Lemonade (NYSE: LMND) published a report with plenty of good numbers. But one of them explains all the others, and it is not revenue.
About the Company
Lemonade is a digital insurance company founded by Israeli entrepreneurs, selling home, renters, car, life and pet insurance. Its model rests on automation: buying a policy, filing a claim and processing it all happen through an app, with minimal human contact.
And that is a claim easy to make and hard to prove. Every traditional insurer will say it invests in technology too. The question is how you measure it.
The Number That Measures It: the LAE Ratio
What the LAE ratio is, and why it is so useful
LAE stands for Loss Adjustment Expense. The ratio measures how much of the premium an insurer collects is spent on actually handling claims - investigators, adjusters, call centres, bureaucracy. It does not include the payout to the customer. It is purely the cost of dealing with the claim. And why it is so useful to an investor: regulators require every insurer to publish it. It is one of the few metrics that allows a direct comparison of operational efficiency between different insurers. The company puts it well itself: the systems that drive claims efficiency are the same systems that run sales, service and underwriting - claims handling is simply the one domain where regulators require everyone to publish a number.
And here is the trajectory:
| Period | LAE ratio |
|---|---|
| A few years ago | 13% |
| When premium crossed $1 billion | 7% |
| Current quarter | 5% (record) |
| Traditional insurers | about 9% |
The direct meaning: for every dollar of premium, Lemonade spends about half of what a traditional insurer spends on the bureaucracy of handling a claim.
And the sharpening point: the improvement is most notable in car insurance - the line where claims are inherently most complex. The LAE ratio in car stood at 7% in the quarter, with most claims initiated through the app.
What Was Reported
| Metric | Q2 2026 | Change |
|---|---|---|
| In-force premium (IFP) | $1.43 billion | +32.4% |
| Revenue | $294 million | +79% |
| Gross profit | $113 million (record) | +76% |
| Trailing-twelve-month gross profit | $404 million | +98% |
| Adjusted EBITDA loss | $(19) million | improved 54% |
| Net loss | $(43) million | roughly flat |
The company describes the premium growth as its eleventh consecutive quarter of acceleration.
Why revenue is growing faster than premium
This is a confusing point worth explaining.
In-force premium is the total annualized premium of all active policies - a measure of the size of the business. Revenue is what gets recorded in the income statement.
And the gap comes from reinsurance: an insurer transfers part of the risk - and part of the premium - to a reinsurer. The less it transfers, the more revenue it keeps.
Lemonade is in the middle of such a transition, retaining a higher share of premium. That is why revenue rose 79% while premium rose 32%.
And it cuts both ways: more revenue - but also more risk on its own balance sheet. A bad year for losses looks very different under this model.
The Growth Is Coming From the Customer, Not From Customer Count
| Metric | Value | Change |
|---|---|---|
| Customers | 3.31 million | +8% |
| Premium per customer | $433 | +23% |
This may be the healthiest point in the report. The company is not growing mainly by acquiring new customers - it is growing because each customer is worth more.
That happens when two things occur: the customer buys more lines (home and car and pet), and the mix tilts toward higher-premium products - mainly car.
And why that matters economically: acquiring a new insurance customer is especially expensive. Selling an additional policy to an existing customer costs almost nothing.
The Path to Profitability - and What Exactly Was Promised
The company guides to its first-ever adjusted EBITDA positive quarter - in the fourth quarter of 2026. Guidance implies roughly $8 million positive.
| Metric | Q3 2026 | Full year 2026 |
|---|---|---|
| In-force premium | $1,537-1,540 million | $1,632-1,639 million |
| Gross earned premium | $356-359 million | $1,374-1,378 million |
| Revenue | $323-326 million | $1,214-1,220 million |
| Adjusted EBITDA loss | $(23)-(20) million | $(51)-(47) million |
Two caveats worth putting on the table
First - "profitable" here means adjusted EBITDA, not net income. Adjusted EBITDA strips out depreciation, amortization, tax, financing and stock-based compensation. Stock-based compensation alone is guided at roughly $95 million for the year - and that is a real cost to a shareholder, because it dilutes them. The net loss is still far from breakeven, and the guidance does not address it. And second - the net loss looks flat, and that is misleading to the downside. It came in at $43 million, similar to last year - but the comparable quarter a year ago included a one-time $12 million tax refund. Excluding it, the underlying improvement is larger than the number shows.
