Impermanent Loss Formula, Stablecoin Risk and LP Ranges

Providing liquidity on decentralized exchanges (DEXs) can earn trading fees, but it also changes the assets you end up holding. If one token moves sharply against the other, the pool automatically sells some of the outperforming token and buys more of the underperforming token. That rebalancing is the source of impermanent loss (IL)².

The short answer to "how do you avoid impermanent loss?" is sober: you usually cannot remove it from a volatile AMM pair. You can reduce it by choosing assets whose price ratio stays close, such as stablecoin pairs, or by using strategies where expected fees, incentives and hedge protection are large enough to justify the loss scenarios⁴.

Stablecoin pools such as USDC/USDT normally have low impermanent loss because both assets target the same value. That does not make them risk-free. Depegs, thin liquidity, smart contract issues and low fees can still turn a quiet-looking pool into a bad position. The right question is not only "what is the IL percentage?" but "can this pool pay enough to cover IL, gas, depeg risk and the time my capital is locked?"

This guide starts with the plain impermanent loss formula, then moves into stablecoin pools, concentrated liquidity ranges, LVR and hedging. If you want to model the position instead of reading the math only, use the Concentrated Liquidity Simulator after the formula section to compare range width, token exposure, fees and out-of-range risk.

How Impermanent Loss Happens in AMM Pools

To grasp impermanent loss, you have to look under the hood of decentralized exchanges. Unlike traditional centralized exchanges (CEXs) and stock markets that match buyers and sellers using an order book, most DEXs run on an automated market maker (AMM) algorithm³. The standard AMM model, first popularized by Uniswap V2, uses a constant product formula.

It's written as a simple equation:

where x is the amount of the first token in the pool (like ETH), y is the amount of the second token (like USDC), and k is the pool's invariant or constant, which must stay mathematically unchanged after any trade, excluding added fees².

When a trader buys ETH with USDC on a DEX, they send USDC to the pool's smart contract and take ETH out. Because the pool now has less ETH and more USDC, the formula automatically bumps up the ETH price to keep the constant k intact⁴. Essentially, the price of an asset inside the pool is dictated entirely by the ratio of its reserves.

The weak point of classic AMMs is that their smart contracts live in a bubble; they don't know the actual market price of assets trading on other exchanges³. If the global price of ETH suddenly jumps due to outside news, the pool's internal price lags behind. This creates an obvious arbitrage opportunity: the pool is selling ETH much cheaper than its fair market value. Automated bots and institutional arbitrageurs immediately buy up the undervalued ETH from the pool, dumping depreciating USDC into it. They keep doing this until the pool's internal price matches the global market price³.

This rebalancing act is exactly what causes impermanent loss. Because of the arbitrageurs, the liquidity provider is left holding less of the asset that went up in price and more of the asset that went down². If the investor compares the final dollar value of the assets in the pool against what they would have had by simply keeping the tokens in a cold wallet (HODLing), they'll see a financial deficit. That deficit is the impermanent loss. It's called "impermanent" because if external prices somehow return to their original ratio, the loss vanishes. But if the investor cashes out their liquidity while the prices are still skewed, the loss locks in and becomes permanent³.

Impermanent Loss Example: ETH/USDC Doubles

Let's run through a classic math example. Imagine an investor provides liquidity to an ETH/USDC pool with a standard 50/50 asset split.

At the start, the external market price for 1 ETH is 3,000 USDC. The investor puts 1 ETH and 3,000 USDC into the pool. At this moment, their total investment is worth $6,000². Then the price shifts: ETH doubles on external centralized exchanges, hitting 6,000 USDC.

Arbitrageurs spot the price gap and aggressively buy ETH from the smart contract, filling it with USDC until the pool's internal price reaches 6,000 USDC per 1 ETH. Once the arbitrageurs finish balancing the pool, the investor's share looks very different. Based on the constant product formula, the position now holds roughly 0.707 ETH and 4,242 USDC. If they withdraw now, those assets are worth $8,484 (0.707 multiplied by 6,000 equals 4,242, plus the 4,242 USDC)².

Now let's compare this to passively holding (HODLing). If the investor had just kept the original 1 ETH and 3,000 USDC in their hardware wallet and avoided smart contracts entirely, their portfolio would be worth $9,000 (one ETH at the new $6,000 price, plus the original 3,000 USDC)¹⁰. The gap between passively holding ($9,000) and providing liquidity ($8,484) is $516. In percentage terms, that's a 5.7% loss. That is your classic impermanent loss¹⁰.

