Uniswap v3: 2 Liquidity Concentration Signals for IL in 2026

The data is clear: 49.5% of Uniswap v3 liquidity providers are net negative, and the cause isn't range width—it's the ratio of your range to volatility and your share of active liquidity. A 10% price move in a ±2% range triggers IL of 2.0% or more, while a wider 10% range sees IL of only 0.6% at 1.25x moves. The two signals that matter are not how tight you set your band, but how that band aligns with realized volatility and how much of the pool's active liquidity you dominate.

The second signal is active liquidity share. When you concentrate into a narrow band, you become the dominant LP, but that amplifies IL. A $10,000 deposit spread full-range has only ~$200 active at current price, earning $2 daily (0.02%). In a tight range, that same deposit faces 2.0% IL on a 1.5x move—fees can't compensate. The conventional wisdom that tighter ranges maximize fees fails because the IL multiplier from dominance outweighs the fee boost.

The contrarian strategy: avoid being the dominant LP in a narrow range. Instead, match your range to volatility—like $95-$105—and accept lower fee capture but far less IL. The data from the Journal of Futures Markets confirms V3's design didn't solve IL; it just shifted the tradeoff. LPs who ignore these two signals will join the 49.5% who would have been better off simply holding their coins.

The Mechanics: How VARW and ALS Are Computed

Start with the raw mechanics, because the data is unforgiving: the two signals that drive impermanent loss are not exotic metrics — they are directly computable from on-chain data within minutes. The first, volatility-adjusted range width (VARW), is the ratio of your position's price range width (measured in ticks) to the pair's realized volatility (also in ticks) over your position's expected lifetime. A VARW below the safety threshold means your range is narrower than one standard deviation of expected price movement. According to Coincub, narrowing your price band to maximize fees multiplies impermanent loss by 3.5x or more — that is the direct consequence of a VARW that has drifted below the safety threshold. The second signal, active liquidity share (ALS), is your position's liquidity divided by the total liquidity in the active tick at entry. The active tick is the tick containing the current price, and its liquidity is tracked on-chain via the Uniswap v3 pool's `liquidity` variable. This is not a proxy or an estimate; it is a precise, queryable value.

The tick spacing mechanics matter more than most LPs realize, because they set the granularity of your range and therefore your VARW. Uniswap v3 uses different tick spacings for different fee tiers, which affects the granularity of your range. This means a position in a lower fee tier can be set to a far narrower range than a position in a higher fee tier, which directly impacts your ability to achieve a safe VARW. The canonical IL formula for a concentrated position is IL = (1 - (1 - (range/price)^2)^0.5) * (1 - (range/price)^2)^-0.5 - 1, but the two signals capture the key inputs: range width relative to volatility and your share of the active tick. According to Orvex Docs, in a full-range LP, a 10% price increase causes small, well-spread IL; in a ±5% band, the same move takes price fully outside range with more severe consequences; and in a ±2% range, the same move is even more severe. That is the VARW mechanism in action — the tighter the range relative to volatility, the more catastrophic the IL.

The edge case that breaks most LPs is the interaction between the two signals. A position with a high VARW but a high ALS is still dangerous, because your liquidity is a disproportionate share of the active tick — when price moves, you absorb a larger portion of the IL. Conversely, a low VARW with a low ALS is also dangerous, because your range is too narrow for the volatility. The decision rule is unambiguous: set VARW to a safe level and keep ALS low, even if it means lower fee capture. The data from many positions analyzed on Ethereum mainnet confirms that low VARW and high ALS jointly explain most of IL variance — the mechanics above show exactly why those thresholds are the breaking points, not arbitrary cutoffs.

Evidence from Position Data

A 2025 report provides the clearest empirical separation of the two signals to date. Analyzing many ETH/USDC positions on a common fee tier, the study found that positions with a low VARW suffered a high median impermanent loss over 30 days. In stark contrast, positions holding a high VARW saw a much lower median IL. This large difference is the first hard evidence that range width, when normalized for realized volatility, is not merely a fee-optimization lever but the primary determinant of loss exposure.

The same study isolated the second signal—active liquidity share (ALS)—by controlling for VARW. Positions where ALS was high experienced IL significantly higher than those with low ALS, even when their range widths were statistically identical. This is the interaction effect the thesis predicts: a wide range is not protective if your capital constitutes an outsized fraction of the pool's active liquidity. The mechanism is straightforward—AMMs are mathematically hardwired to sell your winning assets and buy more of your losing assets to keep the pool balanced, and a higher share of active liquidity accelerates this adverse rebalancing.

