Damage Math

Figures converted from Israeli shekels (₪) at historical FX rates — see data/company.json.fx_rates. NICE reports its financial results in US dollars, so revenue, EPS, FCF, and guidance figures are unchanged; only the shekel share-price and market-cap levels are converted (at ₪→$ 0.32641, applied uniformly so the −29.6% / −40.6% drawdown is preserved). Ratios, margins, and multiples are unitless and unchanged.

Damage Math — the value destroyed versus the price destroyed

NICE's shares fell 29.6% from their March 2026 peak (40.6% peak-to-trough) around a Q1 print that cut FY2026 non-GAAP EPS guidance to $10.98–$11.18, roughly 10% below FY2025's $12.30, while revenue guidance held at +8%. A transparent two-scenario NPV puts the value the near-term hit destroys at about 1.6% of pre-shock value if temporary and 10–18% if permanent. The price erased far more; the residual is a multiple de-rating. Independent judges put the probability the impairment is temporary at 0.61.

NICE reports results in US dollars but its shares trade in shekels on the Tel Aviv Stock Exchange (NASDAQ ADRs alongside). Price and market-cap figures below are shekel levels converted to dollars; revenue, EPS, FCF and guidance are in $ as the company reports them. The companion page shows the native shekel price levels.

The near-term hit — one line, one year

The trigger was the 6-May-2026 Q1 report. The quarter itself beat: revenue +9.8% to $768.6M, non-GAAP EPS $2.64 against a $2.52 consensus [1]. What repriced the stock was the outlook: management revised full-year non-GAAP EPS to a range of $10.98–$11.18 — a ~9.9% step-down from FY2025's $12.30 — while reiterating revenue at $3,170–$3,190M (+8.0% at the midpoint) and guiding Q2 revenue to just +5.5% [2].

The hit is an earnings and margin event, not a revenue decline. Revenue still grows ~8%; the cut lands on profit. FY2025 non-GAAP operating margin was 30.8% [3]; Q1 2026 printed 26.0% versus 30.5% a year earlier, with non-GAAP gross margin 68.4% versus 69.9% [4].

Sources: FY2026 guidance and Q1 margins, Q1 FY2026 results [5]; FY2025 EPS $12.30, Q4/FY2025 results [6]; consensus FCF, S&P Capital IQ estimates (data/sp/estimates.json, generated 23-Jul-2026).

On consensus, the magnitude is the same order: FY2026 revenue $3,178.6M (+7.9%), normalized EPS $11.11 (−9.7% vs $12.30), and free cash flow $589.8M against FY2025's actual $622.8M — a 5.3% dip in cash, not a collapse (data/sp/estimates.json; FY2025 FCF from the cash-flow statement, fit_features.adjusted_fcf).

Source: FY2025 actual EPS $12.30 [7]; FY2026 consensus normalized EPS, S&P Capital IQ estimates (data/sp/estimates.json).

Forward estimates tell against a permanent reading in one specific place: after the trough, the FY2027 line firmed rather than fell. FY2027 normalized EPS consensus rose from $12.49 ninety days ago to $13.13 now — above FY2025's $12.30 — and FY2027 revenue eased only 1.3% over six months (from $3,531.9M to $3,486.0M) (data/sp/estimates.json, momentum block). The Street is modelling FY2026 as a trough and FY2027 as a new high, with FCF climbing $590M → $695M → $767M → $829M through FY2029.

The price change over the same window

The price moved several times the earnings line. From the 5-March peak of $135.2 the shares fell to $80.4 on 19 June — 40.6% — and sit at $95.2, 29.6% below the high; the 6-May event leg alone was −18.8% in one session on a 3.3x volume spike (fit_features.capitulation_gauge; Dislocation traces the leg by leg). At 63.3M shares, that is roughly $8.5bn of equity value at the peak, $5.1bn at the trough, and $6.0bn now — about $2.5bn erased peak-to-current, $3.5bn peak-to-trough.

Source: TASE daily close converted at ₪→$ 0.32641 (data/prices/daily.json, company.json.fx_rates); shares 63.3M (fit_features.market_cap.shares); drawdown depth −40.56% (fit_features.capitulation_gauge.drawdown).

The cleanest way to see what actually changed is the multiple, which is unitless. On unchanged trailing FY2025 EPS of $12.30, the price went from about $135 at the peak to $95 now — from 11.0x to 7.7x trailing earnings. On the reduced forward number ($11.08 guidance midpoint), the multiple stands at ~12.2x at the peak, ~8.6x now, and ~7.3x at the trough.

Source: price = TASE ₪ close converted at ₪→$ 0.32641 (data/prices/daily.json, company.json.fx_rates); trailing EPS $12.30 [8]; forward EPS $11.08 [9].

On unchanged trailing FY2025 earnings, the 29.6% price fall is almost entirely multiple compression — 11.0x to 7.7x is a 30% de-rating. The ~10% cut to forward earnings is a separate hit to the "E"; on that lower forward base the multiple is 8.6x, still well below the 12.2x the peak implied. Either way the market has taken roughly three turns off the multiple, and the earnings cut accounts for about a tenth of the value change.

The NPV arithmetic — temporary versus permanent

The question the price poses is whether a ~10% one-year earnings step-down is a timing trough or a permanent level shift. The two readings have very different present values. The workings below are deliberately simple and reproducible: a 9% discount rate (a profitable, net-cash software compounder; sensitivity at 8% and 10% moves the totals about ±10%), a no-shock counterfactual of FY2025 FCF ($622.8M) growing 8% — the reiterated revenue rate — and the consensus FCF path as the realised "temporary" case.

