Arbitrage and Trading Opportunities Screener

1. What the board shows

For every Solana liquid-staking token (LST), the board answers three questions in one row — in the order you'll use them:

• How to enter cheaper — buy the LST on a DEX, or mint it via the protocol?
• How to exit cheaper — sell it on a DEX, or redeem it via the protocol for SOL?
• Is there an arbitrage — when the LST trades below the SOL value it redeems for, that discount, annualized, as an APY.

Every figure is denominated in SOL, never USD. Redemption settles at the epoch boundary (typically well under a day), not a multi-day validator queue. A second table on this board covers Solana yield-bearing stablecoins — a different mechanic, no epoch projection (see §7). Fair value is read live from on-chain state; nothing here comes from a third-party price feed.

2. Enter — the cheaper way in

To enter you can buy the LST on the open market, or mint it by staking SOL through the protocol. The board quotes both at a standard notional and marks the winner:

• BUY — a DEX aggregator quote (slippage and routing included) gives more LST per SOL than minting does.
• MINT — staking via the protocol beats the market. The rarer case, usually a liquidity gap or a temporary dislocation.

A DEX buy wins on entry the large majority of the time, so this column is usually BUY. The value is the confirmation, and the occasional flip to MINT.

3. Exit — the cheaper way out

To exit you can sell on the market, or redeem through the protocol for SOL. The board compares them — but on Solana the redeem side has a twist (§4): the payout is projected to the epoch boundary, because staking rewards keep accruing while you wait.

• REDEEM — the protocol returns more SOL than selling on a DEX would, once projected over the epoch wait. A real discount to capture.
• SELL — the market pays above the projected redemption value, so selling beats redeeming and skips the epoch wait entirely.

4. The arbitrage — redemption spread as APY (spot vs. projected)

When an LST trades below the SOL value it redeems for, buying it and redeeming it captures the gap. Protocols offer either an instant unstake (capped by reserve liquidity) or a delayed unstake — withdraw a stake account and wait out the epoch. We model the delayed path, since that's where the fair value lives.

Three steps:

1. Buy S with SOL on the market. A DEX aggregator quote for a fixed size Q (default 100 SOL) returns A, slippage and routing baked in. Aggregator is selectable (Jupiter / Kamino).
2. Read the on-chain redemption rate. The pool's fair value, straight from on-chain state (no third-party API): total_lamports / pool_token_supply net of the withdrawal fee (including the pending next_stake_withdrawal_fee when one activates on the epoch we're waiting for), or Marinade's msol_price (Q32.32) minus its live fee for mSOL.
3. Project to redemption. Carry that spot rate forward over the epoch wait (below), then compare to Q.

Spot vs. projected. The on-chain rate is a spot reading — today's rate. But redemption settles at the epoch boundary, and the pool keeps earning staking rewards the whole time you wait, so spot alone understates the payout. The table shows both:

R_spot = fair_redeem_out                                   # today's on-chain rate
R      = R_spot × (1 + staking_apy × redeem_days / 365)    # projected to redemption

spread = R / Q − 1
APY    = spread × 365 / max(redeem_days, FLOOR_DAYS)

• Spot fair redeem / Spot APY % — today's rate, no projection: the honest figure for an instant fill.
• Projected fair redeem / Projected APY % — spot carried forward over redeem_days. The board sorts on the projected value, since that's the trade that's actually there. It's an approximation — hence "projected," not "guaranteed."
• redeem_days — the remainder of the current epoch, read live; network-wide, so near-identical across rows.
• Reserve (SOL) — the pool's current instant-unstake buffer, shown for reference only. It is not used in any spread/APY figure and is not a guarantee of instant redemption; the projected figures assume you waited and earned rewards.

Worked example. JitoSOL / SOL, 100 SOL quote, redeem_days = 0.73:

spot:       R = 99.996313 SOL  →  spread −0.0037%  →  Spot APY −1.84 %
projected:  R = 100.005641 SOL →  spread +0.006%   →  Projected APY +2.82 %

Spot reads as a thin loss; projected — adding the staking rewards earned across the epoch wait — turns it into a thin positive spread. The board sorts on the projected number, so it surfaces the trade that's actually there, not the spot mirage that reads "loss." When the projected spread is negative, the market pays above projected NAV — selling beats redeeming, and you skip the epoch wait.

5. Reading the columns

Every row is on Solana mainnet — a single-network board with no chain chips; filter with the search box and min-spread slider.

6. Pro

The board is free to read: every spread, APY, and the Enter/Exit direction (BUY / MINT, SELL / REDEEM). Pro turns the direction into an action:

• Enter / Exit magnitude — exactly how much each door wins by.
• Quote at your size — enter, exit and buy & redeem priced at your own position size, with real slippage rather than a fixed notional. Because DEX slippage is nonlinear, the cheaper door can flip at your size.
• Direct mint / redeem links on the winning side.
• Telegram alerts — a threshold alert when a spread crosses your APY line, and a flip alert when the cheaper exit changes REDEEM ↔ SELL.

