PGR — Q3 2026 Financial Report Analysis

Overview

Progressive earned $2.54 billion in the third quarter of 2026, down 3% from $2.62 billion a year earlier, or $4.37 per share versus $4.45. On the surface the quarter looks steady: the combined ratio (claims plus operating costs as a share of premiums earned; below 100 means the insurance business made a profit) improved to 88.7 from 89.5. That comparison flatters it. The year-ago quarter carried a one-time $950 million charge for refunds owed to Florida auto customers. Without that charge, last year's ratio would have been about 84.9, so the underlying business got roughly 4 points less profitable. Claims cost more relative to premiums, premium growth slowed to 4%, and a loss on investments held at market value cut another $465 million from pretax income compared with a year ago.

This analysis is based on Progressive's September monthly results release (8-K Exhibit 99.1, October 14, 2026), which covers the quarter. Progressive says it will file the full 10-Q on November 2, 2026, so details such as claim severity, advertising spend and policy retention are not available yet. Quarterly figures that the release does not state directly (revenue, loss ratio, segment ratios, catastrophe losses) are derived below from its year-to-date and monthly tables, and are labelled as such.

At a glance

• Loss ratio 68.2% vs 64.5%: claims ate 3.7 cents more of every premium dollar than a year ago. This is the main reason profits fell once the Florida charge is set aside.
• Net premiums written +4% ($22.3B): growth halved from 10% a year ago, and September alone grew just 2%. Policies in force added 403,000 in the quarter, about half last year's 763,000.
• Book value per share $59.51, down 1.6%: rising bond yields cut the value of the bond portfolio, so comprehensive income was only $645 million against $2.54 billion of net income.

Key figures

Net premiums written, premiums earned, combined ratio, net income, EPS, book value and policies come straight from the release. Revenue and the loss ratio are year-to-date September minus year-to-date June (from the June release). Catastrophe losses are the sum of each month's stated catastrophe loss ratio times that month's premiums earned, so they are accurate to within a few million dollars. LAE means loss adjustment expenses, the cost of handling claims.

Where the profit went

An insurer makes money two ways: underwriting (charging more in premiums than it pays out in claims and costs) and investing the premiums it holds before claims are paid. Here is how pretax income moved between the two quarters:

Underwriting profit is derived as premiums earned times (100 minus the combined ratio) and checked against the income statement lines; it ties to within a few million dollars.

Three forces account for the quarter:

1. Claims rose faster than premiums. Losses and claim-handling costs were $14.9 billion, up 10.9%, while premiums earned grew 4.9%. That pushed the loss ratio up 3.7 points. Progressive's monthly releases also report an accident-year loss ratio, which counts only accidents that happened in the period and so strips out revisions to older claims. Weighting the three monthly figures, it was about 70.5% in Q3, versus about 66.6% in Q3 2025 and 70.0% in Q2 2026. Higher catastrophe losses (storms, hail) explain only about 0.8 points of the 3.9-point rise. The remaining ~3 points is ordinary claims costing more relative to what customers pay. The Q2 10-Q blamed higher severity (the average cost of each claim, especially bodily injury) while personal auto rates were being trimmed; the Q3 10-Q will show whether that is still the cause.

2. Operating costs were flat. The expense ratio (selling and running the business as a share of premiums) was about 20.5%, versus about 20.5% a year ago once the Florida credit is removed. The improvement in the headline combined ratio is entirely the Florida charge dropping out.

3. Investments helped on income but hurt on market moves. Investment income rose 12% to $1.04 billion on a larger portfolio; the pretax book yield was 4.3% in September. But Progressive marks its stocks and some hybrid securities to market through net income, and those marks turned negative: total realized results were a $170 million loss versus a $295 million gain. September alone had a $228 million holding-period loss.

Segment snapshot (derived from monthly releases)

The release does not give quarterly segment ratios, so the combined ratios below weight each month's stated ratio by that month's premiums earned. Treat them as close approximations; the 10-Q will give exact figures.

Commercial Lines (mostly business auto and trucking insurance) grew fastest but had the weakest month of the quarter: a 97.5 combined ratio in August, leaving it just 2.5 cents of profit per premium dollar that month. Personal property got worse mainly because of storms: its catastrophe loss ratio was 10.9% in July and 7.8% in August. The direct auto channel, which drove most of Progressive's growth in the first half (+8% premiums in Q2), slowed to +5% for the quarter and +1% in September.

