Skip the APR Trap: Solana Pools That Win on Risk-Adjusted Yield

31.44 beats 31%. That’s what happens when you price risk.

Chasing the highest APR is the easiest way to underperform as an LP. Your PnL lives or dies on two things: whether fees keep showing up and whether your inventory survives the path to those fees. That’s exactly why we publish farmer_score, risk_score, and a farmer-to-risk ratio (RAR). When you stack this week’s Solana pools by RAR instead of headline APR, the list looks nothing like your DEX front page.

APR is a vanity metric; repeatable fees per unit of risk is the business.

Below, I’ll spell out what the scores actually measure, then walk through the live leaders and the high-APR standouts that fall down the order once you penalize for risk.

What farmer_score and risk_score actually measure

WealthVille distills a large set of on-chain and market signals into two numbers:

• farmer_score (0–100, higher is better): How repeatable your earnings are likely to be over the next trading week or two. It’s driven by realized fee density, turnover persistence, and whether the pool’s structure is aligned with where trades actually happen.
• risk_score (0–100, lower is better): How fragile the position is to tail scenarios you care about as an LP: program risk, inventory blowouts, liquidity cliffs, stale routing, and custody hazards on the quote side.

Concretely, farmer_score combines:

• Recent fee APR quality (not just peaks) vs 7/30-day baselines.
• Turnover consistency: 24h volume relative to TVL and its variance.
• Where fills occurred vs typical CL ranges/bins (are you in the action or off-range).
• Fee tier efficacy (is the pair priced where takers are crossing).
• Emissions reliability, if any, and whether they actually accrue to LPs.

And risk_score penalizes:

• Program surface area and upgrade authority of the AMM (concentrated AMMs differ from bins/DLMMs).
• Token-side risks: depeg profiles for stables, issuer and treasury behavior, and kill-switch governance.
• Liquidity concentration and cliffiness (how far to next meaningful depth).
• Router fragmentation and dependency on a single path.
• Operational quirks that increase IL or miss fees (outages, stuck ranges).

The RAR you see on pool pages is the farmer-to-risk relationship expressed as a single figure. Higher is better. Treat it as a sizing signal, not a binary switch. It rewards pools that pay consistently while scoring clean on tail risk, and it demotes pools that flash but break when they hit turbulence.

This week’s risk-adjusted leaders (not the APR leaderboard)

Sorted by RAR, here’s the live slate and why each ends up where it does when you price risk instead of clickbait APR:

• SOL-USDC on Orca Whirlpool — TVL $32.53M, 24h volume $230.67M, fee APR 0.6%, farmer 50/100, risk 2/100 (pool Czfq3xZZ…) — RAR 31.44. Turnover is 7.09x TVL in a day. That’s a freight train. A midline farmer_score paired with a very low risk_score produces the highest ratio on the board. Your base pair is the network reserve asset vs the dominant stable. Boring. Reliable. Profitable.
• SOL-GUNZ on Orca Whirlpool — TVL $33.84M, 24h volume $56K, fee APR 0.0%, farmer 33, risk 1 (pool EsHdGQHE…) — RAR 22.97. Odd mix: near-zero 24h fees but a featherweight risk read. As a result, the ratio still screens high. It’s a watchlist candidate until fees actually print; allocate small if at all until turnover wakes up.
• SOL-LCAI on Orca Whirlpool — TVL $31.26M, 24h volume $347K, fee APR 0.0%, farmer 34, risk 2 (pool AwdyoUww…) — RAR 19.54. Similar story: clean risk, muted earnings. A low-cost optionality slot if you want exposure and are patient.
• XMR-USDC on Raydium CLMM — Two pools with near-identical stats: TVL $22.64M/$22.41M, 24h volume $297K/$338K, fee APR 0.1%, farmer 34, risk 3 (pools 9RG3MPAd… and AFZnNpoi…) — RAR 11.58 / 11.50. Thin but steady. If you want non-correlated inventory, the ratio is serviceable, not spectacular.
• GFT-USD1 on Raydium CLMM — TVL $19.27M, 24h volume $11K, fee APR 0.9%, farmer 32, risk 4 (pool De8ckf1R…) — RAR 9.23. APR reads OK, but volume is microscopic relative to TVL, so it won’t scale. Glide path only.
• SOL-USDC on Raydium CLMM — TVL $5.00M, 24h volume $10.60M, fee APR 31.0%, farmer 73, risk 8 (SOL-USDC) — RAR 8.64. Headline winner by APR. Not the RAR winner. The 8/100 risk score knocks it down the stack. Great for active LPs who can manage range and inventory tightly.
• Old Slerf-SOL on Raydium AMM — TVL $14.58M, 24h volume $2K, fee APR 0.0%, farmer 30, risk 5 (pool AgFnRLUS…) — RAR 6.54. Legacy liquidity with almost no recent take. Fine for nostalgia; not for fees.
• SOL-USDT on Raydium CLMM — TVL $1.38M, 24h volume $5.34M, fee APR 14.2%, farmer 66, risk 12 (SOL-USDT) — RAR 5.60. Healthy APR, but the highest risk score on the page. The ratio reflects that.
• SOL-USDC on Meteora DLMM — TVL $3.20M, 24h volume $22.58M, fee APR 0.3%, farmer 52, risk 10 (pool 5rCf1DM8…) — RAR 5.51. Huge 24h turnover vs TVL (7.06x). DLMM structure introduces different tail dynamics; risk marks higher, RAR lands mid-pack.

The surprise is the absence of memecoin pairs at the top. Why? Skinny TVLs and lumpy taker flow jack APRs on the day they pump, then disappear. That inflates farmer_score briefly but can nuke risk_score for the next week. RAR catches that whipsaw.

