TLDR: Revenue forecasting models engineered for demand-response businesses produce systematic, directionally predictable error when applied to regulated industries, because the primary revenue driver is the regulatory calendar, a structurally fixed, publicly available input that standard FP&A (financial planning and analysis) frameworks leave entirely unmodelled.
The Structural Mismatch: Forecasting Methodology Built for Demand-Response Businesses Applied to Regulatory-Response Businesses
A Gartner survey of more than 200 CFOs published in August 2025 placed improving forecast accuracy second only to cost optimisation on the executive priority list for 2026. The operational reality behind that ranking is stark: a KPMG and Economist Intelligence Unit study of 544 senior finance executives found that only 22% of firms produce a forecast within 5% of actuals, with an average miss of 13%, and that firms carrying inaccurate forecasts suffered a 6% share-price decline. A 2024 survey of more than 500 finance leaders by Pigment found that 89% make decisions on inaccurate or incomplete data monthly, and only 28% express confidence in planning horizons of one year or beyond.
The CFO community has treated this as a data-quality problem, a technology problem, or a talent problem. Kainjoo’s synthesis of the primary evidence identifies a different root cause in regulated industries: a category error in model architecture. Standard FP&A forecasting was engineered for demand-response businesses, where revenue changes in response to demand signals (consumer preference, pricing, distribution, marketing spend). Regulated industries operate as regulatory-response businesses, where the primary revenue driver is a regulatory decision: an agency approval, a rate determination, a capital adequacy finding. Applying demand-response model architecture to a regulatory-response business creates a structural forecasting error that recurs every cycle.
The distinction is epistemological. In a demand-response business, the forecaster builds a model that translates market signals into revenue projections. In a regulatory-response business, the forecaster must begin with the regulatory calendar, because that calendar is the revenue architecture. The FDA (Food and Drug Administration) PDUFA (Prescription Drug User Fee Act) calendar, the Federal Reserve’s CCAR (Comprehensive Capital Analysis and Review) cycle, and state utility commission rate-case proceedings are public, published, and structurally fixed. A model that treats these as exogenous shocks rather than primary architectural inputs will produce the wrong answer, repeatedly and predictably.
The Pharmaceutical Cascade: How FDA Events Rewrote Pfizer’s Comirnaty Revenue Trajectory
The Pfizer Comirnaty COVID-19 vaccine sequence from 2021 to 2022 offers the clearest documented case of regulatory-calendar architecture driving revenue revision.
Pfizer’s Q1 2021 SEC earnings filing reported guidance of approximately $15.0 billion for Comirnaty revenue, anchored to the then-current EUA (Emergency Use Authorization) for adults aged 16 and older. In May 2021, the FDA expanded that EUA to cover ages 12 to 15. Pfizer’s Q2 2021 8-K reflected a guidance revision to approximately $26.0 billion, an increase of $11 billion or 73%, attributable to a single regulatory action. Through Q2 and Q3 2021, as full BLA (Biologics License Application) approval approached, Pfizer raised guidance further, with subsequent earnings guidance bringing the Comirnaty total toward $33.5 billion per Pfizer’s quarterly filings and analyst tracking by Pharmaceutical Technology. The FDA granted full BLA approval on 23 August 2021, enabling institutional and employer vaccine mandates that broadened the addressable population. Full-year 2021 actual revenue came in at $36.78 billion, a 145% increase over the January 2021 baseline.
The trajectory illustrates directional asymmetry in pharmaceutical revenue: approval events produce structurally upward revisions. Each positive regulatory milestone expands the addressable indication, age group, or mandated-use category. A revenue model calibrated to demand signals alone captures only a fraction of this variance; the regulatory decision itself drove the step-changes, and demand followed.
Peer-reviewed research in the Journal of Pharmaceutical Health Services Research quantifies the forecasting challenge at industry scale: only 24% of pharmaceutical marketing authorisation date predictions are accurate within one month of the actual decision. At the European level, the EMA (European Medicines Agency) carries an average assessment duration of 204 days with an average clock-stop of 198 days, and Papapetropoulos et al. in the British Journal of Pharmacology (2024) found that only 35% of marketing authorisation applications arrive on schedule. Revenue models that treat the approval date as a fixed point anchor to a figure carrying a documented 76% probability of being wrong by more than one month.
