Definition:Interrupted time series analysis (ITSA)
📉 Interrupted time series analysis (ITSA) is a quasi-experimental statistical method that evaluates the impact of an intervention or event by comparing the trend of an outcome variable before and after a clearly defined interruption point. In insurance, ITSA is particularly well suited for assessing the effects of regulatory changes, new underwriting guidelines, claims process reforms, or market-wide events on metrics like loss ratios, claims frequency, premium adequacy, or lapse rates. Because many insurance interventions are implemented at a known point in time across an entire portfolio or jurisdiction — rather than being randomly assigned to individual policyholders — ITSA provides a natural and powerful analytical framework.
📊 The technique works by modeling the pre-intervention trend using a series of regularly spaced observations — monthly claims counts, quarterly incurred losses, or annual combined ratios — and then testing whether the intervention produced an immediate level change, a change in the ongoing trend, or both. For example, when a European regulator mandated gender-neutral pricing under an EU directive, ITSA enabled actuaries to isolate the policy's impact on motor insurance loss patterns by comparing the pre-mandate trajectory with the post-mandate reality. Similarly, a health insurer in the United States could use ITSA to evaluate how the introduction of prior authorization requirements affected specialty drug expenditures. Adding a control series — a comparable portfolio or jurisdiction unaffected by the intervention — strengthens internal validity by accounting for concurrent external trends, a design known as controlled interrupted time series.
🛠️ One of ITSA's greatest strengths is its transparency: the visual depiction of pre- and post-intervention trends makes the analysis accessible to non-technical stakeholders, from board members to regulators, without sacrificing statistical rigor. This makes it a valuable communication tool when insurers need to demonstrate — to supervisory authorities under Solvency II, RBC, or C-ROSS regimes — that a change in operations or market conditions has materially altered their risk profile. Reinsurers examining cedant portfolios can also employ ITSA to evaluate whether shifts in experience are attributable to specific underwriting actions or simply reflect broader market cycles. The method does require a sufficient number of pre-intervention observations and careful attention to seasonality and autocorrelation, but when these conditions are met, ITSA delivers some of the most credible causal evidence available from observational insurance data.
Related concepts: