Progressive Q3 2026: 88.7 combined ratio, policy growth roughly halved

Progressive, which describes itself as a leading seller of personal auto, commercial auto, motorcycle and boat insurance and which reports results every month, posted Progressive Q3 2026 earnings of $2.54 billion, or $4.37 a share, on $22.3 billion of net premiums written. PGR’s combined ratio — the share of each premium dollar spent on claims and expenses — was 88.7, better than 89.5 a year ago and far inside the company’s long-standing goal of 96 or better.

The headline comparison is the least informative number in the release. The year-ago quarter absorbed a $950 million Florida policyholder credit expense, a one-time charge that added about 4.6 points to that quarter’s combined ratio on our arithmetic. Take it out and Progressive’s underwriting margin narrowed by roughly 3.8 points, driven entirely by higher claims costs. At the same time, the business added about half as many policies in the quarter as it did a year earlier. Insurance terms are defined as they come up.

Progressive Q3 2026 results at a glance

Progressive’s release gives a short quarterly table; the rest of the quarter is reported as September-month and year-to-date figures. The quarter column below comes straight from that table except where marked as derived (our arithmetic: year-to-September minus year-to-June, using the June release).

	3Q26	2Q26	3Q25
Net premiums written	$22,301M	$21,077M	$21,384M
Net premiums earned	$21,869M	$21,573M	$20,849M
Net income	$2,542M	$3,311M	$2,615M
Net income per share	$4.37	$5.67	$4.45
Pretax net realized gains (losses) on securities	$(170)M	$604M	$295M
Combined ratio	88.7	87.3	89.5
Investment income (derived)	$1,035M	—	$924M
Average diluted shares	582.2M	584.2M	588.2M
Policies in force, period end (thousands)	40,489	40,086	38,078

Two premium lines matter for an insurer. Net premiums written (NPW) are the premiums on policies sold or renewed in the period, after reinsurance — the best gauge of sales. Net premiums earned (NPE) are the slice of those premiums that corresponds to coverage already provided; because a six-month auto policy is earned month by month, NPE lags NPW. NPW grew 4% and NPE 5%.

Net income fell 3% and per-share earnings 2%, helped by a 1% lower average share count. Pretax income was $3,236 million against $3,323 million (derived, our arithmetic). Excluding realized gains and losses on securities, it rose to about $3,406 million from about $3,028 million (our arithmetic) — but the year-ago figure has the $950 million Florida charge inside it, which is the subject of the next section.

Did Progressive issue guidance for Q3 2026?

No — Progressive does not issue quarterly earnings guidance: its 2025 Form 10-K says “We do not provide earnings estimates to the market,” and its June release (Form 8-K, Exhibit 99, filed July 15, 2026) contains no outlook, only the line that the company has no additional commentary. Against the year-ago quarter, net income fell 3% to $2,542 million, per-share earnings fell 2% to $4.37, and net premiums written rose 4% to $22,301 million. The only standing target is the calendar-year goal of a 96 combined ratio or better, against which the first nine months ran at 87.5.

The $950 million in the year-ago quarter

Florida law caps the profit an insurance group may earn on personal auto insurance over any three-calendar-year period. In its September 2025 release, Progressive said lower loss costs after Florida’s 2023 insurance reform made it probable its 2023–2025 Florida personal auto profit would exceed that limit, and in September 2025 it recorded a $950 million policyholder credit expense — an estimate of the excess it expected to credit to Florida personal auto policyholders active at December 31, 2025. The company booked it in the expense ratio of its Personal Lines vehicle business. No such expense appears in 2026: the year-to-date line is zero.

That one charge explains the September month comparison in the new release: a September combined ratio of 90.0 against 100.4 and September net income of $630 million against $305 million. It also hides what happened over the full quarter.

