues and funding averages 20%, you’d happily pay 10% fixed to receive 20% floating.

2. Leveraged yield seekers: YU gives you exposure to funding without holding spot or perp. Pure rate exposure, amplified.

3. LPs in Boros Vaults: Passive liquidity providers who earn fees for taking the other side of trades.

The 10% fixed rate is the market’s consensus expectation of average funding, minus a risk premium. Ethena pays this premium for certainty.

Part 5: Capital Efficiency

This is the part that confuses everyone.

Pendle advertises “up to 1000x capital efficiency” for Boros. When I first saw that number, I assumed it was marketing fluff. A thousand times? That’s the kind of claim that usually means someone’s either lying or about to blow up.

But the more I dug into it, the more the math actually checked out. Not because Boros invented some new form of leverage, but because of what you’re actually hedging. When you open a position on Boros, you’re not exposed to ETH’s price. You’re exposed to the funding rate. And funding rates, even when they swing wildly in percentage terms, translate to tiny dollar amounts per period.

Let me build it from first principles.

The Question

Ethena has $1B of funding rate exposure from their perp shorts.

To hedge it on Boros, they open a $1B notional short YU position.

How much margin (collateral) do they need to post?

What Margin Is For

Margin exists to cover potential losses before liquidation.

If you deposit $10M margin and your position loses $10M, you’re at zero. The system liquidates you before you go negative.

So the question becomes: What’s the maximum Ethena can lose on their Boros position in a given time period?

Calculating Potential Loss

Ethena is short YU at 10% fixed.

They lose when actual funding exceeds 10%. Let’s stress test with extreme funding: 100% APR.

On $1B notional, short YU at 10% fixed, with funding spiking to 100% APR (extreme):

• Loss rate: 90% APR against them
• Per 8-hour period (1/1095 of a year): ~$822k loss

Under extreme conditions (100% APR funding, which is rare; historical funding has ranged from about -20% to +60% APR in normal markets), Ethena loses ~$822k per 8-hour settlement.

Per day: ~$2.47M

Compare to Traditional Perp Margin

If you held a $1B ETH perp position directly:

ETH can move 10% in a day easily. On $1B notional, that’s a potential $100M loss per day.

A $1B perp position can lose $100M in a day from price movement.

A $1B YU position can lose ~$2.5M in a day from rate movement.

The YU position’s risk is 40x smaller.

Why the Difference?

You’re not exposed to the same thing.

Funding rate differentials move much slower than prices. The “notional” is the same, but the risk per dollar of notional is vastly smaller.

Setting Margin Requirements

Boros’s margin requirement is a design choice based on:

1. How much can you lose per period?
2. How long does the liquidation engine need to act?
3. What’s the safety buffer?

Conservative margin calculation:

In practice, Chaos Labs worked with Pendle to develop dynamic margin formulas that account for rate volatility and time to maturity. The key insight: margin scales with expected payment risk, not notional exposure.

The Visual

Same “notional.” 13x less capital required.

The Circular Logic Made Explicit

The capital efficiency chain:

1. Funding rate differentials are small → Max ~100% APR swing, usually much less
2. Therefore, losses per period are small → $1B notional loses max ~$800k per 8 hours
3. Therefore, margin requirements can be small → Protocol might require only 0.75-1% margin
4. Therefore, capital efficiency is high → $7.5M margin controls $1B notional exposure
5. Therefore, you can hedge $1B with $5-20M margin → 50-200x capital efficiency in practice

Putting “1000x” in Context

At this point you might be thinking: okay, 50-200x capital efficiency makes sense for conservative parameters. But Pendle is claiming “up to 1000x.” Is that real, or is it marketing?

It’s both, honestly.

The 1000x number is achievable under ideal conditions. But more importantly, it’s not even that unusual in the broader world of finance. Interest rate swaps have been around since the 1980s, and they’ve always been wildly capital-efficient compared to trading the underlying assets.

