SaaS Valuation Compression

⚠️ **Trading warning / Avertissement trading** : informational only, not investment advice. A private round price is a negotiated mark at a moment in time, not a market quote: this analysis explains how a multiple moved, it does not tell you what a company is worth today or what it will raise at next, and private investments carry a real risk of loss, including total loss and illiquidity. No output here is a promise of return. Curated by Skill Harbor: an attribution analysis for one recurring private-market puzzle, a SaaS company whose valuation multiple compressed (or expanded) between funding rounds. The skill researches each round rather than answering from memory: round name, date, amount raised, post-money valuation and lead investor, ARR at or near each round (flagged as estimated when it has to be inferred from customer counts and average deal size), growth and retention context, and the narrative around the company, then computes the ARR multiple per round and decomposes the change correctly: valuation growth equals ARR growth times multiple change, additive only in log terms. Attribution follows a structured checklist, macro and rate environment (including the 2020 to 2021 zero-rate premium and the 2022 to 2023 hikes), growth deceleration and net retention, narrative shifts, an AI premium or discount, competitive pressure, and investor supply and demand, each rated primary, contributing or not applicable, benchmarked against dated private-market medians and known peer round pairs in its included reference file. Output is a visualization first (metric cards, multiple versus median over time, decomposition bars, peer comparison) plus a short prose verdict, primary cause and forward implication. From the himself65/finance-skills repository (MIT). Honest caveats: private ARR and valuations are often undisclosed, so parts of the model can be estimates with wide error bars, and the skill is told to flag its data confidence instead of hiding that; a down round, an acqui-hire or a single-round company each change the basis of the analysis and are handled as named edge cases; and median multiples by period are context, not a pricing formula. Skill Harbor never reviews the code, review it yourself before use.