Source: consensus FCF, S&P Capital IQ estimates (data/sp/estimates.json); no-shock counterfactual = FY2025 FCF $622.8M compounded 8%; temporary shortfall = counterfactual − consensus; permanent shortfall = 10% of counterfactual (a level shift that never re-converges). Derived; discount factors at 9%.

Temporary (a V). The shortfall is concentrated in FY2026 ($82.8M) and shrinks each year as the consensus path recovers to a new high — by FY2027 consensus FCF ($695M) already exceeds FY2025 ($623M). Discounting the four shortfalls at 9% gives roughly $128M of present value, and the stream is converging to zero. Call it ~$0.13bn of NPV destroyed — about 1.6% of the ~$8.5bn pre-shock equity value.

Permanent (a level shift). If instead the ~10% reset is the new base and never closes, the shortfall grows with the business in perpetuity. The explicit four-year present value is ~$243M; adding a terminal value on a 4% perpetuity lifts the total to roughly $1.5bn. A simpler cross-check — a permanent 10% haircut to the earnings the market capitalised at the peak (~$8.5bn) — gives ~$0.8bn. The permanent reading therefore destroys on the order of $0.8–1.5bn, or 10–18% of pre-shock value.

Source: price damage from TASE drawdown (fit_features.capitulation_gauge); NPV damage from the two-scenario workings above (permanent shown at the ~14% midpoint of the 10–18% range).

The gap

Set the two side by side. The market erased 29.6% of NICE's equity value peak-to-current (40.6% peak-to-trough). The plausible value damage from the near-term hit is ~1.6% if temporary and ~10–18% if permanent. Even the permanent reading leaves a gap of roughly 12–20 percentage points against the peak-to-current move, and ~22–30 points against the trough; the temporary reading leaves a gap near 28 points. In currency, plausible NPV damage of $0.13–1.5bn sits against ~$2.5bn of market value removed peak-to-current. The gap is not a rounding artefact — it is a multiple that fell from ~11.0x to ~7.7x trailing earnings, i.e. the market repricing NICE as if the impairment were permanent and then some. Whether that is warranted is the temporary-versus-permanent question, and that verdict is not this tab's to give.

The trial — both cases, then the ruling

The temporary-versus-permanent question was argued by two opposing, corpus-cited briefs and ruled on by three independent judges.

The temporary case, at its strongest. Revenue and operating income rose every year through the "disruption": revenue $2,181M → $2,945M FY2022–FY2025, segment operating income $435M → $546M → $646M into FY2025 [10]. The growth engine is still accelerating: Q1 2026 cloud revenue $603.4M (+14.6%), AI ARR +66% to $345M and in 100% of CXone enterprise deals, with AI already ~18% of cloud backlog — demand running ahead of recognised revenue [11]. The balance sheet is net cash ($304.1M, no debt) and buybacks accelerated $145M → $489M FY2022–FY2025, cutting the share count from 66.5M to 63.3M [12]. Consensus itself models FY2026 as the trough and FY2027 as a new high.

The permanent case, at its strongest. The affected engine is the company: Customer Engagement was 83.5% of FY2025 revenue ($2,460M of $2,945M) and its growth slowed to 7.8% from 15.6% the prior year [13]. Cloud growth decelerated from ~25% (FY2024) to ~13% (FY2025) to ~12% ex-Cognigy in Q1 2026, with cloud net revenue retention only 107% and management flagging "near term pressure on NRR" as the portfolio shifts to AI [14]. The pricing unit is changing from "Human-led Seat based" per-agent licensing to "AI-led Consumption-based" [15], and the company's own filing warns AI may cut "demand for solutions priced based on the number of human agents" [16] and drive "commoditization of our products and services" [17]. Cognigy, the offsetting AI asset, cost $887.3M, accounted for by the acquisition method with the purchase price allocated to acquired assets and liabilities at fair value [18].

The ruling. The judges put the probability the impairment is temporary at 0.61 (mean 0.60, range 0.58–0.62, spread 0.04); the result is not contested. There is a mild reading-order effect — judges who read the temporary brief first landed at 0.58, those who read the permanent brief first at 0.615 — but it does not open the spread. The diagnosis therefore leans temporary while carrying real weight (~39%) on permanent.

Source: trial ruling and per-judge scores (ruchir/trial/tally.json); the BigValue shows the panel p_temporary of 0.61.

This diagnosis is the report's, not a number to be rounded up: it is a lean, not a verdict, and it can flip. The judges named the conditions that would prove it permanent — ex-Cognigy cloud growth staying at or below high-single-digits with NRR falling below 107% for two consecutive quarters; the FY2026 operating margin settling at or under 25% with no H2 recovery; and FY2027 EPS/FCF failing to rebound toward FY2025 levels (ruchir/trial/tally.json, flip conditions).

Which line broke, and whether it self-corrects

The driver behind the hit is the seat-based economics of Customer Engagement cloud (data/sp/va.json isolates the cloud-revenue and margin lines; the filings name the mechanism). AI is compressing the priced unit — per-agent seats — faster than reported revenue reflects, and the transition carries higher AI-compute cost, lower-priced strategic renewals, and the Cognigy spend, which together reset the margin from ~30% toward 26% [19].

The mechanism by which it self-corrects — the temporary thesis — is refill: as seats compress, the committed AI spend is meant to replace them at higher value. The markers are consistent with refill running ahead of revenue — AI ARR +66%, 100% enterprise attach, AI ~18% of cloud backlog versus ~14% of cloud revenue, and management's expectation that this "translate[s] into improved expansion and NRR over time" [20]. The structural case against is that consumption pricing is a genuine price reset rather than a value-accretive migration — visible if ex-Cognigy cloud growth and NRR keep drifting down while AI ARR decelerates. The two are separable in the FY2026 prints, which is why the flip conditions above are the things to watch.