UiPath, Inc.

Module 2 of 5

2 · Cost model

Operating costs include cost of revenue plus operating expenses. Their growth responds to simulated revenue growth, recent cost trends and the quarter of the year. Adjusted costs equal revenue less company-reported adjusted operating income.

Open full-size chart. Last 12 reported quarters are plotted; all 19 aligned quarters appear below. The vertical line separates reported history from model predictions. Historical model points use earlier quarters only. Shaded ranges are conditional simulations and have not been calibrated to real-world probabilities.

Next target: 2026-10-31

Four-quarter predictions

Module 3 of 5

3 · Earnings model

Each path subtracts operating costs from revenue, then adds the modeled bridge for other income, taxes and common-stockholder deductions. The result is earnings available to common stockholders—the numerator required for EPS.

Adjusted earnings exclude stock compensation, acquired-intangible amortization, equity-related payroll taxes, restructuring and applicable other reconciliation items, with estimated tax effects. No announced standardized forward tax rate was located. FY2026 valuation-allowance releases boosted GAAP earnings; the historical bridge downweights exceptional results. Current taxes and investment income remain important sources of forecast error. Company reconciliation and tax policy.

Open full-size chart. Last 12 reported quarters are plotted; all 19 aligned quarters appear below. The vertical line separates reported history from model predictions. Historical model points use earlier quarters only. Shaded ranges are conditional simulations and have not been calibrated to real-world probabilities.

Next target: 2026-10-31

Four-quarter predictions

Module 4 of 5

4 · Share count model

Weighted-average shares follow a recent trend with uncertainty; potential dilution is modeled separately. Profitable paths use diluted shares. Loss paths use basic shares to avoid treating anti-dilutive securities as EPS dilution. Buybacks, stock compensation and issuance enter through reported share history.

Open full-size chart. Last 12 reported quarters are plotted; all 19 aligned quarters appear below. The vertical line separates reported history from model predictions. Historical model points use earlier quarters only. Shaded ranges are conditional simulations and have not been calibrated to real-world probabilities. The chart shows basic and adjusted diluted shares; the tables also show GAAP diluted shares.

Next target: 2026-10-31

Four-quarter predictions

Module 5 of 5

5 · EPS model

For every simulated path, EPS equals common earnings divided by the matching weighted-average share count. Revenue, costs, below-operating-income items and shares therefore feed the same EPS distribution. GAAP and adjusted measures remain separate.

Open full-size chart. Last 12 reported quarters are plotted; all 19 aligned quarters appear below. The vertical line separates reported history from model predictions. Historical model points use earlier quarters only. Shaded ranges are conditional simulations and have not been calibrated to real-world probabilities.

Next target: 2026-10-31

Adjusted basis is inferred: the vendor’s prior-quarter $0.15 matches company adjusted EPS, versus GAAP $0.07. Comparisons depend on that basis match. Analyst source.

The vendor reports an analyst EPS range of $0.16 to $0.22 and a revenue range of $442.4M to $444.5M. These describe dispersion among estimates, not probability intervals.

How operating profit and current taxes change the comparison

The historical fit blends tax environments and cannot predict future valuation-allowance releases. Compare its EPS result with the explicit current-tax sensitivity before interpreting a possible surprise.

The last row assumes $100M adjusted operating profit, $17.116M other income (the mean of the two latest quarters), a 26.94% effective adjusted tax rate derived from those quarters, and 522.8M shares from the share-model median. The range changes only the tax rate from 30% to 20%; it is not an 80% prediction interval. Other income includes volatile items and may change. This is a conditional sensitivity, not company EPS guidance or an independently validated forecast. Management’s outlook. The company guides revenue and adjusted operating profit; the EPS sensitivity is calculated here.

Conditional surprise scenarios

These percentages are model scenarios, not calibrated odds. Adjusted EPS: model MAE $0.036, last-quarter baseline MAE $0.091; the nominal 80% range covered 7/7 outcomes. These are reconstructed checks, not archived live forecasts. The historical model estimates $0.24 adjusted EPS versus $0.19 consensus; setting revenue to consensus gives $0.22. Guidance operating profit combined with recent tax economics gives $0.16. The direction changes with these assumptions, so the historical-model gap does not establish a reliable surprise edge.

