Why it matters
Why this tool is essential
Six findings from live measurement — not theory.
The printed price is not the exit price
STONK shows $215M on-chain market cap. A $1M cash exit recovers only 85%. Printed valuation and executable exit capacity are different numbers — and the gap can be enormous.
The metric everyone publishes is the wrong one
Pool TVL ÷ market cap correlates at ρ = −0.66 with measured exit cost. Absolute pool TVL reaches ρ = −0.92. Normalising by valuation actively destroys signal. SPYx ranks worst by TVL/mcap ratio (5.3%) yet executes $100k at 9 basis points.
Cliffs are invisible to every existing dashboard
STRCx executes cleanly at $150k, then returns $0.04 per token at $200k because the router switches paths. TVL, volume, and price are unchanged. Only a size-dependent quote probe reveals this.
Hard walls matter more than big slippage numbers
6 of 18 probed assets return no route at $250k. FTR fails above $25k. A no-route ceiling is a harder capacity limit than any slippage percentage.
You may not hold the asset you think you do
Searching for HOODx returns ~20 tokens. One embeds the genuine mint address inside its own name field to defeat string matching. Mint-level authentication is a prerequisite, not a feature.
Quoted depth is not committed depth
Much of the good depth on majors comes from RFQ venues (Riptide, Byreal, AlphaQ) that can withdraw precisely under stress. The scanner tags RFQ share separately instead of merging it into one depth number.
Exit-cost depth surface (18-asset sweep)
Exit cost at $100k notional (%, absolute)
TVL/mcap ratio vs measured exit cost
STRCx routing cliff
Measured 13 Sep 2026, US equity markets closed · Jupiter router quotes