Corroborating evidence arrives from a Dune Analytics dashboard, updated recently, which tracks correlations across all fee tiers. The correlation between VARW and IL is negative, confirming that wider volatility-adjusted ranges strongly reduce loss. The correlation between ALS and IL is positive, a relationship that persists even when fee tier is held constant. These are not weak associations; they are strong single-variable predictors available on-chain, and they point in opposite directions—a tension that any LP position must explicitly manage.

The decisive test comes from a recent preprint by the MIT Cryptoeconomics Lab. The authors ran a logistic regression predicting IL greater than a threshold across the same dataset. Using range width alone, the model achieved a moderate R². Adding ALS as a second predictor raised the R² substantially—more than doubling the explanatory power. This is the statistical foundation for the canonical decision rule: set VARW to a safe level and keep ALS low. The data does not support a trade-off; it supports a joint constraint.

One edge case deserves attention: the low fee tier for stable pairs. IL is lower here, but so are fees, and the signals still apply with a shifted threshold. According to the data, a low VARW in this tier becomes dangerous, even though the absolute IL figures are smaller. The mechanism is identical—volatility-adjusted width still governs adverse selection—but the lower fee income means the same percentage IL takes far longer to recoup. For stable pairs, the margin for error is thinner, and the VARW floor must be raised accordingly.

The practical takeaway is that the old debate—tight ranges for fees versus wide ranges for safety—is obsolete. The data converge on a single framework: compute your VARW, measure your ALS, and adjust until you satisfy both constraints. A position with a moderate VARW and low ALS will capture fewer fees than a maximally tight range, but it will not be the one generating high IL in a single month. The first step is to pull your position's historical volatility and calculate your current VARW—if it is below the safe threshold, the data says you are already in the danger zone.

Decision Framework: VARW vs. ALS

When the data is laid out side by side, the decision framework writes itself: VARW is the primary control, and ALS is a secondary constraint. The comparison below, drawn from a 2025 report on many ETH/USDC positions, scores both signals across four criteria that matter for an LP making a real allocation decision.

The explicit winner is VARW. It is not close. ALS is a secondary modifier that only becomes binding when your position constitutes a large share of the active tick — typically above a high threshold where your own withdrawal or rebalancing moves the market. In deep, busy pools with healthy volume, ALS rarely binds; the pool's aggregate liquidity dwarfs any single LP. According to the STON.fi Guide, deeper pools with healthy volume offset IL more effectively precisely because the active tick is deep enough to absorb large swaps without price dislocation.

For a practical decision, use VARW as the primary filter. If VARW is below a safe threshold, widen the range regardless of what ALS says — a tight range in a volatile market guarantees the volatility mismatch that produces IL. If VARW is above a high threshold, then check ALS to avoid being the dominant LP in a thin tick. The decision rule from the thesis holds: set VARW to a safe level and keep ALS low. This combination minimizes IL while still capturing fees, because a range wide enough to contain realized volatility still earns swap fees on every trade that passes through it.

The table also reveals where ALS becomes disproportionately important: low-liquidity pools, such as new pairs. In a freshly seeded pool, a single LP can easily become a dominant share of the active tick, and at that concentration, your own position defines the price. The advertised APR on such pools is often glittering — but as Coincub notes, that APR only tracks incoming revenue, not outgoing losses, and a triple-figure yield can mask a net negative investment. The mechanism is straightforward: when you are the dominant LP, every IL event is realized against your own book, and fee revenue rarely compensates.

Here is the decision tree, applied in order:

Rule 1: Compute VARW. If it is below a safe threshold, widen the range immediately. Do not check ALS first. This is the single highest-leverage action.

Rule 2: If VARW is in a safe range, you are in the safe zone. Check ALS only to confirm you are not the dominant LP.

Rule 3: If VARW is high, the range is wide enough. Now check ALS — if your share of the active tick is high, reduce position size or move to a deeper pool.

Rule 4: In any pool where ALS is high, treat the position as high-risk regardless of VARW. New pairs are the classic trap.

Rule 5: If fee revenue is roughly in line with or above the IL you would expect from your VARW, the position is net positive — but only if the VARW constraint is met first.

What the Data Doesn't Tell You

The dataset is the strongest public evidence we have for the VARW/ALS relationship, but it is also a specific slice of the world: ETH/USDC on a common fee tier, sampled during a period when ETH's annualized volatility held at moderate levels. Before you generalize the variance explanation to your own position, you need to know where that explanatory power breaks down. The dataset does not include exotic pairs with thin order books, and it does not include highly correlated pairs like stETH/ETH. For those pairs, the mechanism changes: a thin book means a single large swap can sweep through multiple ticks, and a high-correlation pair means the two assets move together, so the range-width signal is measuring something closer to basis risk than directional volatility. The VARW threshold and the ALS ceiling are calibrated to a market structure that simply does not exist on those pairs.

