Stablecoin Pool Result: 30-Day Standard Full-Range Winner

The 30-day stablecoin-pool winner is standard full range—not because concentrated liquidity cannot win on fees, but because its extra density is conditional. A concentrated range is a fee auction plus an inventory option: while price remains inside it, fees accrue; after an exit, the LP is entirely exposed to one token and fee collection stops. Full range keeps both assets deployed throughout the comparison, making it the passive benchmark.

A 15% yield figure from FinanceFeeds is a generic estimate, not a measured result for this pool. STON says fees can offset or exceed losses, while cited studies report net-negative concentrated positions. Those statements are general claims, not guarantees. Coincub attributes greater impermanent loss to narrow ranges, but that is its claim, not a universal result. The test is realized net performance, not advertised yield.

Skrumble's endpoint price-ratio examples are not a 30-day path and do not measure the unnamed pool; its cited study figures may also be old or unverified. A defensible ledger instead needs cumulative fees, active days, range crossings, withdrawal value, and the HODL benchmark used to define impermanent loss. Until concentrated liquidity clears that benchmark after model error, standard full range remains the rational passive default.

Standard Versus Concentrated Liquidity

For the fixed data freeze, I would define standard as an equal-notional Uniswap v2 full-range LP and concentrated as one fixed-range Uniswap v3 LP over the same 30 UTC days. I would hold the starting P0, x0, y0, direct USD fair value, and every incremental mint, swap, collection, and exit cost constant; only the liquidity geometry changes.

In the v2 constant-product construction, x*y=k and quote value is V=xP+y. The equal-value initializer sets x0P0=y0, while the deployed standard comparison pool charges its selected fee. Each swap’s fee accrues pro rata according to reserves, but the LP does not trade to manage composition: arbitrageurs move the pool toward the external price, shifting its inventory from balanced toward one-sided as price moves.

For v3, liquidity L is indexed to ticks defining the interval [Pa,Pb]. Below Pa, the position consists entirely of token0; above Pb, it consists entirely of token1. It earns no swap fees outside those bounds, and no trade automatically re-centers it. Crossing a boundary therefore does not repair the inventory: the position remains on the side that cannot earn until price returns or an operator intervenes.

According to Uniswap’s official v3 documentation, v3 pools use multiple fee tiers. For equal routed volume V, compare gross v3 fees earned on V_active, the volume executed while the position is in range, with gross full-range v2 fees earned on V, before applying the position’s liquidity share. For the fee-normalized sensitivity, I would rerun the same range, path, inventory, and gas ledger at every official tier; otherwise, a lower swap fee could be mislabeled as a concentration effect.

I would record signed impermanent-loss cost as D = HODL value − LP value before fees. On a common initial-USD scale, v2 has D = 1 − 2*sqrt(r)/(1+r), where r=P30/P0. For v3, I would reconstruct token amounts tick by tick and mark both branches to direct USD fair value, consistent with Algebra’s stablecoin convention, so a depeg cannot be hidden by nominal parity. Net USD excess is fees − D − gas − slippage. Skrumble’s withdrawal-value and HODL benchmark are essential, especially when an ETH move leaves the stablecoin-pair position one-sided.

The defensible deployed result here is structural rather than a fabricated return percentage: the named evidence supplies no pool-level net-return series for the fixed window. STON says fees can offset or exceed divergence, while the catalogued net-negative outcomes point the other way; both are general claims, not guarantees. Unless the concentration exception is documented in the canonical ledger, standard v2 remains the winner.

According to the Uniswap v3 whitepaper, capital efficiency depends on the selected range. That is a protocol-design parameter, not an observed stablecoin return. The boundary that packs inventory into a narrow band also stops fee capture when price exits and leaves the LP one-sided; impermanent loss and gas can reverse the apparent ranking. A capital-efficiency multiplier therefore cannot imply commensurate realized profit.

Endpoint Examples and Missing Evidence

According to Skrumble’s May 19, 2026 methodological note—used only for interpretation, not as a post-freeze return input—the 2x/4x/10x figures are endpoint price-ratio examples, not a 30-day forecast or a measurement of the unnamed pool. No source gives a 30-day path, so a 30-day IL cannot be inferred from those figures. Multipliers belong in a mechanism field; only timestamped fees, inventory, gas, and a matched HODL series can enter the net-return calculation.

The remaining evidence belongs in a versioned ledger, not a collection of disconnected dashboard APYs:

No archived query IDs, transaction hashes, receipt-derived medians, or matched APY snapshots were supplied in the research packet, so this section asserts no fabricated values. Until those artifacts are populated, the table remains an audit specification rather than evidence of concentrated outperformance. Any missing range path, fee series, inventory mark, gas record, HODL comparator, or synchronized timestamp fails the fixed-concentrated case; the canonical standard full-range position therefore remains the choice.

The scorecard records standard full-range as the net-return winner, not because it dominates every path, but because concentration receives no presumption of victory. The source data supplied with the article brief contains no 30-day volume, hourly range ledger, terminal-price series, or net LP P&L with which a concentrated position could clear both gates. “Not supplied” is neither zero nor a pass.

Nexumo’s approximation in “10 LP Hedging Tricks for Vertical Volatility” supports subtracting divergence and execution costs from fees; adding a negative signed IL here prevents double subtraction. According to Binance’s liquidity-provider guidance, CLMM fees stop outside the active range, so elapsed days cannot substitute for hourly observations. The source “What Are Concentrated Liquidity Market Makers (CLMMs)?” treats IL amplification as conditional. Coincub’s assertion that a narrow range multiplies IL is a source-specific claim, not a universal coefficient and never a Winner input.