7. YBS Enter / Exit (Solana yield-bearing stablecoins)

The second table on this board applies the same enter/exit and buy-then-redeem mechanic as above, but for yield-bearing stablecoins rather than LSTs — so no spot/projected split and no reserve column (those are specific to SOL-staking rewards, which don't apply here). Fair price is read live on-chain; Fair redeem / APY % are the plain figures.

Five pairs:

• eUSX (Solstice) — bought with USX; yield from Solstice's delta-neutral strategy. No redemption fee. Redemption window depends on size, so two tiers: Fast (999 USX, ~1 day) and Standard (10,000 USX, ~7 days).
• PRIME / AUTO (Hastra) — bought with USDC, redeemed via wYLDS (two different Hastra programs). Leg 1 (→ wYLDS) is instant; leg 2 (wYLDS → USDC, 1:1, no fee) settles in an app-stated ~1–2 business days. Redeem, days shows a 2.0 max estimate; a ? badge explains the two legs.
• JUICED (Jupiter Lend) — bought with jupUSD; instant redemption, no fee. Fair price reads the documented token_exchange_price field (open-source IDL/SDK), the raw stored rate — same "trust the last on-chain write, don't extrapolate" choice we make for mSOL.
• PST (Huma Finance) — bought and redeemed in USDC; Huma's "Classic Mode" receipt token. Fair price reads the pool's PoolState account, specifically Classic Mode's live assets field (principal + realized/unrealized yield, not a single token account's balance) — the layout and mainnet addresses come from Huma's own published npm package (@huma-finance/shared, which ships their Solana Anchor IDL directly, since Huma's program has no public GitHub repo). An earlier version of this row read a single USDC reserve token account instead, which turned out to hold only a working-capital buffer — about 1/285th of the real NAV — since Huma deploys most of Classic Mode's TVL elsewhere rather than holding it idle.

Fair price for eUSX/PRIME/AUTO is a ratio of two standard SPL reads (reserve balance / staked-token supply), each rate cross-checked against real on-chain transactions rather than an assumed byte layout; PST's is a ratio read from Huma's own published account layout instead (see above).

8. A living, growing list — request a token

The tracked set isn't fixed — we add tokens and keep addresses, fees, and connector logic current as protocols change. Currently: bSOL · pSOL · JitoSOL · hSOL · JupSOL · mSOL · vSOL · BonkSOL · jSOL · hyloSOL · dfdvSOL, and growing weekly. (dfdvSOL and hyloSOL run on Sanctum Single- and Multi-Validator deployments; we byte-verified their layout matches the standard pool, so they use the same connector.)

Not listed yet: sSOL (Solayer restaking — carries AVS exposure that doesn't fit a simple Delayed-Unstake fair value) and INF (Sanctum Infinity basket — need to confirm a true Delayed-Unstake path vs. an AMM swap). Both are to-revisit, not rejected.

If an LST you follow isn't listed, ask — we'll add it once it exposes an on-chain NAV, a documented delayed-unstake path, and enough liquidity to quote. Reach us on X @ArbScreenerDeFi or at arbscreenerdefi@gmail.com.

9. Risks & disclaimer (NFA)

Read-only. ArbScreener never signs, routes, or custodies funds — it reads on-chain values and shows the spread; every action stays with you.

APY is a gross estimate. It ignores network fees; NAV and the withdrawal fee can move during the wait; the epoch projection is an approximation; the quote is a snapshot; and LSTs carry validator and smart-contract risk a spread can't capture. The numbers answer "is this worth digging into," not "here's your exact net yield."

This is Not Financial Advice (NFA).

Arbitrage and Trading Opportunities Screener is provided "as is", for informational and educational purposes only. The app does not provide investment, financial, legal, or tax advice and is not an offer, solicitation, or recommendation to buy, sell, or hold any asset.

All metrics (spread, APY, fair value, etc.) are approximate estimates based on third-party quotes and on-chain data that may be incomplete, stale, or wrong. The gross calculation does not account for gas, bridge fees, slippage beyond the quote, NAV changes, epoch-projection error, validator or smart-contract risk, or depeg risk. Past or projected returns do not guarantee future results.

Contract addresses and protocol parameters must be verified independently before any action; some addresses in the config are marked as needing re-verification. A wrong address can lead to loss of funds.

You use the app and its results at your own risk and are solely responsible for your own decisions and actions. The authors and contributors accept no liability for any direct or indirect losses, lost profits, or other damages arising from use of the app. Interacting with DeFi protocols carries high risk, up to and including total loss of funds.

Before taking any action, do your own research (DYOR) and, if needed, consult an independent licensed professional.