Allstate: How fast should homeowners growth travel?
Download case PDF (5 pages)Request teacher’s pack
Case information cutoff: 30 September 2026, 11:59 p.m. Eastern Time
Allstate had a homeowners growth opportunity and a risk-allocation problem at the same time. Its policy count was rising, new applications were increasing, and its reported first-half homeowners underwriting result had improved sharply from the prior year. Yet catastrophe losses, reinsurance costs, rate regulation and household affordability made it difficult to tell whether more growth would improve returns or add poorly compensated risk.
Assume you are advising Allstate’s board strategy committee at the cutoff date. The case does not report that this meeting occurred. The committee wants a recommendation on the pace and conditions for extending homeowners growth through Allstate’s product and distribution rollout. What should it approve, what should it constrain, and what evidence should trigger a change in course?
A broad insurance proposition meets local risk
Allstate’s 2025 strategy described its “Affordable, Simple and Connected” auto and homeowners offerings as available in 43 and 31 states, respectively, and set a 2026 target to complete the rollout. Its Custom360 middle-market standard and preferred products had reached 36 states by year-end 2025. The filing described local risk evaluation, competitor pricing, loss-cost trends, customer disruption, affordability, regulation, competitive position and profitability as inputs to property pricing and risk management. The 2026 target is a stated plan, not evidence that every rollout milestone was met or that each market cleared a return hurdle. 2025 Form 10-K, pp. 5–7
The company’s geographic posture was already selective. In its second-quarter filing, Allstate said it had no new homeowners business in Florida and completed certain policy non-renewals there during the quarter. It wrote only a limited amount of new California homeowners business through North Light Specialty, its excess-and-surplus-lines carrier. Management said growth in some states could require regulatory or legislative reforms to permit coverage at appropriate risk-adjusted returns. The filing did not quantify state-level policy counts, lost premium, rate-approval status or returns. Q2 2026 Form 10-Q, p. 51
Growth in the first half
At 30 June 2026, Allstate reported 7.819 million homeowners policies in force (PIF), compared with 7.596 million a year earlier. It received 769,000 new issued applications in the first half, 14.3% more than in the first half of 2025. Management attributed the 2.9% PIF increase primarily to direct and exclusive-agent channels, partly offset by independent agency. Applications describe new business activity; they are not a conversion rate, retention measure or the net contribution to PIF.
Homeowners premiums written were $8.493 billion in the first half, versus $7.848 billion a year earlier. Premiums earned were $8.365 billion, versus $7.428 billion. Written premium is charged when policies are issued; earned premium enters results over the policy term. Management attributed premium growth to higher average premiums from rate changes and higher insured-home replacement costs, as well as PIF growth. Allstate implemented homeowners rate increases of 4.9% in the first half and estimated their total insurance-premium impact at 1.3%, excluding changes in insured-home replacement costs. These are different measures; the 4.9% should not be added to premium growth as though it were a uniform realized price increase. Q2 2026 Form 10-Q, pp. 49, 51
Exhibit 1. Homeowners scale and new applications
| Measure | H1 2026 | H1 2025 | Change / context |
|---|---|---|---|
| Premiums written | $8.493bn | $7.848bn | Reported; timing differs from earned premium |
| Premiums earned | $8.365bn | $7.428bn | Reported; recognized over policy terms |
| Policies in force at June 30 | 7.819m | 7.596m | +223,000; +2.9% reported |
| New issued applications | 769,000 | 673,000 | +14.3% reported |
| Implemented rate increases | 4.9% | — | Estimated total insurance-premium impact: 1.3%, excluding insured-home replacement-cost changes |
Source: Allstate Q2 2026 Form 10-Q, pp. 49 and 51. Interim financial information is unaudited. Policies are policy counts, not households. Applications are not bound policies or net additions. Rate action and estimated premium impact have different timing and scope.