The patent-cliff dimension compounds the directional asymmetry in the reverse direction. Evaluate Group’s analysis places $236 billion to $400 billion in branded pharmaceutical revenue at risk from patent expirations between 2025 and 2030. LOE (loss of exclusivity) events are as structurally fixed as approval events; the expiration date is public and generic-entry timing is legally defined. Revenue models that treat LOE as a demand-side competitive event systematically underestimate the precision of the cliff timing. The regulatory calendar governs both the upside (approval) and the downside (expiration) with equal structural force.
PG&E’s 2023 General Rate Case Reveals the Cash-Collection Gap in Utility Revenue Timing
The California utility revenue model rests on a GRC (General Rate Case) determination by the CPUC (California Public Utilities Commission). The authorised revenue requirement sets what a utility may collect from ratepayers; prior to a final CPUC decision, the utility operates on interim rates that may diverge from the authorised level.
Pacific Gas and Electric’s 2023 GRC illustrates the cash-timing dislocation this creates. The CPUC issued its final decision on 17 November 2023, authorising a revenue requirement of $13.52 billion for test year 2023. That decision arrived 10.5 months into the rate year it was intended to govern. Collection through rates at the authorised level commenced on 1 January 2024. The 2023 undercollection was amortised over a 24-month period ending 31 December 2025. The CPUC authorised subsequent year requirements of $14.24 billion (2024), $14.60 billion (2025), and $14.80 billion (2026).
The PG&E Form 10-K for fiscal year 2024 disclosed the operational consequence directly: expenditure during the period of delay may exceed the authorised amount, leaving the utility exposed to unrecoverable costs until the next rate case. This is a structural timing dislocation, recurring by design. A utility’s revenue model that treats the authorised requirement as the 1 January revenue figure for the rate year will produce a timing error every cycle in which the GRC decision arrives after the year has started. CPUC docket data demonstrates that GRC decisions routinely arrive mid-year; the 2023 experience is structurally representative rather than anomalous.
The directional-asymmetry analysis applies here with the opposite sign relative to the pharmaceutical case. Utility revenue revisions triggered by delayed rate decisions point toward cash-collection shortfalls in the current period, with recovery deferred to future periods. The regulatory calendar in this context is a cash-timing architecture: the model that treats rate authority as coterminous with the rate year will systematically misstate when revenue is collectable.
The CCAR/DFAST Cycle Gates Banking’s Capital Distribution Income to a Predictable Q4 Window
Banking revenue forecasting encompasses a category of capital-deployment income (dividends, buybacks, and the share-price support that repurchases provide) that the Federal Reserve’s stress-testing regime gates to a structurally predictable annual window.
The annual sequence operates as follows. In the fourth quarter of the prior year, banks submit capital plans to the Federal Reserve, calibrated to regulator-approved scenarios. In June, the Fed publishes DFAST (Dodd-Frank Act Stress Testing) and CCAR results, establishing each institution’s capital distribution headroom. In August, the SCB (Stress Capital Buffer) is announced, determining the maximum dividend and buyback envelope. On 1 October, the SCB becomes legally effective and capital distribution decisions become executable. In the fourth quarter, banks announce dividend increases and buyback programmes.
The 2025 cycle illustrates the specifics. Goldman Sachs received an SCB of 3.4%, producing a CET1 (Common Equity Tier 1) capital requirement of 10.9% effective 1 October 2025. Bank of America announced a $40 billion buyback authorisation in August 2025, following the June stress test results. JPMorgan Chase published its 2025 DFAST results on 1 July 2025, covering a nine-quarter projection period.
A standard FP&A model for a major bank that projects capital-distribution-related income on a linear or demand-driven basis will misattribute the timing of that income. The capital distribution decision is gated, by regulation, to a post-October window. The distribution envelope is bounded by the SCB outcome published in August. Both the window and the ceiling are structurally fixed by the CCAR/DFAST calendar. A revenue model that treats capital distribution as a continuous variable governed by management discretion alone will produce a timing error that repeats every year on precisely the same schedule.
