Devnet only · no mainnet TGE · no buy button

Paper deal · 0002

NO — price · Saber AMM

Live-NO. TVL is LPs’, not ours. Counterfactual ask $25k, cap ~$7k. Buying the float is a trade, not a cell.

Paper deal 0002 — PASS

Target: Saber AMM (SBR) — Solana DEX for pegged assets (USDC/USDT, SOL/mSOL, …)
Source: DefiLlama / Saber
Memo date: 2026-09-19
Cell type: protocol (hybrid with the SBR token)
Outcome: PASS. The ask is counterfactual ($25k for upgrade authority). The cap on transformed earnings is ~$7k. Buying the float at $217k mcap is a trade, not a cell.

Exercise. No LOI, no listing. On-chain facts are locked. The handover price is labeled counterfactual.


1. Public facts (locked)

Item Value Note
TVL $5.53m Users’ LP. Not ours, does not go into the treasury.
Fees 30d $513
Revenue 30d $266 Half of fees, retained by the protocol.
Holders revenue 30d $0 SBR takes no cash.
Volume 30d $7.05m Still alive, small.
Cumulative volume $6.56b Historical PMF, not a pitch.
SBR mcap $216,912 $0.000082 vs ATH $0.96
Ask $25k Counterfactual: upgrade-authority handover. Not a listing.
AI disruption low AMM for pegged assets. AI does not eat USDC/USDT swap.

Run-rate revenue ≈ $266 × 12 = $3.2k/year. We underwrite this, not the $6.56b cumulative volume.


2. Screen (constitution §3)

Filter Outcome
Proven PMF Pass. Cumulative volume and residual TVL.
Room to operate Pass. Holders revenue $0, fee take, UX, Jupiter routing.
Predictable earnings Low pass. On-chain fees, small, persistent.
AI risk after integration Pass. Category is not an “AI wrapper.”
Not pre-PMF Pass.

No hard fail. We go to price. (Different from 0001, where AI closed the screen.)

Extra constraints, locked:

  • TVL $5.53m is user capital. Constitution / RAILS: we do not LP dry powder into the target, we do not “socialize” a hole.
  • Buying SBR at mcap for “control” without upgrade authority is a position, not a cell (PROTOCOL.md).

3. Transformed earnings

90 days, if the keys arrived (they do not):

  • Upgrade authority on Squads.
  • Fee switch: holders revenue today $0 → an honest on-chain share, not Harvest.
  • Clean client/IDL, Jupiter listing if missing.
  • Steward = agent. Low opex.
  • Dry powder does not provide Saber liquidity (conflict + IL).
P Year-1 revenue Cost Earnings
Down 0.35 $1.0k (volume dies) $0.6k $0.4k
Base 0.50 $3.2k (run-rate) $1.0k $2.2k
Up 0.15 $8.0k (fee+UX) $1.5k $6.5k

Expected ≈ 0.35*0.4 + 0.5*2.2 + 0.15*6.5 = $2,215.

Haircut 20%: uncertain keys, old program, TVL that can leave the day of the announcement.


4. Price

Hurdle 4×, hold forever, no leverage.

  • 4 × $2,215 = $8,860
  • after 20% haircut → cap $7,088

Counterfactual ask $25k > cap. Mcap $217k ≫ cap.

No offer. One offer, little movement: we do not walk up toward $25k and we do not buy SBR “because it is cheap vs ATH.” ATH is a fossil.

If a founder arrived with the keys under cap (~$7k), the memo reopens. Today no.

Governance: even the counterfactual ask is ≥ $25k → it would be Realms. Irrelevant: PASS before the vote.


5. Solana rails (why this deal exists)

Saber is a Solana financial instrument (pegged AMM). For us it is a target, not a venue. Treasury venues stay: USDC, Kamino/Save (cash), Jupiter (spot/buyback). Putting dry powder into Saber pools while we underwrite it is IL + conflict. RAILS.md.


6. Curriculum

  • A neglected protocol with real PMF and real P&L can pass the screen and fail the price. That is the healthy case.
  • Token mcap ≠ protocol price. Without upgrade authority, $SBR is a chart.
  • Holders revenue $0 is room to operate, not an invitation to gouge.