The combined ratio has two parts. The loss and loss adjustment expense (LAE) ratio is claims paid and reserved for, plus the cost of handling them, as a share of earned premium. The expense ratio is the cost of acquiring and running the business. Rebuilding both from the income statements (our arithmetic, treating fees and other revenues as an offset to underwriting expenses — the treatment that reproduces the reported 88.7 and 89.5):

Third quarter, our arithmetic	3Q26	3Q25 as reported	3Q25 excluding the Florida credit
Loss/LAE ratio	~68.2	~64.5	~64.5
Expense ratio	~20.5	~25.0	~20.5
Combined ratio	88.7	89.5	~84.9

On this basis, expenses were flat as a share of premium and the deterioration sits in claims: the loss/LAE ratio rose about 3.7 points, which accounts for the roughly 3.8-point combined-ratio change once rounding is allowed for (our arithmetic). Losses and LAE were $14,904 million in the quarter against $13,445 million a year earlier, up 11%, while earned premium grew 5% (our arithmetic). An 88.7 is still a very profitable quarter for a property-casualty insurer, where 100 is break-even on underwriting; the change is in direction, not in level (our reading).

Current-year losses, catastrophes and reserve releases

Three forces move a loss ratio, and the monthly supplements let us separate them roughly.

Prior-year reserve development is the change in what an insurer expects to pay on claims from earlier years; “favorable” development means old claims are costing less than reserved, and the release flows into current profit. Progressive reported $1,536 million of favorable development for the first nine months of 2026, against $1,002 million through June — so about $534 million in the third quarter, or about 2.4 points of earned premium (our arithmetic). A year earlier the equivalent figure was about $461 million, about 2.2 points. Reserve releases were therefore slightly larger, not smaller, and do not explain the weaker loss ratio.

That means the accident-year loss ratio — losses on claims that occurred in the current year, which strips out reserve releases — moved even more: about 70.6 against about 66.7 a year earlier (our arithmetic).

Catastrophe losses — losses from events such as hurricanes, hail and wildfire — explain only part of the gap. The release gives catastrophe loss ratios only for September (1.4 points companywide) and for the year to date (2.2 points, against 2.3 a year ago). Backing out the year-to-June figures implies catastrophes cost roughly $0.4 billion, or about 1.8 points, in the third quarter, against roughly 1.2 points a year earlier. Because those ratios are published rounded to one decimal, treat the quarterly split as approximate (our arithmetic). It still leaves about 3 points of accident-year deterioration that is not weather.

The release does not explain the rise — September carries the line “The company has no additional commentary,” and there are no executive quotations in the filing. This piece does not attribute it to any particular cause. The third-quarter Form 10-Q, which the company plans to file on November 2, typically carries management’s discussion.

Policies in force: still growing, but more slowly

Policies in force (PIF) is the count of active policies at period end — a measure of the customer base, separate from the premium charged on each policy.

Policies in force (thousands)	Sep 30, 2026	Jun 30, 2026	Sep 30, 2025	YoY	Net adds, 3Q26	Net adds, 3Q25
Agency auto	11,347	11,211	10,630	7%	~136	~207
Direct auto	16,848	16,721	15,619	8%	~127	~374
Special lines	7,390	7,297	6,980	6%	~93	~130
Property	3,652	3,631	3,651	0%	~21	~43
Total Personal Lines	39,237	38,860	36,880	6%	~377	~754
Commercial Lines	1,252	1,226	1,198	5%	~26	~9
Total	40,489	40,086	38,078	6%	~403	~763

Net adds are our arithmetic from the period-end counts in the three releases. Agency means policies sold through independent agents; direct means policies sold by Progressive itself online, by phone or in its app. Special lines are motorcycle, boat and RV policies.

The slowdown is concentrated in direct auto, the largest channel: about 127,000 net new policies in the quarter against about 374,000 a year earlier. Year-over-year growth in direct auto PIF eased to 8% from 10% at June 30, and total PIF growth to 6% from 7%. Commercial Lines was the one segment adding more policies than a year ago.