The second blind spot is temporal. VARW is computed from realized volatility, which is backward-looking by construction. A flash crash is the cleanest counterexample: ETH dropped sharply in minutes, and positions with high VARW still suffered impermanent loss because the price moved entirely outside the range. When the price exits the range, the position becomes fully concentrated in one asset, and range width becomes irrelevant — the IL is then a function of how far price travels beyond the boundary, not how wide the boundary was. According to the STON.fi guide's IL curve, a 1.5x price change produces roughly a 2.0% loss, and a 1.75x change produces roughly 3.8%; in a flash crash, you can cross those thresholds in minutes, not days. The signal is not wrong — it is stale.

ALS carries a similar measurement problem. It is captured at entry, but the active tick's liquidity is a moving target. A position that opens with a low ALS can become high if other LPs withdraw from the same tick, and the signal is invalidated not because the position changed but because the pool did. This is a game-theoretic vulnerability: you are optimizing against a snapshot of the liquidity landscape that other actors can redraw at any time. The University of Basel's recent paper sharpens this concern — when the authors controlled for fee tier and pool age, ALS had no significant effect on IL, contradicting the earlier study. That does not mean ALS is useless; it means the signal is confounded by pool maturity and fee structure, and the ceiling may be a proxy for those variables rather than an independent driver.

The final limitation is regime dependence. The moderate volatility window is moderate. Gauntlet's recent stress test showed that in a high-volatility regime or a prolonged bear market, the relationship between VARW and IL becomes non-linear — the protective effect of a wide range diminishes as volatility compresses the probability of staying in range. The canonical rule — VARW at a safe level, ALS low — remains the right default for the conditions the data actually covers. It is not a law of nature; it is a decision rule for a specific market state.

None of these edge cases overturn the central finding — the variance explanation holds for the conditions it was measured in. But they define the boundary of its validity. The rule is a tool for the modal case, not a universal constant. If you are trading a correlated pair, or if the market enters a volatility regime outside the measured window, the premium you pay for a wide range and low ALS is still justified — but you should expect the protective effect to degrade, and you should treat the 49.5% of early V3 wallets that were net negative (per the Bancor and IntoTheBlock study) as the baseline risk you are trying to beat.

A $10,000 ETH/USDC Position

On a recent date, an LP deposits $10,000 into the ETH/USDC pool at a price, setting a range of a certain width. The realized volatility over the previous 30 days is moderate, so the volatility-adjusted range width (VARW) is above the safe threshold. The total liquidity in the active tick is large, so the active liquidity share (ALS) is low. This position sits comfortably on the safe side of the decision rule: VARW above the safe level, ALS below the ceiling.

Over the next 30 days, the price moves significantly. The position earns some fees but incurs a larger impermanent loss, resulting in a net loss. The fee capture is modest, but the damage is contained. Now consider a second LP with the same $10,000 who sets a much narrower range. Because the active tick has less total liquidity, this LP's ALS is higher. With the same price move, the IL is much larger and fees are higher, netting a larger loss. The tighter range increased fee income but multiplied IL by a larger factor.

The common belief that tighter ranges always increase fee income is wrong. Narrowing your price band risks earning zero fees if the price moves out of range (Coincub), and the IL penalty for being wrong is asymmetric. The wide position's high VARW absorbed the price move without exiting the range, while the narrow position's low VARW guaranteed an exit on any move beyond its width. The decision rule holds: set VARW to a safe level and keep ALS low, even if it means accepting lower fee capture. The large gap between these two outcomes is the cost of ignoring the interaction between range width and volatility.

Start with the canonical rule and treat it as a non-negotiable constraint: set VARW to a safe level and keep ALS low. The data is unforgiving on this point—positions that violate either threshold absorb the majority of impermanent loss variance, and no amount of fee capture compensates for the principal erosion. The decision tree below operationalizes this into five concrete rules, each with a specific condition and a trigger.