This highlights an important psychological point: the investor still made a profit overall compared to their initial $6,000 (they now have $8,484), but they made noticeably less than they would have by staying out of the liquidity pool. At its core, impermanent loss is an opportunity cost, not a direct deduction from your starting deposit¹.

Impermanent Loss Formula for a 50/50 Pool

The size of an impermanent loss depends entirely on how far the price ratio between the two assets has drifted since you deposited them. The direction of the move doesn't matter: a 50% price drop in one asset causes the exact same percentage loss relative to holding as a 100% price jump⁶. The AMM algorithm doesn't care about the trend; it just balances the reserves mathematically.

You can derive the impermanent loss formula for a classic 50/50 pool from the k invariant principles. It looks like this:

where d is the price ratio change, calculated as the new price divided by the old price². The formula is symmetric: if one token doubles against the other, the IL percentage is the same as if that token falls by half. Fees, incentives and gas are separate from this formula and should be added after the IL estimate.

These numbers show why the "APR is high" argument is incomplete. A 5x move means the LP gives up more than a quarter of the value they would have had by holding before fees are counted¹. In the extreme case where one asset collapses, the pool keeps buying the falling token to maintain the constant product. That can leave the LP with mostly the weaker asset while arbitrageurs remove the stronger one.

Impermanent Loss With Stablecoins

Stablecoin pairs usually have lower impermanent loss because the intended price ratio is close to 1:1. If USDC and USDT both trade near one dollar, the value ratio barely changes and the 50/50 IL formula stays close to zero.

The hidden risk is that "stable" is not the same as guaranteed. A temporary depeg can create IL, slippage, withdrawal stress and poor execution at the same time. A stablecoin LP should still check pool depth, historical depeg behavior, fee tier, protocol risk and whether the expected fees compensate for the scenario where one side trades below peg.

For live pool selection, start with Liquidity Pool Discovery to compare TVL, volume, fee tier and simulator readiness before deciding whether a stable pair or correlated pair is worth modeling.

The DeFi industry has also tried to reduce IL by changing pool weightings. Unlike standard Uniswap pools that use a 50/50 value split, protocols like Balancer can support weighted pools such as 80/20. Heavier weighting can reduce exposure to the smaller volatile asset, but it changes the portfolio trade-off rather than removing risk. If the heavily weighted asset falls, the position can still perform poorly¹⁰.

Concentrated Liquidity Impermanent Loss

A real technological breakthrough in market maker architecture arrived with Uniswap V3, which introduced concentrated liquidity. This massive upgrade changed the rules of the game entirely: it made position management much harder mathematically, but allowed hedging and providers to use their capital with unprecedented efficiency².

"In the old AMM design, a provider's capital was spread out thinly and evenly across the entire price curve, from zero to infinity. This meant that the vast majority of their money was always sitting idle. If the market price of ETH is bouncing between $2,000 and $3,000, the liquidity locked up for a hypothetical price of $10 or $100,000 never sees any action and earns zero fees." ¹⁵

Uniswap V3 gave providers the power to concentrate their capital within tight, custom price ranges. This lets you put your money exactly where the actual trading volume is happening⁷. The trade-off is sharper range risk. A narrow range can earn more fees while active, but it also reaches one-sided exposure faster when price leaves the band.

That is why concentrated liquidity should be modeled as a range decision, not only as an impermanent loss percentage. Compare the fee tier, expected volume, price range, time in range, token composition at the range edges and cost of rebalancing. DeFiRiskSim's LP range simulator is built around that workflow.

The Deep Math of Uniswap V3: Ticks and Liquidity

To make price limits work technically and avoid heavy floating-point math, the Uniswap V3 developers rebuilt the protocol's core. Instead of tracking exact reserves, the contract runs on the concepts of liquidity L and the square root of price √P. Liquidity is defined as the geometric mean of the reserves: L = √(x × y), and the price of token 0 in terms of token 1 is P = y / x¹⁶. Using √P lets the contract calculate token balance changes using simple linear equations:

To slice up the price curve, the infinite range was chopped into discrete intervals called ticks⁵. Every tick marks a specific price level. The relationship between tick index i and the current price relies on a logarithmic formula:

This scale guarantees a crucial financial property: the distance between any two adjacent ticks is always exactly one basis point (0.01%). Using whole-number ticks instead of heavy on-the-fly math drastically cuts down on expensive rounding errors on the Ethereum Virtual Machine (EVM). For extreme accuracy, prices are stored in Q64.96 format—rational numbers with a fixed point that use 64 bits for the whole number and 96 bits for the fraction¹⁶. To pull the actual price out of the sqrtPriceX96 variable, the contract divides the value by 296 and squares the result¹⁸.

