Allstate Reports $220 Million of September Catastrophe Losses, Taking Third-Quarter Losses to $1.65 Billion and Six-Month Losses to $3.4 Billion

Allstate, the largest publicly traded personal-lines insurer in the United States, reported on October 15, 2026 that its catastrophe losses for September were $220 million before tax, or $173 million after tax. The figure comes from the company's monthly press release filed with the Securities and Exchange Commission as an exhibit to a Form 8-K. The same release put catastrophe losses for the third quarter of 2026 at $1.65 billion before tax, or $1.3 billion after tax, and noted that the company had 216 million policies in force across all of its products. The release does not break policies in force down between auto and homeowners, and it does not state the number of catastrophe events behind the monthly total, so those details are not covered here.

Allstate's monthly disclosure is unusual among large insurers and is a useful data point for anyone trying to follow the cost of weather. Catastrophe losses, in the company's definition, are the claims costs from events that cause large numbers of claims, such as a hailstorm, a tornado outbreak or a wildfire. Because the company reports them every month rather than waiting for a quarterly earnings call, investors can track the trend through the year and adjust expectations for the quarter before the full results arrive. Artemis, the catastrophe bond and reinsurance news site, compiled the monthly figures in its coverage on October 16, 2026. According to that report, Allstate's catastrophe losses were $870 million in April, $289 million in May, $563 million in June, $682 million in July, $748 million in August and $220 million in September. Added together, those six months come to $3.372 billion, which Artemis reported as the total since April 1. The three summer months of July, August and September add up to $1.65 billion, which matches the third-quarter figure in the company's release and is a useful cross-check.

The shape of the series is worth reading carefully. April, at $870 million, was the most expensive month, which is consistent with the spring severe convective storm season in the central and southern United States, when hail and tornadoes tend to produce the largest claims counts. May was much lighter at $289 million. June, July and August then ran between $563 million and $748 million, and September dropped to $220 million. A low September does not by itself mean the year has been light: even though it was the lowest of the six months, the six-month total of $3.372 billion is a large number by the standards of a single insurer. It is also true that the monthly pattern is lumpy, because one large hailstorm over a densely populated metropolitan area can add several hundred million dollars in a single week.

The more technical part of the story concerns how Allstate protects itself against that volatility. Artemis reports that the $150 million Class B tranche of the Sanders Re III Ltd. (2023-1) catastrophe bond attaches at $4.78 billion and exhausts at $5.28 billion. A catastrophe bond is a security sold to capital markets investors. If losses in the defined period reach the attachment point, investors begin to lose some of their principal, and the money goes to the insurer to pay claims; if losses reach the exhaustion point, the tranche is fully used. Artemis states that the cover applies to all US states except Florida and that a $50 million per-event deductible applies, which means that individual events below that size do not count toward the aggregate. The bond is described as an aggregate structure, meaning it responds to the accumulation of many smaller events during the period rather than to one big hurricane.

Comparing the figures requires care. Allstate's reported catastrophe losses of $3.372 billion for April to September are below the $4.78 billion attachment point reported for the bond, a gap of about $1.4 billion on simple arithmetic. However, the exact definitions of the covered losses, the measurement period and the way events are counted are set out in the bond's offering documents, which were not reviewed for this article, and the reported monthly catastrophe losses may not map one-to-one onto the losses that count under the cover. The comparison is therefore an indication of how far losses are from the layer rather than a statement of whether the bond will be triggered.

Artemis also reports that Allstate has bought a new aggregate excess-of-loss reinsurance layer of $1 billion. The layer attaches at $8.5 billion and carries a $1 million event deductible. An aggregate excess-of-loss contract pays when the sum of covered losses across a period passes a threshold, as opposed to a per-event contract that pays when a single event passes a threshold. The $1 million event deductible reported means very small events are excluded from the aggregate count. At an $8.5 billion attachment, the layer is positioned far above the losses reported so far this year and would respond only in a very severe year for accumulated catastrophe losses; the reporting does not say whether it covers the same lines or the same period as the catastrophe bond.

For policyholders, the connection between monthly catastrophe losses and what they pay is indirect but real. Homeowners insurers in states with frequent hail and wind have raised premiums, increased deductibles for wind and hail, and in some places stopped writing new policies in the highest-risk areas. Auto insurance is affected too, because hail damages vehicles and floods destroy them. Reinsurance and catastrophe bond layers matter because they determine how much of a bad year an insurer carries on its own balance sheet versus passing to reinsurers and investors; the more cover a company has above its retention, the less a run of severe months threatens its capital, and the more stable its pricing can be. The release and the trade coverage do not give Allstate's retention or its full reinsurance programme, so no conclusion can be drawn here about how much of the $3.372 billion was recovered from reinsurers.

What to watch next is Allstate's third-quarter earnings release, for which the date was not given in the sources reviewed, and the October catastrophe figure, which would follow in the company's monthly series. Investors will look at whether the low September reading was an exception or the start of a quieter autumn, and at how the company describes reserve development, auto margins and homeowners rate actions in its commentary. For the wider industry, Allstate's monthly series is one of the clearest public indicators of whether 2026 will be remembered as an average or a costly year for US storm losses.