The channel mix sharpened the question. First-half 2026 applications were 501,000 through exclusive agents, 101,000 through independent agents and 167,000 direct. Direct applications rose 75.8% from a smaller base; exclusive-agent applications rose 3.7%, and independent-agent applications rose 6.3%. The filing did not disclose channel-level PIF contribution, retention, acquisition expense or policy profitability. Q2 2026 Form 10-Q, p. 51
Better reported results, but a difficult comparison
Allstate’s homeowners combined ratio was 89.1% for the first half of 2026, compared with 107.1% for the same period in 2025. Its catastrophe-loss effect was 29.3 ratio points, compared with 46.3 points. First-half homeowners underwriting income was $911 million, versus a $527 million loss. The company reported $2.962 billion of first-half catastrophe losses, compared with $4.192 billion a year earlier; the 2025 period included $1.11 billion of California wildfire losses. Changes in catastrophe losses and prior-year reserve effects affect the ratios and underwriting income. These consolidated results do not isolate the economics of newly issued policies or the markets where growth occurred. Q2 2026 Form 10-Q, pp. 48, 52–53
After the quarter, a 17 September Form 8-K furnished an August monthly release. It estimated August catastrophe losses at $748 million before tax ($591 million after tax), across 21 events, with approximately half associated with one wind-and-hail event. The release separately estimated combined July–August losses at $1.43 billion before tax ($1.13 billion after tax). These are management estimates in a furnished exhibit, not audited or final quarter-end results. The cumulative July–August estimate includes August and must not be added to the August estimate. The cited paragraph does not identify line-of-business allocation or gross-versus-net basis. 17 September 2026 Form 8-K and Exhibit 99
Exhibit 2. Underwriting results and catastrophe exposure
| Measure | H1 2026 | H1 2025 | Reading the comparison |
|---|---|---|---|
| Homeowners combined ratio | 89.1% | 107.1% | Reported ratio; includes catastrophe and prior-year reserve effects |
| Catastrophe effect in homeowners combined ratio | 29.3 points | 46.3 points | 2025 includes the California wildfire comparison noted below |
| Homeowners underwriting income (loss) | $911m | ($527m) | Reported result; not the return on new policies |
| Total catastrophe losses reported for the period | $2.962bn | $4.192bn | 2025 included $1.11bn of California wildfire losses |
| Modeled 1-in-100 annual aggregate PML, net of reinsurance | $3.2bn at 30 June 2026 | About $3.1bn at 31 December 2025 | Hurricane, earthquake and wildfire; modeled tail-loss measure, not expected annual loss or a cap |
| Catastrophe reinsurance program cost, excluding reinstatement premiums | $686m | $562m | H1 2026 cost was $124m, or 22.1%, higher year over year (derived) |
Source: Allstate Q2 2026 Form 10-Q, pp. 48, 52–53; 2025 Form 10-K, p. 7. The filing says about 83% of the reduction in net written and earned premium caused by catastrophe-placement premiums relates to homeowners. That percentage is not the homeowners share of the reinsurance cost increase. PML is a modeled loss at a return period, not expected loss. The September 17 furnished estimate is discussed separately above and is not included in this H1 table.
Allstate’s modeled 1-in-100 annual aggregate probable maximum loss (PML) for hurricane, earthquake and wildfire, net of reinsurance, was about $3.2 billion at June 30. First-half catastrophe reinsurance cost, excluding reinstatement premiums, was $686 million, up from $562 million a year earlier. The company said about 83% of the reduction in net written and earned premium caused by catastrophe-placement premiums related to homeowners. These figures describe the program and modeled portfolio exposure; they do not reveal the marginal reinsurance cost, capital need or return from adding a policy in a particular state.
A competitor’s selective approach
Progressive reported that it selectively increased product availability by state in the second quarter of 2026, using considerations that included concentration, product segmentation, rate adequacy, cost sharing, geographic diversification, regulation and market conditions. It lifted some restrictions in selected markets while retaining them where it sought better profitability and lower weather exposure. Its homeowners new applications rose 11% in Q2; PIF was down 2% year over year in its company-defined “growth-oriented” markets and down 17% in its “volatile-weather” markets. Progressive’s categories and measures differ from Allstate’s, so the comparison shows that selective opening and restriction are credible choices, not a transferable performance benchmark. Progressive Q2 2026 Form 10-Q, pp. 27–29, 39–40
The decision
Allstate could accelerate growth where its current underwriting and distribution systems indicate adequate returns, using the rollout to build scale and capture applications. It could instead constrain or reduce exposure until rate approvals, catastrophe trends, reinsurance costs and capital capacity become clearer. A third path would authorize conditional growth: define market- and peril-level entry gates, limit concentration, and continue only while observed pricing, loss emergence, reinsurance and customer measures meet agreed thresholds.
Each path has a cost. Faster growth risks adding exposure before price and capital catch up. A broad hold can protect against underpriced tail risk but may forfeit business in markets where risk selection, distribution or mitigation can support acceptable economics. Conditional growth preserves choices but requires granular data and a willingness to pause markets when evidence turns. The filings do not provide state-level returns, rate-approval timetables, policy contribution margins, retention, customer affordability measures or marginal reinsurance costs. The committee must decide how to act despite those gaps and what information it needs before the next rollout step.
Questions for discussion
- What should Allstate’s homeowners growth posture be at the cutoff: accelerate selectively, constrain exposure, or use explicit conditional gates? State the main reason for your choice.
- Which of the reported first-half indicators best inform that decision, and which are weak proxies for the return on incremental policies? Explain the limits.
- How should Allstate account for catastrophe volatility, regulation, reinsurance and affordability when comparing markets? What would you measure at the state, peril and channel levels?
- What thresholds or review triggers would you set, and what additional data would you request before expanding further?
- What does Progressive’s approach add to the decision, and what can it not tell Allstate?
Source note
The case uses Allstate’s 2025 Form 10-K filed 20 February 2026; unaudited Q2 2026 Form 10-Q filed 5 August 2026; and 17 September 2026 Form 8-K with a furnished Exhibit 99 estimate. It also uses Progressive’s Q2 2026 Form 10-Q. Page and exhibit locators appear above; source versions, archive identities, access depth and further limitations are recorded in the source register. The academic evidence and its external-validity limits are summarized in the literature review.