Since January 1, 2026, every candidate pursuing the ACAS designation must pass both parts of the CAS's Property and Casualty Predictive Analytics requirement, a $300 Pearson VUE exam and a separate $700 take-home project, and 67.5% of the 126 candidates in PCPA's first mandatory administration passed both components (CAS Exam Statistics, Spring 2026). No prior CAS requirement has combined an on-demand computer-based test with a quarterly, human-graded independent project, and that structural break is what the first cohort is now navigating without a large pool of prior completers to benchmark against.

From tracking CAS exam progression patterns and candidate sequencing choices since the credential restructuring was first announced, the dual-component structure creates a scheduling problem that candidates consistently underestimate going in. The exam is forgiving: fail it, wait two weeks, sit again, up to three times in a rolling year. The project is not. Fail the project and the next attempt does not open until the following quarterly window, which means a badly timed failure can add three to six months to an ACAS timeline depending on where in the cycle it lands. That asymmetry, not the modeling content itself, is the operational risk the 2026 cohort is actually managing.

A Requirement Built From Two Different Assessment Instruments

The CAS finalized the mandatory transition after extending the original November 1, 2025 deadline to January 1, 2026, giving candidates a final window to complete the legacy ACAS requirements, Exams 1, 2, MAS-I, MAS-II, 5, and 6 plus the DISC courses and Course on Professionalism, without touching PCPA at all ("PCPA Transition Date Extended to January 1, 2026," CAS). Anyone who cleared that full slate before the deadline is grandfathered; everyone else now has PCPA inserted into the pathway as a required, sequenced pair of assessments (CAS PCPA FAQ, April 2026).

The exam half is a two-hour, computer-scored test administered continuously at Pearson VUE centers, with an estimated 80 hours of study and preparation time, no formula sheet provided, and a $100 no-show fee if a candidate registers and skips it (CAS PCPA FAQ, April 2026). It covers GLM diagnostics and distributional selection (Tweedie, Poisson with log-exposure offset, Gamma), tree-based model tuning across Random Forest and XGBoost, variable selection methods including LASSO, and a situational ethics section built around ASOP 12, 23, 41, and 56 alongside the NAIC Model Bulletin on the Use of Algorithms and state instruments like Colorado SB 21-169. A candidate can sit it up to three times within a rolling twelve months, with a mandatory two-week wait between attempts.

The project is a different kind of instrument entirely. Once a candidate passes the exam, they have up to one year to register for and complete the project before the exam credit expires and they would need to pass it again. The project itself runs on four fixed windows a year: CAS distributes a dataset and a business problem, and the candidate has two weeks, working independently with no hands-on help from another person, to submit a technical report capped at exactly 1,000 words, working model code in R, Python, or SAS, and up to five supporting tables or exhibits (CAS PCPA FAQ, April 2026). The word cap is not a style guideline. "Any project submission that goes beyond 1,000 words will not be scored and will result in a failed attempt," the CAS states in its FAQ (CAS PCPA FAQ, updated April 15, 2026). Generative AI tools may be consulted for general syntax or conceptual questions, but candidates are explicitly barred from sharing the proprietary project instructions or dataset with any AI system, and the submitted work must be the candidate's own. Results take six to eight weeks after the submission deadline, so a candidate sitting the project's opening window of a quarter will not learn the outcome until roughly two months into the next one.

What the First Mandatory Cohort's Numbers Show

The Spring 2026 sitting is the only data point available so far for candidates who faced PCPA as a true requirement rather than a beta or optional add-on: 85 of 126 candidates passed, a 67.5% combined rate, tracked on actuary.info's PCPA pass-rate page from CAS's own exam statistics. That figure covers candidates who cleared both the exam and project; the CAS has not yet published a component-level breakdown showing how many failures came from the CBT versus the technical report, and it will not begin reporting project statistics on a quarterly cadence, with exam statistics reported annually, until enough sittings accumulate under the mandatory regime (CAS PCPA FAQ, April 2026). That reporting lag is itself informative: it means candidates entering the project in late 2026 are still calibrating against a single published data point rather than a multi-year trend line, a very different starting position than a candidate approaching, say, Exam 5 with a decade of pass-rate history to study.

The CAS's own time estimates frame the asymmetry between the two components clearly. The exam carries an estimated 80 hours of study and preparation; the project is estimated at roughly 20 hours spread across preparing, drafting, revising, and finalizing the submission inside its two-week window (CAS PCPA FAQ, April 2026). On paper, that makes the project the lighter lift. In practice, compressed timelines punish process failures more than knowledge gaps: a candidate who has not pre-planned how to allocate a 1,000-word budget across problem framing, model selection rationale, and validation discussion will burn disproportionate time mid-window discovering that their draft is 1,400 words with the most important section, why this model over the alternatives, still unwritten.

Why the CAS Wants MAS-I, MAS-II, and Exam 5 First

PCPA has no formal prerequisites, but the CAS is explicit that candidates are "highly encouraged to pass MAS-I, MAS-II and Exam 5 before attempting the exam" (CAS PCPA FAQ, April 2026), and the sequencing logic is more than a suggestion to spread out the workload. MAS-I and MAS-II supply the GLM diagnostics and ensemble-model statistical foundation the PCPA exam tests at a materially deeper level than either MAS exam alone. Exam 5 supplies something different: the ratemaking and classification framework that lets a candidate correctly interpret what a model's output actually means in an insurance pricing context.