ASC 606 and IFRS 15 Embed Regulatory-Cycle Variance Directly into the Income Statement
Beyond the forecasting dimension, the regulatory-cycle overlay carries an accounting-disclosure imperative. Both ASC (Accounting Standards Codification) 606 and IFRS (International Financial Reporting Standards) 15 tie revenue recognition to the satisfaction of performance obligations. In regulated industries, the performance obligation is frequently the regulatory-approval event itself.
Under ASC 606, pharmaceutical milestone payments and licensing revenues are recognised at the point when the performance obligation, commonly the receipt of regulatory approval, is satisfied. This produces recognition as a single event on the approval date rather than as a smooth accrual over the development period. IFRS 15 applies the same five-step model: identify the contract, identify the performance obligations, determine the transaction price, allocate it to the obligations, and recognise revenue as each obligation is satisfied. Where regulatory approval is the performance obligation, recognition concentrates at the approval event.
The accounting consequence is that regulatory-cycle variance embeds directly in the income statement timeline. A pharmaceutical company whose approval is delayed by three months will report a three-month deferral of milestone or licensing revenue attributable to regulatory timing rather than to product quality, demand, or pricing. The CFO whose forecast treated the recognition date as fixed will report a miss on the income statement that is, in origin, a regulatory timing miss. The regulatory-cycle overlay is therefore also a revenue-recognition planning tool and a disclosure imperative under both US GAAP (Generally Accepted Accounting Principles) and IFRS, extending its relevance from the FP&A function to the external reporting function.
Revenue Driver Comparison: Standard Model Versus Regulated-Industry Adjusted Model
Table 1: Revenue Driver Comparison, Standard Model Versus Regulated-Industry Adjusted Model
| Revenue Driver | Standard Model Approach | Regulated-Industry Adjusted Model | Primary Evidence |
|---|---|---|---|
| Product Launch Revenue | Demand-curve projection from market research; launch date treated as a fixed, internally determined input | Probability-weighted PDUFA-date scenarios (base, accelerated, delayed); revenue recognition date mapped to ASC 606 and IFRS 15 approval-event performance obligation | Pfizer Comirnaty guidance revised from $15.0B to $36.78B (2021) across four FDA milestone events; 145% upward revision driven by regulatory actions, with demand following |
| Annual Revenue Baseline and Rate Authority | Authorised revenue requirement assumed effective 1 January of rate year; cash collection modelled as continuous from year-start | Rate-case docket timing model: probability distribution over decision date; cash-collection commencement date and undercollection amortisation schedule modelled separately from the income-statement authorised requirement | PG&E 2023 GRC: CPUC decision issued 17 November 2023 (10.5 months into rate year); authorised $13.52B collectable from 1 January 2024; 2023 undercollection amortised across 2024 to 2025 |
| Capital Distribution Timing in Banking | Dividend and buyback income projected as a continuous variable governed by management discretion and earnings capacity | CCAR and DFAST overlay: capital-distribution income assigned to post-October window by default; distribution envelope capped by August SCB announcement; scenario branches built for adverse and severely adverse DFAST outcomes | Goldman Sachs SCB 3.4% effective 1 Oct 2025; Bank of America $40B buyback authorised August 2025 post-CCAR; JPMorgan Chase 2025 DFAST results published 1 July 2025 |
| Patent and Exclusivity Cliff Revenue | Revenue decline modelled as competitive demand erosion post-LOE; generic-entry timing treated as a market variable subject to competitive dynamics | LOE date-anchored cliff model: expiration date as structurally fixed regulatory input; generic-entry timing derived from Paragraph IV challenge data and FDA Orange Book; revenue step-down mapped to statutory exclusivity end-date | Evaluate Group: $236B to $400B in branded pharma revenue at risk from patent expirations 2025 to 2030; LOE events legally determined and publicly available in advance |
| Regulatory Timeline Uncertainty as Base-Case Variance | Approval or decision date treated as a point estimate; variance in outcomes attributed to market or operational factors | Empirical regulatory-clock distribution as base-case input: EMA average assessment 204 days, clock-stop 198 days; only 24% of MAA predictions accurate within one month; variance budget explicitly allocated to regulatory timing risk | Papapetropoulos et al., British Journal of Pharmacology, 2024; Mahinc et al., Journal of Pharmaceutical Health Services Research (PMC5021138) |
Sources: Pfizer SEC 8-K filings Q1 and Q2 2021; CPUC Press Release 17 November 2023; PG&E Form 10-K FY2024 (SEC); Goldman Sachs SCB filing (SEC, 2025); Bank of America Form 10-Q Q3 2025 (SEC); JPMorgan Chase 2025 DFAST Disclosure; Evaluate Group patent-cliff analysis; Papapetropoulos et al., British Journal of Pharmacology, 2024; Mahinc et al., Journal of Pharmaceutical Health Services Research, PMC5021138.
