Damage Math

Damage Math

Salesforce is a damage-math anomaly: the price fell about 54% from its peak while the near-term cash engine grew. Free cash flow rose from $9.5B (FY2024) to $14.4B (FY2026) and forward consensus held flat to higher through the drawdown. So the usual numerator — a near-term earnings cut — is roughly zero. The de-rating instead reprices the durable growth rate. Whether that repricing is an overshoot or correct is the trial's contested question, ruled 0.62-probable-temporary.

The near-term hit, quantified

The framework's usual trigger is a short-term earnings cut the market then anchors to. Here the numerator barely exists. Across the window that carries the drawdown — the December 2024 peak to today — Salesforce's reported cash flow rose and forward consensus did not fall.

Free cash flow compounded straight through the decline. Operating cash flow was $10.2B in FY2024, $13.1B in FY2025 and $15.0B in FY2026 [1]; on the profile's feed that nets to free cash flow of $9.5B, $12.4B and $14.4B — a 16% rise in the last year and 52% over two years, entirely inside the price fall (fit_features.adjusted_fcf).

Source: FY2024–FY2026 actuals derived from company filings (operating cash flow less capex), FY2026 10-K [2]; FY2027e–FY2029e are S and P Capital IQ consensus, data/sp/estimates.json (vendor vintage 2026-07-21). Feature: fit_features.consensus_forward_yield.

Forward consensus tells the same story. Over the roughly 180 days spanning the June 2026 trough, FY2027 revenue consensus held near $46.1B and FY2027 normalized EPS rose from $13.20 to $14.13 (+7.1%); FY2028 revenue held near $50.5B and FY2028 EPS edged from $15.14 to $15.51. Estimates did not impair; several inched up.

Source: S and P Capital IQ consensus revision snapshots, data/sp/estimates.json (vendor vintage 2026-07-21). Snapshots dated 2026-01-22 (180d) and 2026-07-21 (now).

The prints that landed after the fear peaked confirmed earning power rather than eroding it: normalized EPS beat consensus by 13.6% in Q3 FY2026, 24.9% in Q4 FY2026 and 24.1% in Q1 FY2027 ($3.88 vs $3.13), on S and P Capital IQ consensus (data/sp/estimates.json, vendor vintage 2026-07-21). On the framework's own numerator, then, the near-term hit is not negative — it is slightly positive.

The price and value moves, side by side

Set that against the price. Salesforce peaked at $367.87 on December 4, 2024 and trades at $170.77 today, with a $150.12 intraday trough on June 22, 2026 — a 59.2% peak-to-trough fall and 53.6% peak-to-today (fit_features.capitulation_gauge.drawdown). On ~975M shares at the peak and 956M today, market capitalization fell from roughly $359B to $163.3B — about $196B, or 54.6%, of equity value erased (fit_features.market_cap).

Source: EPS and FCF from data/sp/estimates.json and company filings [3]; price and market cap from fit_features.capitulation_gauge and fit_features.market_cap.

Enterprise value moves with it. The profile's feature carries net debt of $7.1B, putting current EV near $170.4B (fit_features.balance_sheet_class); on a marketable-securities-inclusive basis consensus actually shows a small net-cash position (FY2026 net debt −$4.6B in data/sp/estimates.json), so Salesforce is roughly balance-sheet-neutral and EV tracks market cap within a few billion. Either way the arithmetic is the same: consensus FY27 EPS rose ~7% and FCF rose 16%, while the market cap fell ~55%. That divergence is the whole subject of this tab.

The NPV arithmetic

Because there is no near-term cash hit, the standard "N years of reduced cash, then recovery" model returns essentially zero NPV damage — the temporary-impairment floor here is that the problem destroyed no measurable cash flow at all. The de-rating is therefore a repricing of the durable growth rate and terminal value, not of near-term earnings. Two transparent tools bound it.

Assumptions, stated. Discount rate r = 9% (cost of equity for a mega-cap software franchise; sensitivity shown). Base free cash flow = FY2026 actual $14.4B. Ten years of explicit FCF growth plus a Gordon terminal value, discounted to equity value, compared to today's $163.3B market cap and the ~$359B peak. Salesforce's FCF is levered (post-interest) and net debt is immaterial, so the discounted stream is compared to equity value directly.

Tool one — a two-scenario DCF-lite. The scenarios differ only in the durable growth path.

Source: derived; ten-year explicit FCF projection off the FY2026 $14.4B base plus a Gordon terminal value, discounted at the stated rate. Reproducible from the assumptions above and the base FCF in fit_features.adjusted_fcf.

The current $170.77 sits between the two pessimistic rows — it requires FCF growth to fade toward flat-to-declining, or an AI-risk discount rate of 10–11%, before plausible NPV matches the price. A temporary re-acceleration values the equity near $432 a share; even a moderate permanent maturation to mid-single-digit growth values it near $278. The price is consistent only with the bearish end of the permanent reading.