Think about what a swap actually is. Two parties agree on a notional amount (say, $1 billion) but neither party actually sends $1 billion anywhere. They just exchange the difference between two rates, applied to that notional. If the fixed rate is 10% and the floating rate is 12%, the only money that changes hands is 2% of $1B, prorated for the period. The notional is a reference number, not capital at risk.

This is why banks can have hundreds of billions in notional swap exposure without the financial system collapsing. JPMorgan reportedly has something like $50 trillion in notional derivatives exposure. That sounds insane until you realize most of it is interest rate swaps where the actual cash flows are tiny fractions of the notional.

Boros isn’t inventing new physics. It’s applying the same structure that’s been battle-tested in traditional finance for forty years.

So how does Pendle get to 1000x specifically? Optimistic assumptions:

• Short maturity (less time for rates to move against you)
• Normal rate volatility (not stressed conditions)
• Efficient liquidation (quick response to margin calls)
• Maximum allowed leverage

Aggressive margin calculation: $365k max period loss × 1 period × 1.2 buffer = ~$440k margin. That’s ~2,270x capital efficiency on $1B notional.

The “up to 1000x” is reasonable under favorable conditions. In practice, with conservative parameters, you’re probably looking at 50-200x. That’s still dramatically better than the ~10x you typically get on price-exposed perp positions, where a 10% move can wipe you out. The difference is what you’re exposed to: price volatility versus rate volatility. One moves fast. The other doesn’t.

Part 6: The Full Picture

Let’s zoom out and see how all the pieces connect.

The Value Chain

Where Does the Money Come From?

Trace it back to the source:

1. Retail degens pay funding because they want leveraged long exposure
2. Ethena collects that funding by providing short liquidity
3. Ethena pays a premium to speculators for taking their variance risk
4. Speculators either profit (if funding stays high) or lose (if it drops)

The ultimate source of yield is the structural demand for leveraged long exposure in crypto. As long as people want to go long with leverage, this money flow continues.

Is This Sustainable?

The funding rate premium isn’t a bug or a temporary inefficiency. It’s a structural feature of how crypto markets work.

As long as:

• More people want leveraged long exposure than short exposure
• Perpetual futures remain the dominant leverage instrument
• Market participants value convenience over capital efficiency

…funding rates will remain structurally positive.

This doesn’t mean they’re always positive, or that the yield is risk-free. Bear markets happen. Funding flips negative. That’s exactly why hedging exists.

But the underlying economic reality (that leverage has a cost, and shorts get paid to provide it) is unlikely to change.

Conclusion

We’ve covered a lot of ground:

1. Funding rates are the price of leverage, structurally positive due to market composition
2. Ethena harvests funding via delta-neutral basis trading, but faces volatile income
3. Interest rate swaps let you exchange fixed for floating (or vice versa)
4. Boros brings swaps to crypto funding rates via Yield Units
5. Capital efficiency comes from the fact that rate differentials move slower than prices

The key insight: you’re not hedging principal, you’re hedging a derivative of a derivative. The funding rate is already a small percentage of notional. The differential from your fixed rate is a fraction of that. Your margin only covers the fraction of the fraction.

That’s why rate swaps are capital-efficient everywhere, not just crypto. Interest rate swaps are the largest derivatives market in the world ($500+ trillion in notional) precisely because you can take big positions with modest capital. Boros just brings this structure to crypto funding rates.

When I first encountered this, it felt like genius. Now it feels like a logical combination of well-understood concepts: basis trading, delta hedging, interest rate swaps, margining.

The building blocks were all there. Someone just had to assemble them.

Further Reading

• Ethena Documentation: Particularly the risks section, which is unusually thorough
• Pendle Boros Documentation
• Chaos Labs: Risk Infrastructure for Boros
• Hull, Options, Futures, and Other Derivatives: Chapter 7 on Swaps
• Harris, Trading and Exchanges: Market microstructure fundamentals