The Regulatory-Cycle Overlay: Architecture, Data Inputs, and Implementation
Incorporating the regulatory calendar into a revenue model requires a structural change to the model architecture: adding a regulatory-event layer that functions as the primary variable for each revenue category, treated at the same level of precision that demand-side models give to pricing and volume.
For pharmaceutical FP&A, the overlay begins with the PDUFA target action date for each asset in the pipeline. Because only 24% of marketing authorisation predictions land within one month of the actual decision, the model carries probability-weighted scenarios for approval timing (base, accelerated, delayed), each mapped to the revenue recognition date under ASC 606 or IFRS 15. The Papapetropoulos et al. empirical data on EMA clock-stops provides a grounded distributional input for the delay scenario. The LOE calendar supplies the downside boundary for each product. Together, these two regulatory anchors (the approval horizon and the exclusivity expiration) frame the full revenue arc for every asset in the portfolio.
For utility FP&A, the overlay is a rate-case docket timing model: track the CPUC (or relevant commission) filing dates, statutory decision timelines, and historical decision-lag distributions. The model builds a probability distribution over the decision date and maps each scenario to the cash-collection commencement date. Undercollection and amortisation schedules then feed the cash-flow projection separately from the income-statement authorised requirement, eliminating the timing conflation that standard models embed.
For banking FP&A, the CCAR/DFAST overlay is the most structurally deterministic of the three, because the Federal Reserve publishes its calendar in advance. The model assigns capital-distribution income to the post-October window by default, uses the August SCB announcement as the distribution-envelope input, and builds scenario branches for adverse and severely adverse DFAST outcomes to bound the downside.
In all three cases, the regulatory calendar inputs are publicly available in SEC filings, agency press releases, commission dockets, and Federal Reserve stress-testing documentation. The implementation gap is architectural rather than informational: the data is present; standard FP&A frameworks have historically treated the model structure to receive it as beyond their scope.
The Regulatory Calendar as Competitive Intelligence, Extending Well Beyond Internal Finance Discipline
The PDUFA calendar, CCAR results, and state utility commission dockets are public records. A CFO who builds a regulatory-cycle overlay gains an external intelligence advantage that extends well beyond internal forecasting accuracy, because the same structural visibility that sharpens internal projections also illuminates the revenue architecture of every regulated competitor.
On the pharmaceutical side, PDUFA target action dates are published on the FDA’s website. A CFO who tracks the approval calendars of competitors can anticipate when a competitor’s revenue will step up on approval of a rival indication, or step down on LOE, and adjust competitive resource allocation accordingly. The pipeline visibility that the regulatory calendar provides functions as a forward-looking competitive landscape map, updated by every FDA action letter and EMA assessment clock-stop across the industry.
On the banking side, CCAR results are published and ranked by institution. An institution carrying a lower SCB than its peer group holds a structural capital-distribution advantage in the October-to-December window. A CFO who reads competitors’ SCB outcomes can calibrate share-repurchase aggressiveness with knowledge of the competitive distribution envelope alongside the institution’s own constraints, transforming a regulatory compliance output into a capital-strategy input.
On the utility side, GRC proceedings are public docket matters. A utility’s CFO who monitors the rate-case schedules of peer utilities can anticipate which competitors will face undercollection periods and when their rate authority will reset, informing partnership, acquisition, or infrastructure-investment timing with a precision that demand-side analysis alone would be unable to provide.
The Evaluate Group’s documentation of the $236 billion to $400 billion patent cliff is, from this perspective, a competitive-intelligence asset as much as an internal forecasting input. The CFO of a pharmaceutical company who has mapped competitor LOE events against the institution’s own pipeline approval calendar holds a structural foresight advantage in capital allocation and business development decisions.