By squeezing liquidity between two ticks (lower and upper), a provider massively increases their cut of the fees. The capital efficiency of providing liquidity between price a and price b, compared to the old zero-to-infinity method, is roughly:

For a range that sits just fractions of a percent away from the current price, this multiplier hits the tens of thousands¹⁷. But this financial superpower has a dark side. Higher capital efficiency automatically scales up the risk of impermanent loss². Because liquidity is artificially crushed into a short span, any sharp price move triggers a much faster and more aggressive rebalancing of the assets inside that corridor.

The critical failure point hits when the market price crosses the boundary and falls out of your chosen range. In that exact second, your position converts entirely into one asset (whichever one is now cheaper relative to the market), and the smart contract stops paying you trading fees until the price decides to return⁵. The investor gets stuck holding 100% of a losing asset, stripped of the cash flow that was supposed to offset the loss. This mechanic forces professionals to ditch passive investing in favor of active, algorithmic management systems¹⁹.

The Art of Setting Price Ranges

Picking the right price range is a constant balancing act between aggressively farming fees and shielding yourself from market volatility. Professional strategies demand an analytical approach rooted in probability models, historical data, and projected price swings¹².

Volatility Analysis and Distribution Strategies

To figure out the best range width, analysts usually lean on classic volatility metrics like the Average True Range (ATR) or Implied Volatility calculated from options pricing¹². If the indicators point to heavy market turbulence, setting a hyper-concentrated range is a strategic blunder: your position will instantly drop out of bounds, eating maximum impermanent loss at the edge without having had the time to farm enough fees to justify the risk.

In the academic paper "Strategic Liquidity Provision in Uniswap v3" by Gamma Strategies, researchers suggest using expected price range strategies. They crunched historical ETH price data in 10-minute blocks and proved that liquidity provision should be based on the statistical probability of the price landing in a specific corridor¹². They highlight two main ways to distribute capital:

1. Uniform strategy: You spread your liquidity evenly across the entire selected range. This is how the default Uniswap V3 interface does it.
2. Proportional strategy: You break your capital down into many smaller sub-ranges, and you inject funds based on the statistical odds of the price landing in that specific sub-range. The middle zones get the most liquidity, while the edges get the least. Research proves this method delivers the best risk-to-reward ratio for most providers²³.

Choosing between a tight position and a wide spread comes down to the fundamental nature of the assets. For stable pairs, the risk of a severe price depeg is tiny, so tight ranges let you max out your APR without constantly worrying about falling out of the earning zone¹. On the flip side, highly volatile pairs demand wide ranges. This ensures you stay in the active zone longer, drastically cuts down how often you have to pay Ethereum gas fees to rebalance, and slows down the math curve of your impermanent loss¹⁰.

Single-Sided Provision and Limit Orders

Uniswap V3's unique design lets you use ranges to neatly mimic limit orders (Range Orders)²⁰. If an investor wants to buy ETH below the current market price, they can set a price range strictly below the current spot price and fill it entirely with stablecoins (like USDC)²⁰. As the price of ETH drops and slides through that range, the smart contract steadily swaps the USDC for ETH. The massive advantage here over centralized exchange limit orders is that while the price is moving inside the range, the investor is also collecting trading fees for providing liquidity². But there is a strict catch: if the price slices all the way through the range and then bounces back before the investor closes the position, the trade essentially "cancels," automatically swapping the ETH back into USDC. You have to pull your assets out on time²⁶.

Paradigm Shift: From IL to LVR (Loss-Versus-Rebalancing)

As DeFi infrastructure matures, institutional players, market makers, and quants are moving away from the classic IL concept. Instead, they track a much harsher financial metric: LVR, or Loss-Versus-Rebalancing⁶. Truly understanding LVR is the hallmark of deeply grasping AMM macroeconomics.