Premium growth slowed in step. Companywide NPW growth was 4% in the third quarter after 5% in the second, and only 2% in September alone, when Personal Lines NPW grew 1%. By channel, third-quarter NPW (derived, our arithmetic) rose about 2.3% in agency auto, 4.7% in direct auto, 5.2% in Property and 8.9% in Commercial Lines.

Personal Lines vs Commercial Lines

The release does not give quarterly segment ratios, but they can be approximated from the year-to-June and year-to-September tables. Because each input is rounded to a tenth of a point, these quarterly figures carry an error of roughly a quarter of a point (our arithmetic):

• Personal Lines — about $19.9 billion of NPW in the quarter, about 3.7% more than a year earlier — ran a combined ratio of about 88.0, with a loss/LAE ratio of about 67.5 against about 63.6 a year ago. The year-ago Personal Lines combined ratio is not comparable because the Florida credit sat inside it.
• Commercial Lines — mostly small-business auto, trucking and related coverage, about $2.4 billion of NPW — ran about 92.6 against about 89.2 a year earlier, with the loss/LAE ratio up to about 71.9 from about 67.5. The Florida credit did not touch this segment, so this is a like-for-like weakening.
• Property — homeowners and renters — was the best-performing book at about 72.8, against about 63.5 a year earlier, as quarterly catastrophe losses in Property rose to roughly 8 points from roughly 1 point. For the year to date, the Property combined ratio is 76.3 against 78.0, but the June release says Property’s ratios include a one-time favorable 11.7-point impact from an actuarial methodology process change, so the year-to-date comparison flatters the underlying trend. The quarterly figures above are unaffected, because the adjustment sits in both year-to-date totals and cancels.

For the first nine months, the companywide combined ratio is 87.5 against 87.3 a year earlier — a year that included the Florida charge.

Investments, unrealized losses and capital

Investment income was about $1,035 million in the quarter, up about 12% from about $924 million (derived, our arithmetic), and the pretax annualized book yield on the portfolio was 4.3% in September. The fixed-income portfolio has a duration of 3.6 — a measure of how much its value moves when interest rates change — and a weighted average credit quality of AA-.

The swing in realized results was the main reason net income fell: pretax net realized losses on securities were $170 million against $295 million of gains a year earlier. September alone carried $231 million of losses, mainly holding-period losses — changes in the market value of securities still held — partly offset by gains earlier in the quarter (our arithmetic).

Unrealized marks moved more. The net unrealized pretax position on fixed-maturity securities was a loss of $3,455 million at September 30, against a $1,053 million loss at June 30 and a $249 million gain a year earlier; it fell $1,656 million in September alone. On our arithmetic, third-quarter comprehensive income — net income plus the change in those marks — was about $645 million, a quarter of net income. The total portfolio’s fully taxable equivalent return was -1.5% for September and -0.3% for the year to date.

Book value per share was $59.51, up from $59.05 at June 30 and down from $60.45 a year earlier. Trailing 12-month return on average shareholders’ equity was 34.4% on net income. Debt was $8,389 million and the debt-to-total-capital ratio 19.6%, unchanged from June and up from 16.3% a year ago. Common shares outstanding fell to 579.5 million from 581.4 million at June 30; the company repurchased 969,262 shares in September.

Progressive’s dividend policy, per its 2025 Form 10-K, is a quarterly dividend plus consideration of a variable dividend at least annually. The September release says nothing about a variable dividend for 2026.

What to watch in Q4

• The loss/LAE ratio. About 68.2 in the third quarter against about 64.5 a year earlier, with about 3 points of accident-year deterioration not explained by catastrophes on our arithmetic. Whether that persists is the central question for the next three monthly releases.
• Direct auto policy growth. Net adds of about 127,000 in the quarter were about a third of last year’s pace. October results are scheduled for November 18, before the market opens.
• The calendar-year combined ratio against the 96 goal. At 87.5 for nine months there is wide room, but the company’s goal is to grow “as fast as we can, at or below a companywide 96 combined ratio on a calendar-year basis,” per its 10-K — so on our reading, how much margin it chooses to spend on growth is a strategic choice rather than a constraint.
• Commercial Lines. About 92.6 for the quarter and unaffected by the Florida comparison.
• Unrealized losses on bonds. A $3,455 million unrealized loss at September 30 does not hit net income, but it does reduce book value.
• The Form 10-Q (planned November 2) and the November 3 conference call, which may explain the loss-ratio move that the release leaves uncommented.