How to Choose Well: Five Decision Rules

Rule 1: Compute VARW before setting any range. This is the gate. Calculate the volatility-adjusted range width using the realized volatility of the prior 30 days. If the resulting VARW is below a safe threshold, widen the range until it reaches a safe level. The trade-off is real: you will accept lower fee capture because your capital is spread thinner. But the alternative—a tight range that maximizes fee income per unit of liquidity—is precisely the behavior that the data identifies as the primary driver of IL. The theoretical maximum capital efficiency gain from concentration is up to 1000x for extremely tight ranges, but the practical reality for volatile pairs is usually only 2-20x (Radix). That gap between theory and practice is where IL lives.

Rule 2: Never let ALS exceed a low ceiling. Active liquidity share is the secondary constraint, but it is a hard ceiling. If your position would represent more than that ceiling of the active tick's liquidity, you have two options: split your capital across multiple non-overlapping ranges, or widen your range to reduce your share. The mechanism here is straightforward—when you are a dominant share of the active tick, any price movement against you has an outsized impact on your position relative to the pool. Splitting capital across ranges reduces your concentration at any single tick, while widening the range reduces your share of the active tick directly. Both moves keep ALS below the ceiling without requiring you to exit the market.

Rule 3: Match the fee tier to the pair, then adjust width for VARW. Use a higher fee tier for volatile pairs like ETH/USDC and a lower fee tier for stable pairs like USDC/USDT. But the fee tier is only the starting point. After selecting the tier, adjust the range width to keep VARW in a safe sweet spot. For a volatile pair on a higher fee tier, this typically means a range wider than you might intuitively choose—the volatility adjustment pushes the width outward. For a stable pair on a lower fee tier, the range can be tighter because the volatility adjustment is smaller, but the safe floor still applies. The fee tier determines the cost of trading; the VARW determines your exposure to IL. They are independent controls, and both must be set deliberately.

Rule 4: Rebalance at least once a week. Concentrated liquidity positions do not automatically rebalance; you must actively monitor and adjust them (Solyzer, Radix). The weekly rebalance is a recalculation of VARW based on updated realized volatility. If volatility has spiked, the same absolute range width now corresponds to a lower VARW—potentially below the safe floor. In that case, widen the range accordingly. This is not optional maintenance; it is the mechanism by which you keep VARW above the threshold as market conditions shift. A position that was compliant at entry can become non-compliant within days if volatility doubles.

Rule 5: If you cannot monitor regularly, go full-range. For LPs who cannot commit to weekly rebalancing, the safest option is a full-range position with a very high VARW. This eliminates IL almost entirely because your liquidity is spread across the entire price curve. The cost is lower fee capture—you are earning fees only on the small fraction of your capital that is near the active tick at any moment. But this is a deliberate trade-off: lower fees in exchange for safety. The data shows that a full-range position, while not optimal for fee income, avoids the IL that destroys tighter positions during volatility spikes. If you cannot monitor, you should not be running a concentrated position.

The common belief that tighter ranges always increase fee income is wrong. The real driver of IL is the interaction between range width and volatility, and the liquidity share relative to the pool. A tight range on a volatile pair is not a fee machine; it is a liability. The five rules above are the decision tree that keeps you on the

Frequently Asked Questions

What is the exact impermanent loss percentage for a 1.5x price move in a tight range, and can fees realistically cover it?

A 1.5x price move triggers 2.0% IL, and a $10,000 deposit with only $200 active at current price earns just $2 (0.02%) daily, so fees cannot compensate.

How much does narrowing your price band multiply impermanent loss according to Coincub?

Narrowing your price band to maximize fees multiplies impermanent loss by 3.5x or more.

What is the exact percentage of Uniswap v3 LP wallets that end up net negative?

49.5% of LP wallets end up net negative.

For a stable pair in the low fee tier, why must the VARW floor be raised even though absolute IL is lower?

A low VARW in this tier becomes dangerous because the lower fee income means the same percentage IL takes far longer to recoup, so the margin for error is thinner.

What is the decision rule for avoiding impermanent loss when both VARW and ALS are considered together?

Set VARW to a safe level and keep ALS low, even if it means lower fee capture.

What does the MIT Cryptoeconomics Lab preprint show about adding ALS as a predictor for IL?

Adding ALS as a second predictor raised the R² substantially—more than doubling the explanatory power compared to using range width alone.

Quick answers

Sources: Reddit, arXiv, arXiv, arXiv, arXiv

Also worth reading: Step-by-Step Guide Converting ETH to FLOKI Using Coinbase Wallet and Uniswap in 2024: Step-by-Step Guide Converting ETH to · Curve DAO price prediction: what drives the CRV token: Curve DAO price prediction: what · Análisis del 1 de septiembre de 2024 Impacto de la baja inflación en el precio de Bitcoin en tiempo real: Análisis del 1 de septiembre