Net 30-Day Scorecard

Both rows must use one named USD-close source, the same scoped incremental-approval policy, and identical actual-gas treatment: count every required approval, entry, exit, and rebalance transaction at its execution-time USD mark without netting. The brief neither identifies the USD source nor supplies gas and slippage ledgers, so those inputs cannot be silently imputed. Deployed fees remain primary; any fee-normalized counterfactual stays separate and cannot overwrite the all-in decision.

Next, replace every “not supplied” cell with a reproducible hourly-close or transaction-ledger reference and rerun both tests. Failure, a tie, or missing evidence keeps the standard row’s Winner label.

The verdict is a conditional default, not a law of automated market makers. The supplied source set is flagged as potentially incomplete: without path-level prices, finer bars, and swap-time liquidity histories, it can support a rule-bound result, not a universal claim. A concentrated premium earns consideration only if the replay survives the falsification tests below; every failed, tied, or unresolved test leaves standard full-range liquidity selected.

In the prescribed peg round trip, price crosses the range’s upper boundary and returns to its starting level. Re-entry can leave concentrated inventory one-sided, although an endpoint-only Uniswap v2 comparison still reports no divergence. The position must therefore be reconstructed from the complete path. The capital-efficiency multiplier is not a realized-profit multiplier: fee capture stops at the boundary, while gas and one-sided inventory can reverse the ranking.

Counter-Evidence

A close-sampled month can conceal the event that matters most. A short peg dislocation could dominate fees and impermanent loss in an otherwise quiet interval. Results should therefore report dispersion across stablecoin pairs and peg regimes; one calm interval cannot support a general stablecoin claim.

Observation frequency is model structure, not cosmetic preprocessing. Finer sampling can change both the active-hour count and reconstructed exit inventory. Archive both resolutions, pre-designate the canonical hourly-close aggregation, and treat the alternative resolution as a robustness replay before calculating returns.

Perturb both decision cutoffs in opposite directions. If the selected position changes within the sensitivity band, the finding is threshold-sensitive rather than definitive. This does not promote the challenger: the exact cutoffs remain the decision rule, and uncertainty resolves to standard.

Fee share depends on contemporaneous competition, not a static TVL headline. More competing Uniswap v3 liquidity can dilute fees even when volume is flat. Replay active liquidity at every swap, including liquidity births and exits; treating a current TVL snapshot as constant would bias the fee-share estimate.

An active strategy that rebalances both sides after a range break is genuine counter-evidence to passive implementation, but not an automatic exception. Simulate the resulting swap fees, Ethereum gas, price-impact slippage, latency, and execution risk. Deep liquidity and routing may reduce slippage, but they do not make execution free. Credit an improvement only after the challenger clears the fixed-range activity and net-edge gates on a gas-inclusive, impermanent-loss-adjusted basis; otherwise, standard remains the winner.

Standard full-range liquidity is the decision winner, but this case reaches that verdict through the article’s evidentiary fail-safe rather than a verified return spread. According to the source-data notice supplied with the article brief, the available material contains no numerical fee, impermanent-loss, or execution-cost outputs. The required Dune query, raw exports, CoinGecko hourly series, and Etherscan receipts are also absent. A contract address is not an archive, and treating missing observations as zero would manufacture evidence.

The v3 test must remain preregistered: the fixed lower and upper bounds around P0 must be declared before examining any hourly close. The lower bound is rounded down and the upper bound rounded up to valid ticks; those ticks, P0, and the starting token amounts then determine starting liquidity. For each swap, the fee credit is the actual fee multiplied by the position’s pre-swap liquidity share, followed by the corresponding inventory change. Without the pool-event export, neither the active-hour count nor a potential one-sided terminal balance can be reproduced.

USDC/USDT Case

Capital-efficiency multipliers cannot replace this cash-flow audit: fee accrual can stop at the boundary and leave inventory one-sided. The operational lesson is narrow but important—preserve the query, raw rows, hourly marks, and receipts before a concentrated position can pass; any missing artifact is a failed test, not permission to assume parity.

For the fixed data freeze, this is an evidence-gated decision tree, not a forecast: the first failed gate selects standard full-range liquidity. Concentrated liquidity receives no presumption from capital efficiency. A higher efficiency multiplier does not create a proportional increase in realized profit because fee capture stops at the range boundary, while terminal inventory, impermanent loss, and execution costs remain. Evaluate the 30-day candidates in sequence; an exact tie selects standard.

The v2 calculation is independently auditable: mainnet reserve events establish starting reserves and the exact USDT-per-USDC price; each fee uses the position’s changing reserve share; token inventories remain in native units. Gas is separate from impermanent loss: Etherscan receipts supply gas used and the historical effective gas price, while the fixed passive position assumes no rebalance transaction but still incurs entry and exit costs. Marking a terminal balance does not create slippage; slippage enters only if an actual conversion is executed.

Capital-efficiency multipliers cannot replace this cash-flow audit: fee accrual can stop at the boundary and leave inventory one-sided. The operational lesson is narrow but important—preserve the query, raw rows, hourly marks, and receipts before a concentrated position can pass; any missing artifact is a failed test, not permission to assume parity.

Decision Tree: Range Survival, Then Net Economics

For the fixed data freeze, this is an evidence-gated decision tree, not a forecast: the first failed gate selects standard full-range liquidity. Concentrated liquidity receives no presumption from capital efficiency. A higher efficiency multiplier does not create a proportional increase in realized profit because fee capture stops at the range boundary, while terminal inventory, impermanent loss, and execution costs remain. Evaluate the 30-day candidates in sequence; an exact tie selects standard.

What to do next