The Fundamental Flaw of Classic IL

The main academic problem with impermanent loss is that it's path-independent. It only takes a static snapshot of your position at two points in time: when you deposit and when you withdraw⁶.

Picture a hypothetical scenario: you put money into an ETH/USDC pair when ETH is at $3,000. Over the next month, the market goes crazy, swinging wildly between $2,000 and $4,000, but by the time you withdraw, the price miraculously lands exactly back at $3,000. Classic IL math will spit out a flat zero, suggesting you lost nothing³. But the laws of financial physics dictate that something destructive happened. During every single one of those hundreds of price swings, arbitrageurs executed trades against your pool at lagging, unfavorable prices, relentlessly bleeding microscopic slivers of value out of it³⁰. The standard IL calculation completely ignores this hidden drain on your working capital, feeding you a dangerous illusion of breaking even.

The Math and Essence of LVR

Loss-Versus-Rebalancing puts a hard number on the real value that arbitrageurs extract from the pool by exploiting the AMM's lagging prices³⁰. Wall Street calls this the cost of "toxic flow" or adverse selection⁸. On centralized exchanges, high-frequency market makers can pull and update their quotes in milliseconds when the market shifts. But smart contract pools sit there passively. They are forced to wait for an external agent to execute a trade before they update their internal price²⁹. This architectural flaw guarantees that every single trade that aligns the pool's price with the global market is profitable for the informed arbitrageur and a direct loss for the liquidity providers²⁹.

Conceptually, LVR is framed as the difference between the actual financial value of the assets in the pool and the value of a hypothetical "rebalancing portfolio." This rebalancing portfolio is an idealized math model of a portfolio that makes the exact same restorative trades as the pool, but executes them on a zero-fee centralized exchange perfectly—hitting the fair spot price with zero slippage or lag²⁸.

In continuous-time models, the instantaneous LVR for a constant product pool is elegantly approximated as:

where σ is the asset's volatility and Δt is the time interval. The equation makes it obvious that the loss doesn't accumulate linearly; it grows strictly proportional to the square of volatility⁶. Unlike impermanent loss, which can roll back to zero if the price returns, LVR is a constantly growing, predictable sunk cost. Every arbitrage trade permanently extracts a chunk of the pool's value, and you never get that money back⁶. Consequently, for liquidity provision to actually make economic sense, the trading fees your pool earns must consistently beat not just theoretical IL, but this endlessly bleeding LVR metric²⁰. Researchers emphasize that modern analysis now focuses on evaluating a pool's "revenue density"—checking if the invested capital generates enough real economic activity to pay for the damage done by toxic flow³³.

Advanced Position Management: The Futures Illusion and Options Power

Faced with inevitable IL and the toxic drain of LVR, major market players build hedging (insurance) strategies to defend their positions. One of the most widespread myths in DeFi is that you can protect your position just by opening a short trade on futures. A deep structural teardown proves this approach is fundamentally flawed and actually accelerates the burn of your deposit²⁴.

The Futures Hedging Illusion

A beginner's logic usually goes like this: "Putting liquidity into an ETH/USDC pair gives me something similar to a long position on Ethereum. So, if I open a short on perps for the equivalent amount of ETH, my position becomes market-neutral, and I can safely farm 50% APR."⁹

The fatal algorithmic error here lies in the geometry of the yield curves. Futures are strictly linear financial instruments³⁴. The profit or loss on a futures contract goes up and down in a straight line. Impermanent loss, however, has a distinctively convex, non-linear (quadratic) profile²⁴. The further the price runs from your entry point, the faster your losses compound exponentially compared to just holding⁹.

Look at what happens if ETH doubles while you hold a futures-hedged position:

1. The linear futures short zeroes out the direct paper profit from the base asset (ETH) going up.
2. Meanwhile, as the price climbs, the AMM smart contract auto-rebalances, dumping your appreciating ETH to arbitrageurs and stacking depreciating USDC in your reserves. The mathematical scissors of impermanent loss start cutting.

With this strategy, you end up taking a heavy loss. The linear hedge simply cannot offset the exponentially curving losses of the pool⁹. Hedging with futures successfully kills the directional risk of the asset itself, but it does absolutely nothing against the curvature of the pool's algorithmic rebalancing. You just end up paying funding rates on your futures while watching your balance melt away⁵.