The regulatory-cycle overlay reframes the CFO’s role in regulated industries. The regulatory calendar functions simultaneously as the primary architectural input to the revenue model and as a public intelligence resource that standard FP&A frameworks have routinely left unread. The CFO who installs the overlay produces more accurate forecasts and, simultaneously, gains a structural intelligence edge over peers whose models remain calibrated to demand-side signals alone. The blind spot is architectural; the correction is available in the public record.
References
- Gartner. “Gartner Survey Shows Top Priorities for CFOs in 2026 Include Cost Optimization,” August 2025. https://www.gartner.com/en/newsroom/press-releases/2025-08-12-gartner-survey-shows-top-priorities-for-cfos-in-2026-include-cost-optimization
- KPMG and Economist Intelligence Unit. “Forecasting with Confidence,” 2025. https://assets.kpmg.com/content/dam/kpmgsites/se/pdf/2025/forecasting-with-confidence.pdf
- Pigment. “Office of the CFO 2024 Report,” 2024. https://www.pigment.com/newsroom/office-of-the-cfo-2024-report-from-pigment
- Pfizer. Form 8-K Q1 2021, SEC filing, April 2021. https://www.sec.gov/Archives/edgar/data/78003/000007800321000062/pfe-04042021xex99.htm
- Pfizer. Form 8-K Q2 2021, SEC filing, July 2021. https://www.sec.gov/Archives/edgar/data/78003/000007800321000090/pfe-07042021xex99.htm
- Pharmaceutical Technology. “Pfizer Full Year 2021 Revenues.” https://www.pharmaceutical-technology.com/news/pfizer-full-year-2021-revenues/
- Mahinc M et al. “Prediction of marketing authorisation dates for new medicines in Europe,” Journal of Pharmaceutical Health Services Research. https://pmc.ncbi.nlm.nih.gov/articles/PMC5021138/
- Papapetropoulos S et al. “Novel drugs approved by the EMA, the FDA, and the MHRA in 2023,” British Journal of Pharmacology, 2024. https://bpspubs.onlinelibrary.wiley.com/doi/10.1111/bph.16337
- Evaluate Group. “Portfolio Tactics to Scale the $300bn Patent Cliff.” https://www.evaluate.com/thought-leadership/portfolio-tactics-to-scale-the-300bn-patent-cliff/
- California Public Utilities Commission. “CPUC Prioritizes Safety, Reliability and Affordability in PG&E Rate Case 2023,” 17 November 2023. https://www.cpuc.ca.gov/news-and-updates/all-news/cpuc-prioritizes-safety-reliability-and-affordability-in-pge-rate-case-2023
- PG&E Corporation. Form 10-K FY2024, SEC filing. https://www.sec.gov/Archives/edgar/data/1004980/000100498025000010/pcg-20241231.htm
- Goldman Sachs. Exhibit 99.1, Stress Capital Buffer announcement, SEC filing, 2025. https://www.sec.gov/Archives/edgar/data/886982/000119312525154097/d86055dex991.htm
- Bank of America. Form 10-Q Q3 2025, SEC filing. https://www.sec.gov/Archives/edgar/data/70858/000007085825000405/bac-20250930.htm
- JPMorgan Chase. “2025 DFAST Results and Methodology Disclosure,” 1 July 2025. https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2025/2025-dfast-results-and-methodology-disclosure/2025-results-methodology-disclosure.pdf
- Financial Accounting Standards Board (FASB). Accounting Standards Codification 606: Revenue from Contracts with Customers. https://asc.fasb.org/606
- International Accounting Standards Board (IASB). IFRS 15: Revenue from Contracts with Customers. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-15-revenue-from-contracts-with-customers/
- US Food and Drug Administration. “Coronavirus (COVID-19) Update: FDA Authorizes Pfizer-BioNTech COVID-19 Vaccine for Emergency Use in Adolescents,” May 2021. https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-authorizes-pfizer-biontech-covid-19-vaccine-emergency-use
- US Food and Drug Administration. “FDA Approves First COVID-19 Vaccine,” August 2021. https://www.fda.gov/news-events/press-announcements/fda-approves-first-covid-19-vaccine
- Federal Reserve. “Comprehensive Capital Analysis and Review (CCAR).” https://www.federalreserve.gov/supervisionreg/ccar.htm
