The Core Numbers
Infrastructure Gap Figures
836,120×ICP avg TPS vs. Stripe 1B TPS target
4,768×ICP theoretical max vs. Stripe 1B TPS
83×Bear Case floor — minimum gap regardless of scenario
$650BTotal ACSM TAM across 5 layers by 2030
23,611EPS baseline — 1M agents × 1 wf/hr × 85 micro-events
ACSM Framework · Five Layers
$650B Est. TAM · Composability is the moat
L1 $180B Agent Execution — Stripe ACP, OpenAI. Winner-take-most. Now.
L2 $95B Off-Chain Mesh — State channels. Oligopolistic. 2025–26.
L3 $120B Rollup/Netting — ZK rollups. Stripe Tempo. Duopolistic.
L4 $85B Risk/Compliance — Smart contract AML/KYC. GENIUS Act + MiCA.
L5 $170B Finality Anchor — ICP/Solana/ETH. 2–3 dominant by 2030.
Three Scenarios · 2030
Bear Case Floor Has Permanently Moved
55%Base Case1M+ TPS · Ordered transition
30%Bull Case1B+ TPS · Hypergrowth break
15%Bear Case100K+ TPS · Still 83× above today
Original Frameworks · iProDecisions Research
Three Original Frameworks
ACSM Five composable infrastructure layers. $650B total TAM. Composability is the moat.
ACML Five maturity stages: Evaluate → Pilot → Deploy → Scale → Native. Stage 3 = 2025–27 window.
EED Economic Event Density. Baseline: 23,611 EPS. Stage 4: 23.6M EPS.
Verbatim Quotes · Stripe Annual Letter 2025
Patrick & John Collison
"Agents will most likely soon be responsible for most internet transactions."
"Meeting AI demands will likely require a horizontal architecture of multiple, interacting chains."
"1M or even 1B TPS"
Stripe Annual Letter 2025 · Feb 25, 2026
Strategic Window · Time-Critical
Act in 2025–2027
L1/L2 decisions now are path-dependent for 10–15 years
EU AI Act August 2026 — L4 compliance forcing function
GENIUS Act (July 2025) + MiCA (Dec 2024) — legal framework live now
L4 compliance retrofit at Stage 4 costs 3–5× vs. building at Stage 3
Bear Case still demands 83× infrastructure — pessimism requires re-architecture
Read Full Report →Book Advisory