Options Power and The Greeks

The only class of derivatives that can mathematically map and insure against the non-linear curve of AMM pools is options³⁴. Options carry a unique parameter called "Gamma"—a measure of the contract's curvature (convexity), which allows their payout graph to perfectly mirror the impermanent loss graph²².

Academic studies, relying on the famous Carr-Madan spanning formula, prove that providing liquidity in concentrated pools is algorithmically identical to selling an infinite continuum of cash-secured strangles or put options³⁵. By handing capital to a smart contract, the investor acts as a classic volatility seller, earning a regular "premium" via trading fees. Therefore, a market maker's position holds negative Gamma¹⁶.

To hedge this properly and neutralize the non-linear risk, the investor has to become an active volatility buyer using these advanced strategies:

• Long Strangle Strategy: You simultaneously buy out-of-the-money Call and Put options. If the market price makes a violent swing in either direction, the options deliver a non-linear profit that easily covers the impermanent loss from the pool's rebalancing¹⁰.
• European and Asian Puts: Research indicates that hedging an LP portfolio with classic European put options can be optimized by swapping to Asian puts. Because they rely on averaged prices, Asian puts smooth out the shock of sudden volatility spikes on Gamma and Delta²².
• Collars: This strategy involves selling a Call option (capping your upside profit) to pay for a protective Put option. This drastically slashes the overall dollar cost of the hedge¹⁹.

The main algorithmic headache with options is their high internal cost. Options suffer from relentless time decay, known as Theta⁹. Holding an option in your portfolio costs you real money every single day. In reality, the premium you pay for the insurance often entirely eats up the massive APR from the pool. That exposes the fundamental law of the market: sky-high DeFi yields are not free money; they are direct compensation for taking on convex price risk⁹.

Maximum Profit and Minimum Loss: Position Management Rules

By fusing academic research, LVR analysis, and the advanced algorithmic strategies of decentralized market makers, we can outline a core playbook for managing positions professionally during extreme market volatility²⁴:

1. Range Laddering: One of the safest tactics is skipping a single pool and splitting your capital into tranches. Instead of one range, you build three: a tight core range around the spot price to capture heavy fee flow, buffer ranges above and below to cut rebalancing costs during market noise, and a super-wide insurance range to survive flash crashes. This stops your portfolio from bleeding out on gas fees when you constantly get knocked out of bounds²⁴.
2. Delta Risk Awareness and Hedging: Aggressively tightening your price range without setting up a counter-hedge is basically a naked bet that the price will freeze in place. If you run a tight range on a volatile pair, you must pair it with delta hedging (dynamically shorting futures to control your delta) or put options to protect against a dump²⁴.
3. Control Your Insurance Budget: Options are expensive. You should pay for regular protective contracts using a strict, predefined cut of your trading fees (usually 20–40%), rather than draining your core capital. This builds a stable, self-funding risk management loop²⁴.
4. Fail-Fast Rules to Protect Deposits: Hardcode algorithmic emergency exits. If funding rates flip violently and price swings blow past historical norms, the safest move is to temporarily pull your liquidity into stablecoins. This stops catastrophic LVR buildup and prevents emotional "revenge trading"²⁴.
5. Focus on Revenue Density and Fee Tiers: Picking the right fee tier is make-or-break. A fee that is too low (0.05%) on a highly volatile pair is mathematically guaranteed to fail against your accumulated LVR. High revenue density happens in pools with maximum trading volume but moderate total value locked (TVL)¹⁵.
6. TWAP Execution (Time-Weighted Average Price): If you are rebalancing massive positions during thin overnight liquidity, your transfers and swaps must be algorithmically sliced into smaller chunks. This prevents brutal price slippage and blocks MEV (Maximal Extractable Value) bot attacks²⁴.

For most retail investors who don't have the resources to build options models or pay for expensive Ethereum rebalancing transactions, sticking to stablecoin pools or strictly correlated assets (like USDC/DAI, ETH/stETH) is the simplest way to reduce IL and LVR. These pairs can still carry depeg, liquidity and smart contract risk, so the lower IL profile should be treated as one input, not a blanket safety label⁴.

Until LP management tools can reliably account for fees, volatility, gas, slippage and adverse selection in one place, mathematical discipline remains the liquidity provider's best defense. A position should pass a simple test before deposit: projected fees should exceed plausible IL and LVR scenarios after costs.