What would erase the modeled upside?

At consensus EPS, the median break-even adjusted operating-cost level across the simulated revenue, earnings-bridge and share paths is $359.6M, versus a cost-model median of $333.2M. The median cost headroom is $25.3M. Replacing next-quarter revenue with analyst consensus and rerunning the linked cost response produces a conditional adjusted EPS median of $0.22.

Four-quarter predictions

Analysts, realized surprises and market context

Last closing share price: $12.98 on 2026-11-09. Historical market prices.

ApexVol reports an options-implied stock move of ±8.9% for the December 3 report, dated November 9, 2026. This is a vendor indication; the public page does not expose the selected contracts, expiry or raw quotes needed to rebuild it. It measures magnitude, not direction. Market-implied EPS is not identifiable from this move. No directional market EPS forecast is presented. An earnings beat can accompany a falling share price when guidance, valuation or other news dominates.

Historical analyst estimates, retrieved 2026-11-10; reported actuals checked against linked company releases. Returns use the release-date close and the next trading close, exclude dividends, and do not isolate the earnings effect. Historical expectations can be revised by the vendor. Six historical vendor rows passed the actual-EPS and revenue checks. The December 2024 vendor row failed the primary-source adjusted-EPS check and was excluded; the January 2025 comparison was unavailable from the same vendor. Intraday November 10 prices are excluded from the closing-price comparison.

Does the linked model improve predictions?

Adjusted EPS: model MAE $0.036, last-quarter baseline MAE $0.091; the nominal 80% range covered 7/7 outcomes. These are reconstructed checks, not archived live forecasts. Compare mean absolute error (MAE) with carrying forward the last reported value and with the same quarter a year earlier. Lower is better. The model is not consistently superior across measures and horizons.

Model and source notes

The revenue module supplies 6,000 simulated paths. Accounting equations fit 19 aligned quarters with recency weighting (four-quarter half-life), regularization and robust residuals. The accounting layer uses approximate Gaussian parameter uncertainty and correlated Student-t shocks; it is not a second fully Bayesian fit.

Cost growth responds to revenue growth and seasonality. Common earnings are linked to operating income through an estimated other-income/tax bridge. No fixed normalized tax rate is imposed for UiPath. Robust estimation reduces the influence of exceptional tax benefits; the current-tax sensitivity separately illustrates the more recent environment.

Share growth is modeled on a log scale, with a separate nonnegative dilution premium. Forecast medians of individual components need not add or divide to the median EPS; accounting identities hold on every individual path.

Seven expanding historical origins train on earlier quarters only and produce 22 outcomes across one to four quarters. Revenue is refitted at each origin. Latest filing revisions and reconciled release data remain in history, and model choices were reviewed using these results: this is retrospective validation, not an untouched holdout or point-in-time trading test.

80% and 95% ranges are central pointwise simulation ranges. Small, overlapping backtests do not calibrate them. Surprise percentages are conditional simulation fractions, not demonstrated real-world probabilities. A beat means more than $0.01 above consensus; a miss means more than $0.01 below. Everything within ±$0.01 is treated as in line.

Analyst consensus is shown only where the target period and EPS basis can be aligned. No verified analyst cost, common-earnings or share-count forecasts were located; unavailable values stay blank. Management-derived values are scenarios, not independent analyst consensus. Forecast horizons beyond the next quarter have no verified comparable consensus.

Historical analyst estimates are retrospective vendor observations checked against company actuals, not pre-release snapshots captured by us. Stock returns run from the after-close release-date close to the next session’s close. Archived options magnitudes use the vendor’s methodology and have not been independently rebuilt from contract quotes.

Revenue range and approximately $100M adjusted operating income are issued company guidance. Adjusted costs of $342.5M are derived from the revenue midpoint less operating income. The release supplies no quarterly EPS, net-income or share-count guidance. Guidance and sensitivity ranges are not probability intervals. Company outlook and reconciliation.

Computed 2026-11-10 · 6000 simulated paths · 19 accounting quarters through 2026-07-31 · model version path-linked-accounting-v1-2026-11-10. Figures are frozen to this research vintage.