The Pair That Looked Safe Until It Wasn't

You picked the boring trade. USDC/USDT, maybe DAI/USDC, two stablecoins both pegged to a dollar. You deposited, watched the APR ticker, and went to bed feeling responsible.

Then you checked the position six months later.

Not catastrophically bad, not the 50%-drawdown bad that haunts volatile pairs, but a quiet, persistent drain that compounded month by month until the fees never quite caught up. This is the specific problem: impermanent loss in stable pairs is small per unit of time, but fee revenue is also small, and which one wins is far less settled than most liquidity mining guides will admit.

Three things cause impermanent loss to exceed fee revenue in stable pairs: a depeg event of even moderate size, a pool with chronically low volume relative to its total value locked, and the particular shape of the AMM curve in use. Understanding all three explains why two liquidity providers can deposit into the same protocol on the same day and walk away with meaningfully different outcomes six months later.

The Math Underneath the "Stable" Label

Impermanent loss is a function of price divergence between the two assets in a pool. For a standard constant-product AMM using the x\*y=k formula, the loss is fully deterministic: if one asset moves to 1.25x its original price relative to the other, an LP loses roughly 0.6% compared to just holding. Sounds trivial.

Here's the wrinkle. Stable pairs don't use x\y=k. Protocols like Curve Finance use a hybrid invariant that concentrates liquidity near the 1:1 price. The curve is nearly flat around the peg, which means the pool barely rebalances when prices are close, which means lower impermanent loss under normal conditions. Fine. The catch: that same concentrated shape means the pool rebalances aggressively* the moment price moves even slightly away from 1:1. The tighter the concentration, the sharper the response.

A 2% depeg in a Curve-style stable pool causes more rebalancing activity, and therefore more impermanent loss, than a 2% move in a standard pool. The AMM is doing exactly what it was designed to do, just not in your favor when prices drift.

Run the numbers on a concrete scenario. You deposit $50,000 in DAI and $50,000 in USDC into a Curve-style pool. A stablecoin stress event pushes DAI to $0.97 relative to USDC. On a concentrated stable curve, the pool has already sold a significant portion of your USDC for DAI to maintain balance, so you now hold more of the cheaper asset. The price recovers to $1.00. Round-trip complete, and the pool's rebalancing bought high and sold low on your behalf the entire time. Impermanent loss on that round trip for a highly concentrated pool can reach 0.3% to 0.8% of position value. Annualized fee revenue on a low-volume pool might be 1% to 3%. One moderate depeg and recovery consumes weeks of fee accumulation in a single event, like a slow leak that drains faster than the drip you're collecting in the bucket.

When Volume Dries Up and Fees Stop Covering the Gap

Fee revenue is trading volume times the fee rate times your share of the pool. That's it. No other income source exists for a basic AMM LP.

This creates an asymmetry most people don't think about carefully enough. Impermanent loss accrues based on price movements, which happen whether or not anyone is trading. Volume, on the other hand, can vanish. During low-activity periods, a stable pair pool might see its 24-hour volume drop to 2% or 3% of TVL. At a 0.04% fee tier, common on Curve, that generates roughly 0.001% in daily fees, maybe 0.03% monthly, 0.36% annualized. That is the baseline fee income.

One bad weekend can erase it.

A single 3% depeg event on a concentrated curve can produce impermanent loss of 0.5% or more. Consider Marcus and Priya, two LPs who each deposit $100,000 into the same DAI/USDC pool. Marcus enters during a period of high stablecoin demand, lots of cross-protocol arbitrage, volume running at 15% of TVL daily. He earns 4.2% annualized and exits before any depeg. Priya enters three months later, after the narrative has moved on, volume has dropped, and a minor stablecoin wobble clips her position. Her realized return is negative 0.8%. Same pool, same assets, different timing. The fee-to-loss ratio is not a fixed property of the pair. It shifts with market conditions, and the dashboard won't warn you when it turns.

What People Get Wrong About "Low Risk" Stable Pools

The most persistent misconception is that impermanent loss is negligible in stable pairs under all conditions. It is negligible when prices stay within a very tight band and volume stays high. Neither condition is guaranteed, and treating one as permanent while ignoring the other is how positions quietly bleed out.

The second mistake is ignoring pool concentration parameters. Not all stable pools are equally tight. Some protocols allow pool creators to set amplification coefficients that control how concentrated the liquidity is. A higher amplification coefficient means more capital efficiency near the peg and sharper impermanent loss when the peg breaks. Checking this parameter before depositing is the step most guides skip entirely, and skipping it is, frankly, negligent.

The third mistake: treating fee APR estimates from protocol dashboards as forward-looking guarantees. Those figures are typically calculated from the previous 24 or 168 hours of volume, then annualized. They are a rearview mirror, not a windshield. A pool that showed 8% APR last week because of a specific arbitrage event might show 0.4% APR next week. The impermanent loss you incur, meanwhile, is locked in the moment it happens.

Still, this is not an argument that stable pairs are traps. For large, high-volume pools with sustained organic demand, fee income can genuinely and consistently outpace impermanent loss. The USDC/USDT pool on a major protocol with billions in TVL and hundreds of millions in daily volume is a different animal from a small pool with $2 million TVL and intermittent volume. The underlying mechanics are identical. The realized outcome is not.

The Ratio That Actually Matters

The number to watch is fee APR divided by the maximum realistic impermanent loss from a plausible depeg. Comfortably above 3:1, and fees have a genuine buffer. Below 2:1, a single moderate stress event flips the math.

Below 1:1? You are paying for the privilege of providing liquidity.

Ask yourself: would you sell insurance on a house in a flood zone for less than the expected annual payout? Because that is the position you are in when this ratio inverts. Concentrated liquidity was designed to help LPs earn more per dollar deployed. It does that. It also concentrates the downside. The stable pair that looked like a savings account is closer to a short-volatility position: you collect small premiums in calm conditions and absorb outsized losses when conditions stop being calm. Understanding that framing does not make the trade bad. It makes the trade honest.