Tool two — reverse the DCF. For a stable grower, price/FCF ≈ 1 / (r − g), so today's price implies a single perpetual FCF growth rate the reader can judge directly.

Source: derived; implied g = r − (FY2026 FCF $14.4B / market cap), with current cap $163.3B and peak cap ~$359B on $12.4B trailing FCF. Reproducible from fit_features.market_cap and fit_features.adjusted_fcf.

At a 9% discount rate, today's price embeds roughly zero perpetual FCF growth; the peak embedded about 5.5%. The de-rating has stripped out five points of assumed long-run growth from a business whose cash flow is currently compounding double digits and whose consensus FCF keeps rising — a forward FCF yield of 9.1% (FY27), 10.0% (FY28) and 10.9% (FY29) on today's cap (fit_features.consensus_forward_yield), cross-referenced in the Yield tab.

The gap. The price destroyed about $196B (−54.6%) of market cap. The near-term cash flow destroyed nothing — it grew. So under a temporary reading the NPV damage is roughly zero and the gap is nearly the entire drawdown; under a moderate-permanent reading the equity is worth ~$265B, a ~$94B (26%) haircut from the peak, still leaving ~$100B of unexplained price damage versus today's $163B. The gap only closes — price damage roughly equals plausible NPV damage — at the bearish-permanent end: terminal FCF growth near zero-to-negative, or a 10–11% AI-risk discount rate. Which of those readings holds is not this tab's to decide.

The trial, presented fairly

The temporary-versus-permanent question was argued by two opposing, corpus-cited briefs and ruled on by three blind judges. Both cases are strong.

The case for temporary. The impairment is essentially pure multiple compression: at the peak Salesforce traded ~29x trailing FCF, today ~11x, while FCF grew $12.4B → $14.4B. Consensus never impaired the NPV — FY2027 revenue held flat and EPS rose; forward FCF is modeled to keep rising to $17.8B by FY2029. Q4 FY2026 revenue re-accelerated to +12% year over year and Q1 FY2027 EPS beat by 24% [4]. Management raised its FY2030 revenue target to $63B and, calling the tape "obviously not a rational market," lifted the buyback authorization to $50 billion after returning 99% of FY2026 free cash flow [5]. Cannibalization would show as falling ARR and FCF; both are rising.

The case for permanent. The old 20%+ compounding engine has reset to high-single digits, and the NPV of a 9% grower is structurally below that of a 25% grower even with cash flow intact. Revenue growth fell from five straight ~25% years to 18.3%, 11.2%, 8.7% and 9.6% in FY2023–FY2026 (fit_features.revenue_trajectory). The core is mature: in FY2026 Agentforce Sales grew 8%, Service 8% and Marketing and Commerce just 3% [6]. cRPO growth was "somewhat disappointing" at 9% [7]. And the latest acceleration is partly bought: Q1 FY2027 revenue growth was "bolstered by the acquisition of Informatica" and roughly two points of FX, with Agentforce Apps up only 9% against Data 360 up 25% [8][9]. A reset terminal growth rate lowers the NPV permanently, cash flow notwithstanding.

The ruling. The judges put the probability the impairment is temporary at 0.62, with individual seats at 0.62, 0.42 and 0.68 — a range of 0.42 to 0.68 and a spread of 0.26. The ruling is formally contested: reading order moved the mean only 0.07, so the disagreement is substantive, not an artifact. Read plainly, the panel leans temporary but does not settle it — the permanent case commands a real minority. This tab reports that number; it does not override it. The moat evidence behind the durability question is developed in Durability.

Which line broke, and whether it self-corrects

The driver behind the hit is not a cash shortfall but a growth-rate reset concentrated in the mature core. Subscription revenue grew 10% in FY2026, held back by Marketing and Commerce at 3% and the Sales and Service clouds at 8% each [10]; the forward-looking cRPO line grew 9% [11]. Consecutive-decline years on the framework's own test are zero (fit_features.revenue_trajectory), so this is deceleration off a high base, not decline — but the level has reset.

The self-correction mechanism, if the reading is temporary, runs through two channels. First, monetization of the same AI shift that is feared: in Q1 FY2027 the newly disclosed Data 360, Headless Platform and Other category grew 25%, versus 9% for Agentforce Apps [12]. Second, the buyback flywheel: Salesforce returned more than $14B in FY2026 — 99% of free cash flow — and authorized $50B more at what management called low prices [13], retiring float from 974M (FY2022) to 956M (FY2026) and lifting per-share value regardless of when sentiment turns (fit_features.share_count_trend); the capital-return record is examined in Self-Help.

The mechanism against self-correction is that the growth engine now leans on acquisitions and FX rather than clean organic repair — the Informatica contribution and two points of currency in Q1 FY2027 [14] — and that agentic, consumption-priced AI could displace seat-based subscription before it fully replaces it. If organic growth ex-Informatica and ex-FX proves sub-8% and still decelerating, or FY2027 FCF falls below $14.4B, the permanent reading gains the arithmetic. Those are the lines to watch; the trial's flip conditions name them precisely.