That distinction shows up concretely in the project. Two candidates can build technically identical GLMs on the same supplied dataset, one who has taken Exam 5 and one who has not, and produce materially different technical reports. The candidate without ratemaking context is more likely to describe model fit purely in statistical terms, deviance reduction, AIC comparison, residual patterns, without connecting the selected variables to what they mean for a rate indication or a classification plan. Graders evaluate the report against a rubric that rewards actuarial communication quality alongside technical correctness, so a project that reads as a statistics exercise rather than an actuarial one is a real risk for a candidate who jumped ahead of the recommended sequence, even if the underlying model code is sound.

PCPA and CSPA Are Not the Same Credential

The CAS Institute's Certified Specialist in Predictive Analytics, CSPA, predates PCPA and remains a separate, deeper credential rather than a required step. iCAS designs its programs for experienced practitioners already working in quantitative roles, not as an entry-level pathway, and CSPA's own project component is substantially more demanding than PCPA's, as the CAS lays out directly in its FAQ comparison:

Feature PCPA Project CSPA Project
Technical report length Up to 1,000 words Up to 3,000 words
Time to complete 2 weeks 60 days
Business case and dataset Basic More advanced
Mentor option Not available Available, with resubmission window

Source: CAS PCPA FAQ, updated April 15, 2026.

The two credentials intersect at exactly one point: a waiver. Candidates who passed CSPA Course 3, Predictive Modeling: Methods and Techniques, along with its case study project, and did so by actually sitting the course rather than obtaining credit through an IFM exam waiver, are eligible to petition for a full waiver of both PCPA components (CAS PCPA FAQ, April 2026). That path matters most for candidates who came up through a data science or analytics track before formally entering the ACAS pipeline; for a traditional exam-sequence candidate, pursuing CSPA purely to waive PCPA would mean taking on a harder, longer credential to avoid an easier one, which is rarely the efficient move.

Why the Quarterly Window Is the Real Risk, Not the Content

Run the timeline on a badly sequenced failure and the stakes become concrete. A candidate who passes the PCPA exam in January and registers for the next available project window sits in that quarter's two-week submission period; if that attempt fails, they cannot register again until the following quarterly window opens, roughly three months later, and then wait another six to eight weeks after that window's deadline for a result (CAS PCPA FAQ, April 2026). A single failed project attempt at the wrong point in the calendar can therefore push a target ACAS date back three to six months, not because the candidate needs that long to fix a modeling gap, but because the CAS's own administration calendar does not offer a faster retry path. That is a materially different risk profile than a written-answer exam like MAS-II, where a fall failure simply means sitting again the following spring on a known, fixed schedule candidates plan around a year in advance.

The practical hedge is to treat the exam-to-project handoff as a single planned sequence rather than two independent events. Passing the exam with room to spare before a project window's registration deadline, which closes roughly three weeks before that window opens to account for the prior window's grading release, gives a candidate the option to enter the very next administration instead of waiting for the one after it. Candidates who pass the exam close to a registration cutoff, by contrast, frequently miss the near-term window entirely and lose a full quarter before their first project attempt even begins.

What the Mandate Signals to Employers

The PCPA mandate lands at the same moment DW Simpson's 2026 salary survey shows actuaries fluent in Python, R, and SQL earning 10 to 15% more than peers without those skills, a premium that stacks on top of the 15 to 25% jump tied to an ASA or ACAS credential itself (DW Simpson, 2026). Making predictive analytics a baseline requirement for ACAS, rather than a specialty differentiator layered on top of it, effectively folds part of that skills premium into the credential's floor. A candidate who previously distinguished themselves by pursuing CSPA or building a portfolio of side projects is now competing in a pool where every ACAS holder has demonstrated the same GLM diagnostics and model-selection competency through a standardized, externally graded instrument.

That shift changes what is left to differentiate on a resume. Employers running pricing and reserving rotations are increasingly parsing PCPA status as two separate signals rather than one binary credential line: exam-passed but project-pending candidates are read as having the theoretical foundation, while project-complete candidates have additionally demonstrated the ability to frame a business problem, defend a model choice in exactly 1,000 words, and produce reproducible code under a deadline, closer to a compressed consulting deliverable than a classroom exercise. Listing "PCPA Exam Passed" and "PCPA Complete" as distinct line items, rather than folding the requirement into a generic "pursuing ACAS" note, gives a hiring manager that resolution without requiring them to look up transcript status.

Sequencing Strategy for Candidates Now in the Pipeline

The candidates most exposed to the quarterly-window risk are those treating PCPA as an on-demand extension of their existing exam habits rather than a two-stage process with its own calendar logic. The lower-risk sequence, consistent with what the CAS itself recommends and what the project's actuarial-context dependency argues for, is to clear MAS-I, MAS-II, and Exam 5 first, sit the PCPA exam with enough runway before a project registration deadline to enter the next window rather than the one after it, and budget the CAS's own estimated 80 hours of exam prep and 20 hours of project work as two separate, non-overlapping study blocks rather than one continuous push. Candidates who already hold CSPA Course 3 credit earned by sitting the course, not through an IFM waiver, should confirm their waiver eligibility with CAS admissions before building a PCPA study plan at all, since the waiver removes both fee and calendar exposure entirely.

The candidates most likely to lose a full quarter are, based on the pattern so far, those who pass the exam late relative to a registration deadline, attempt the project before finishing Exam 5's ratemaking content, or misjudge the 1,000-word report as a place to narrate methodology chronologically instead of defending a specific model selection decision. None of those failure modes are about statistical knowledge. They are about treating PCPA's administrative structure with the same casualness that has generally worked for CAS's fixed-sitting exams, and the first mandatory cohort's early experience suggests that approach does not transfer.

CAS Exams SOA vs CAS Actuarial Careers

Further Reading

Sources