High APR that fades on risk: two clear examples

If you scroll DEX front pages, you’ll be tempted by double-digit APR on majors. Two headliners this week, and how the ratio reframes them:

• SOL-USDC on Raydium CLMM is posting 31.0% fee APR on $5.00M TVL and $10.60M 24h volume. That’s attractive and very LP-able. On RAR, it drops to 8.64 thanks to a risk_score 8/100. You’re paid, but you’re working for it: concentrated ranges, faster inventory cycling, and program surface.
• SOL-USDT on Raydium CLMM shows 14.2% fee APR on $1.38M TVL and $5.34M 24h volume. The risk_score 12/100 is the outlier on the page, pulling RAR down to 5.60. The quote-side profile and routing dependencies matter here.

Contrast those with SOL-USDC on Orca Whirlpool (RAR 31.44) where the risk_score 2/100 and a 7.09x daily turnover-to-TVL make a modest fee APR (0.6% for the day) look far more investable on a risk-weighted basis. I’d rather fund the boring train for two weeks than chase a flashing sign for two days.

If this sounds familiar, it is. We’ve been making the same case since Boring Wins: The Solana Pools Beating 500% APR Traps.

How to read and use RAR like a pro

Think in bands, not absolutes

• RAR 20–35: Core. Fund it with size. You’re getting repeatable fees against a very clean risk read.
• RAR 10–20: Satellite. Track closely; scale with fee density persistence.
• RAR 6–10: Opportunistic. Short-dated deployments, active range management, and stricter stop rules.
• RAR <6: Watchlist. Needs a catalyst or a structural change before funding.

Position sizing and checks

• Start with your base spread: e.g., 60% in RAR ≥20, 30% in 10–20, 10% in 6–10.
• Run a turnover sanity check: 24h volume/TVL ≥2x scales; <1x means your fees rely on range micromanagement.
• Prefer pools where your expected fill zone matches realized fills (farmer_score helps you see that).
• Cut quickly when RAR degrades by two bands or risk_score jumps ≥3 points on a maintenance update or token-side event.

Rebalancing cadence

• Weekly for majors (SOL-USDC, SOL-USDT) unless volume regime shifts earlier.
• Twice weekly for thinner pairs that live in the 6–10 band.

If you want a curated starting point, the Best Solana pools page applies these filters live, and AI Signals flags when a pool’s farmer_score or risk_score crosses your thresholds.

Execution notes: AMM structure and fee math matter

Three AMM structures appear in this week’s board. Each treats price paths and fee capture differently:

• Orca Whirlpools concentrate liquidity into ticks with classic CL behavior and clear fee tiers. See the Whirlpools docs for mechanics.
• Raydium CLMM has its own CL implementation, routing pathways, and vaulting. The CLMM docs outline tick spacing, fee tiers, and positions.
• Meteora DLMM uses dynamic bins. Fee accrual depends on bin placement and bin size, which shifts how you think about “range.”

Two implications for RAR users:

• Range discipline beats heroics. On CLMMs, most of your earnings come from being in-range during high-volume windows. If your farmer_score is high but your realized fills keep missing where takers cross, your personal RAR will lag the pool’s headline RAR.
• Don’t ignore quote-side specifics. USDC vs USD1 vs USDT vs other wrapped stables carry different operational and governance profiles. That’s a direct input into risk_score and why SOL-USDT reads riskier than SOL-USDC counterparts.

Finally, size with turnover. The top RAR this week (SOL-USDC on Orca) posted $230.67M against $32.53M TVL; the DLMM variant did $22.58M against $3.20M TVL. Both are 7x turnover regimes that scale capital and still catch fees without perfect micromanagement.

Where to find and act on these faster

We keep a running stack of live picks on Best Solana pools and a macro view on Top Solana pools by TVL. If you’re building a cross-chain book and want context, the Cross-chain yield reference puts Solana’s fee regimes next to Ethereum and L2s. New ideas hit the Opportunities feed first, and if you want the mechanics behind the metrics, start with WealthVille Learn.

FAQ

How is RAR calculated, exactly?

RAR expresses the relationship between farmer_score (earnings quality) and risk_score (fragility), with smoothing to avoid division quirks at very low risk readings. Higher is better. Use it to rank and size, not as a binary yes/no.

Why does a pool with 31% APR rank below one with 0.6% APR?

Because the latter has much lower risk and more repeatable fee capture. SOL-USDC on Raydium shows 31% APR but carries an 8/100 risk score; SOL-USDC on Orca prints only 0.6% for the day but sits at 2/100 risk and far higher turnover-to-TVL. On a return-per-unit-of-risk basis, the Orca pool wins.

How often do farmer_score and risk_score update?

We refresh intraday as new fills, TVL moves, and route patterns land on-chain. Big structural changes (fee tier changes, program updates) can move scores immediately; smaller drifts tend to reflect in the next few data windows.

Does zero 24h fee APR make the RAR useless?

No. Farmer_score is not just today’s APR; it includes trend persistence and where volume tends to hit. A zero-fee day can still score if the structural read is solid, but we wouldn’t size those until fees reappear and turnover confirms.

How should I size positions using RAR?

Allocate more to RAR ≥20, keep satellites in 10–20, and limit 6–10 to short-dated, hands-on deployments. Rebalance when a pool moves by two RAR bands or when risk_score jumps three points.

What’s the difference between Orca Whirlpools, Raydium CLMM, and Meteora DLMM for LPs?

Whirlpools and CLMMs are tick-based concentrated liquidity; DLMM is bin-based with dynamic sizing. Fee accrual and range mechanics differ, which changes your realized fee profile and your tail risk. Read the Orca Whirlpools and Raydium CLMM docs if you’re moving size.