Quick answers

How much did Progressive earn in Q3 2026?

Net income was $2,542 million, or $4.37 per share, against $2,615 million and $4.45 a year earlier, on net premiums written of $22,301 million, up 4%.

What was Progressive’s combined ratio in Q3 2026?

88.7, against 89.5 a year earlier. The year-ago quarter included a $950 million Florida policyholder credit expense; excluding it, the year-ago ratio would have been about 84.9 on our arithmetic.

How many policies does Progressive have in force?

40.5 million at September 30, 2026 (40,489 thousand), up 6% from 38.1 million a year earlier, including 16.8 million direct auto and 11.3 million agency auto policies.

---

Sources and method. Current-period company figures — the quarterly table (net premiums written and earned, net income, per-share earnings, realized gains and losses, combined ratio, average diluted shares), policies in force, the September and year-to-date income statements, segment premiums and ratios, catastrophe loss ratios, reserve development, investment results, balance sheet items, book value, return on equity, debt-to-capital and September share repurchases — are Progressive’s own disclosures in Exhibit 99 to the Form 8-K filed 2026-10-14 (accession 0000080661-26-000367) for the month and quarter ended 2026-09-30, read directly from SEC EDGAR and linked above. Second-quarter and June 30 figures and the year-to-June 2026 totals come from the June release (8-K filed 2026-07-15); the year-ago September figures and the $950 million Florida policyholder credit disclosure from the September 2025 release (8-K filed 2025-10-15); the year-to-June 2025 totals from the June 2025 release (8-K filed 2025-07-16). The 96 combined ratio goal and the dividend policy are quoted from the 2025 Form 10-K. The second-quarter column, and Property’s year-to-date ratios, include the one-time favorable actuarial methodology adjustment disclosed in the June release. The release contains no executive quotations; the company states it has no additional commentary on September’s results. The “before the open” timing reflects the August release’s stated schedule and the filing’s 9:00 a.m. ET acceptance time.

No consensus estimate is included. Whether the quarter beat or missed is judged only against the year-ago quarter, because Progressive issued no quarterly guidance (its 2025 Form 10-K states that it does not provide earnings estimates to the market). This piece makes no claim about the share-price reaction.

Our arithmetic, labelled as such: all third-quarter figures not in the release’s quarterly table — investment income, pretax income with and without realized results, losses and LAE, the loss/LAE and expense ratios, prior-year development, the accident-year loss ratio, the catastrophe split and the segment combined and loss ratios — are year-to-September minus year-to-June figures from the two releases, and ratios rebuilt from rounded published ratios carry an error of roughly a quarter of a point. The expense ratio treats fees and other revenues as an offset to underwriting expenses because that treatment reproduces the reported 88.7 and 89.5; it is our reconstruction, not a company definition. Also ours: the ~4.6-point effect of the Florida credit and the ~84.9 year-ago ratio excluding it, the ~3.7-point loss-ratio increase, the 11% rise in losses, policy net adds, segment NPW growth, and third-quarter comprehensive income. The interpretation — that the headline improvement in the combined ratio masks weaker underwriting, and that the slowdown is concentrated in direct auto — is our analysis, not a company statement, and is not a forecast.

Related: Bank of America Q3 2026, Morgan Stanley Q3 2026 and BlackRock Q3 2026, the other reports we covered the same day; and UnitedHealth Q3 2026, another insurer, a day earlier.

Tradune shows what you actually hold across every account, and whether a position has beaten a plain index fund since you bought it. Educational analysis, not investment advice.