Impermanent loss is the price risk embedded in automated liquidity. By understanding the formula, checking stablecoin and correlated-pair assumptions, modeling concentrated ranges and comparing fee income against LVR, you can treat LP exposure as a measurable DeFi risk rather than a vague yield promise.

FAQ

Source Materials & Links

1. SpeedrunEthereum - IL Math Explained, https://speedrunethereum.com/guides/impermanent-loss-math-explained
2. Chainlink - Understanding Impermanent Loss, https://chain.link/article/impermanent-loss-defi
3. ChainUP - A Guide for DeFi Liquidity Pools, https://www.chainup.com/blog/impermanent-loss-explained-a-guide-for-defi-liquidity-pools/
4. Binance Academy - Impermanent Loss Explained, https://www.binance.com/en/academy/articles/impermanent-loss-explained
5. Changelly - What is Impermanent Loss, https://changelly.com/blog/what-is-impermanent-loss/
6. Uniswap V3 Book - Different Ranges, https://uniswapv3book.com/milestone_3/different-ranges.html
7. Nansen - What is Uniswap V3, https://nansen.ai/post/what-is-uniswap-v3
8. Gamma Strategies - Expected Price Range Strategies, https://medium.com/gamma-strategies/expected-price-range-strategies-in-uniswap-v3-833dff253f84
9. Arxiv - Strategic Liquidity Provision, https://arxiv.org/html/2106.12033v5
10. Jobcannon - IL Hedging, https://jobcannon.io/skills/impermanent-loss-hedging
11. Mikhail Liublin - The Illusion of Hedging with Futures, https://medium.com/@mikhailliublin/the-illusion-of-hedging-impermanent-loss-with-futures-e977678d4165
12. Dev.to - Why You Can't Hedge IL with Futures, https://dev.to/mikhailliublin/why-you-cant-hedge-impermanent-loss-with-futures-5f2f
13. Aperture Finance - Automating Tools for Uniswap V3, https://github.com/Aperture-Finance/uniswap-v3-automan
14. Gelato Network - Liquidity Management on Uniswap V3, https://gelato.cloud/blog/liquidity-management-on-uniswap-v3
15. Medium - Aperture Finance Tools & APY Mechanics, https://medium.com/@dreykerr/aperture-finance-univ3-tools-b6703c4eb95e
16. Typefully - Aperture Finance Automated Actions, https://typefully.com/FLOAT_LOCKER/aperture-finance-automated-uniswap-v3-jhKY20H
17. Bankless - Automating Uniswap V3 Management Safely, https://www.bankless.com/how-to-automate-uniswap-v3-liquidity
18. Uniswap Docs - Range Orders Strategy, https://developers.uniswap.org/docs/get-started/concepts/liquidity-providers/range-orders
19. Uniswap V3 Book - Setting Optimal Price Ranges, https://uniswapv3book.com/milestone_3/different-ranges.html
20. RareSkills - Concentrated Liquidity Mathematics Guide, https://rareskills.io/post/uniswap-v3-concentrated-liquidity
21. Uniswap V3 Book - Comprehensive Math and Code Mechanics, https://uniswapv3book.com/milestone_0/uniswap-v3.html
22. Uniswap Blog - AMM Volatility and Mathematical Primer, https://blog.uniswap.org/uniswap-v3-math-primer
23. Medium - Capital Efficiency metrics of Concentrated Liquidity, https://tienshaoku.medium.com/uniswap-v3-maths-explained-capital-efficiency-86257c44405a
24. Binance Square - Returns Convexity & Option Volatility, https://www.binance.com/en/square/post/706655748562
25. Eco - Modern DeFi Lending and Interest rate structures, https://eco.com/support/en/articles/14800882-best-defi-lending-protocols-2026-tvl-rates-risk
26. Eco - Solana Defi Ecosystem Overview, https://eco.com/support/en/articles/15083170-best-solana-defi-protocols-2026
27. FinTech Weekly - High revenue density and yield options, https://www.fintechweekly.com/magazine/articles/defi-capital-efficiency-tvl-revenue-density-institutional-2026
28. Moallemi - Loss-Versus-Rebalancing (LVR) Paper (2022), https://moallemi.com/ciamac/papers/lvr-2022.pdf