What the headline numbers hide

• The "improved" combined ratio is a comparison artifact. The 0.8-point improvement exists only because Q3 2025 included the $950 million Florida policyholder credit, about 4.6 points of that quarter's ratio. On a like-for-like basis the business went from roughly 84.9 to 88.7.
• Older claims are still settling cheaper than expected, and that help grew. Favorable prior-year reserve development (claims from earlier years costing less than reserved) totaled about $534 million across July–September, or 2.4 points of premiums, versus about $461 million (2.2 points) a year ago. Without it, the combined ratio would have been about 91. Releases like this are not guaranteed to continue.
• Growth is slowing faster than the quarterly number shows. Monthly premium growth went 5% (July), 6% (August), 2% (September). Total policies in force grew 6% year over year at September 30, down from 7% at June 30, and personal auto policies grew by 263,000 in the quarter versus 581,000 in the same quarter of 2025.
• EPS fell less than net income because of buybacks. Average diluted shares fell 1% to 582.2 million, so EPS dropped 1.8% versus 2.8% for net income. Progressive bought 969,262 shares in September at an average $207.92, against $242.04 a year earlier. The tax rate was steady at about 21.4% versus 21.3%, so taxes did not move the result.
• Book value went backwards. Unrealized losses on the bond portfolio grew to $3.46 billion pretax at September 30 (from a $249 million gain a year earlier), including a $1.66 billion drop in September alone. Together with the $13.60-per-share annual dividend paid in January (disclosed in the Q2 10-Q), that leaves book value per share 1.6% below a year ago, even though it edged up from $59.05 at June 30. These are paper losses on bonds with a 3.6-year average duration (a measure of sensitivity to interest rates) and AA- average credit quality; they reverse as the bonds approach maturity if they are held.
• Balance sheet items look in line. Premiums receivable grew 5%, in step with premiums. Loss reserves grew faster, up 11% to $46.8 billion; with reserve releases still favorable, this does not look like under-reserving. Debt rose to $8.4 billion from $6.9 billion after a March bond issue, and quarterly interest expense rose to $87 million from $70 million. Debt-to-total capital is 19.6%. Cash flow is not in the release.

Takeaway: Strip out last year's $950 million Florida refund and Progressive's underwriting profit fell about 21% in Q3. Claims now take about 4 points more of each premium dollar on current-year accidents than a year ago, and growth slowed to 2% in September. The business is still very profitable at an 88.7 combined ratio, but margins are clearly shrinking, not holding steady.

Did last time's read hold up?

Our Q2 2026 analysis said margins were "narrowing gradually, not collapsing" and named three things to watch:

• Whether June's 90.0 combined ratio was one weak month or a trend: a trend. July came in at 86.8 (85.3 a year earlier), August at 89.3 (83.1) and September at 90.0. September 2025's 100.4 included the Florida credit; Progressive's footnote put personal lines at 86.5 without it.
• Accident-year loss ratio against pricing: it kept rising, from 70.0% in Q2 to about 70.5% in Q3.
• Hurricane season: catastrophe losses were higher than a year ago at about $394 million, but far below Q2's $773 million and not a major hit. The season runs into late November, so Q4 is still exposed.

The "gradual" framing still fits in that the business remains comfortably profitable. The pace of decline is faster than Q2 suggested, though: like-for-like underwriting profit fell 21% this quarter versus 3% in Q2.

Outlook

The monthly release gives no guidance and no management commentary ("no additional commentary regarding September's results"). Progressive's stated goal is a calendar-year underwriting margin of at least 4%, which means a combined ratio of 96 or lower. Year-to-date, it stands at 87.5, so the target is not in question for 2026.

Our read: Progressive is deliberately giving up margin to grow, with rates trimmed and heavy advertising according to the Q2 10-Q. The Q3 numbers show margin falling while growth also slows, which is the less comfortable combination. What to watch:

• The Q3 10-Q (November 2) and the November 3 call: claim severity trends, advertising spend, and whether management still describes personal auto as "adequately priced". A shift toward rate increases would be the clearest sign it sees the loss trend as lasting.
• October results (scheduled for November 18): whether premium growth stays near September's 2%, and whether the combined ratio stays around 90.
• Commercial Lines: after a 97.5 combined ratio in August, any month above 100 would mean that segment is losing money on underwriting.
• Reserve releases: about a fifth of this quarter's underwriting profit came from older claims settling cheaper. If those releases shrink while the current-year loss ratio keeps rising, the combined ratio would move into the low 90s.