Why this is in the vault
CJ uses Prospect Theory (Kahneman/Tversky, via a Jay Ritter podcast interview) to explain why founders and boards don't get angry about Day 1 IPO pops even though 21% of all IPO capital raised since 1980 ($250B) gets "left on the table" — worth keeping as a clean applied-behavioral-economics case study, plus the recurring Koyfin valuation/efficiency benchmark data.
The core argument
Money left on the table = shares sold × (Day 1 close − IPO price); Ritter's dataset puts the long-run average at ~21 cents per $1 raised, and CJ's own 15-company recent-IPO sample (ex-SpaceX) comes in at ~49 cents per $1. The behavioral explanation: founders anchor on early banker valuation estimates during the ~6-month IPO process, so each subsequent upward revision (even one still below fair value) registers as a windfall, not a loss — classic Prospect Theory reference-point framing. CJ also steelmans the other side: oversubscription insurance, imperfect price discovery, Day 1 illiquidity (lockups), and the fact that founders aren't selling the whole company all mitigate the "incompetence" read. Bottom line: some underpricing is rational, but Figma pricing at $33 and closing at $115 is hard to wave away as pure discount.
Curation section
- Run the Numbers podcast: full episode with Jay Ritter ("Mr. IPO," 40+ years of IPO data) on Apple/Spotify/YouTube — the source interview for the main piece.
- Weekly Valuation & Efficiency Metrics (Koyfin-sourced): revenue multiples (NTM EV/Revenue) and efficiency benchmarks (CAC Payback, Revenue per Employee, Rule of 40) across a 9-sector, ~130-company tracked universe (Security, Data/AI Infra, Dev Tools, Horizontal SaaS, GTM/MarTech, Vertical SaaS, Take-Rate Platforms, Payments, Consumer Fintech/Crypto).
- OPEX breakdown: COGS/S&M/R&D/G&A as % of revenue across the same universe, GAAP basis (includes stock comp).
Mapping against Ray Data Co
The mapping is weak on the literal subject — RDCO has no IPO path, and this issue doesn't touch AI services, agent economics, or SMB acquisition, the three lanes that usually connect Mostly Metrics content to RDCO's actual work. The one real thread worth naming: Prospect Theory's anchoring mechanism (people evaluate outcomes relative to a moving reference point, not an absolute) is directly relevant to the founder's own "accumulate→enjoy inflection" tension already tracked in the vault (user_money_values_potential_tension memory) — the same psychological mechanism that makes a founder feel rich off a below-market IPO print is the one worth watching for in his own read of RDCO's growth numbers or any future liquidity event. That's a personal-decision-psychology note, not a strategy one — don't force a business-model connection that isn't there.
⚠️ Sponsorship
Three disclosed relationships in this issue: (1) Abacum — sponsor ad + survey/webinar plug, standard recurring sponsor for this newsletter (also seen in prior mostlymetrics notes); (2) Mostly Talent — CJ's own recruiting arm, self-promotional, not a third-party sponsor, flagged as house-promo; (3) Koyfin — recurring data-source credit plus an explicit affiliate link ("Please check out our data partner, Koyfin... It's dope"), so the weekly valuation charts are sourced from a paying affiliate partner. None of the three bias the core IPO/Prospect Theory argument, which cites independent academic data (Ritter/Loughran) rather than sponsor material.
Related
[[2026-09-06-mostlymetrics-oura-ipo-s1-breakdown]] [[2026-05-21-mostlymetrics-spacex-ipo-s1-breakdown]] [[2026-05-12-mostlymetrics-cerebras-ipo-s1-breakdown]] [[user_money_values_potential_tension]]