And one measure that is improving consistently: net loss as a percentage of gross earned premium fell from 29% two years ago, to 17% last year, to 13% this quarter.
The Bull Thesis
Whoever reads it positively will point to the consistency: eleven consecutive quarters of accelerating premium growth, gross profit up 98% over twelve months, and an LAE ratio improving steadily - 13% to 7% to 5%.
Beyond that: the growth is high quality. It comes from higher premium per customer (+23%) rather than an expensive acquisition race.
And the positioning against the industry: 5% against 9% is a measurable efficiency gap, in a field regulators require everyone to disclose. That is not a technology promise - it is a result.
And the target is close: a first adjusted EBITDA positive quarter marked for Q4.
The Bear Thesis
Whoever reads it critically will note first that the company is still losing money - $43 million in the quarter, and $47-51 million of adjusted EBITDA loss expected for the full year. And adjusted EBITDA profitability is not profitability.
Second, risk that moved onto the balance sheet. Retaining more premium raises revenue - and raises exposure. An unusual catastrophe year looks very different under this model.
Third, the industry itself. Insurance is cyclical and depends on the frequency and severity of losses - variables outside the company's control, and increasingly affected by extreme weather.
And fourth, dilution. Stock-based compensation of roughly $95 million a year, against a company this size, is a meaningful share.
The debate in one line
The bulls see eleven consecutive quarters of acceleration, gross profit nearly doubling in a year, a measurable efficiency ratio at about half the industry's, and growth coming from value per customer. The bears see a company still losing money, underwriting risk rising with premium retention, dependence on weather, and ongoing dilution. Both sides are reading the same report.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
What catches me in this report is that it contains a number that is hard to argue with.
Technology companies in traditional industries always tell the same story: "we do this more efficiently." The problem is that you almost never get to check. The metrics differ, the definitions differ, and everyone presents what suits them.
The LAE ratio is the exception - because regulators require everyone to publish it under the same definition. 5% against 9% is not a marketing claim. It is a measurement.
And what strengthens it, to me, is car insurance specifically. Claims there are far more complex - there is a third party, potential bodily injury, an adjuster's assessment. And even there the ratio was 7%. It is easy to be efficient in renters insurance; doing it in car is a more genuine proof.
And what I hold as caution: the difference between adjusted EBITDA and profit. The company guides to a first positive quarter in Q4 - and that deserves marking. But it is on a measure that strips out roughly $95 million a year of stock-based compensation. From a shareholder's perspective, dilution is a cost - even when it is not cash.
And what I will follow is premium per customer. $433, up 23%. As long as it rises faster than customer count, the company is growing by deepening rather than acquiring - and that is cheaper growth. The day the ratio inverts, marketing spend will have to work far harder for the same growth.
Summary
Lemonade delivered a quarter of consistent growth: in-force premium of $1.43 billion (+32.4%) - acceleration for an eleventh consecutive quarter - revenue of $294 million (+79%), and record gross profit of $113 million (+76%). The adjusted EBITDA loss narrowed 54%.
And the number that explains the thesis: the LAE ratio fell to a record 5% - about half the roughly 9% typical of the industry - and 7% even in car insurance, the most complex line.
And the target that was marked: a first adjusted EBITDA positive quarter, in Q4, of roughly $8 million. With the caveat that this is profitability on an adjusted measure, not net income.
The question for the investor is not whether the technology works - the LAE ratio answers that, in a number regulators force everyone to publish. The question is how long it takes for that operational efficiency to reach the bottom line itself.
Sources: Lemonade's shareholder letter for the second quarter of 2026, as filed with the U.S. Securities and Exchange Commission on Form 8-K on July 29, 2026, including in-force premium, revenue, gross profit, GAAP loss and adjusted EBITDA, the LAE ratio, customer and premium-per-customer figures, and guidance for the third quarter and full year. Data accurate as of the time of writing. The chart is shown in real time via TradingView. Nothing herein constitutes a forecast, recommendation or advice - see the full disclaimer at the bottom of the page.