Frontdoor sells protection against the most expensive surprise a homeowner can face, and roughly three of every four revenue collections arrive from customers who already signed up a year ago.
The most important recent development is the second quarter of 2026, in which the installed base finally turned positive again. The count of active home warranties rose for the first time in four years, and the real estate channel, tied to home closings, kept adding first-year policies while renewal pricing actions lifted the value of the contracts that came back. Growth remains thin, but it reverses a two year erosion of the base that had been the single largest threat to the earnings stream.
The central tension is that the company is shrinking its customer base even as it grows revenue and profit, and the buyback is now doing more work than new sales. Management repurchased well over a million shares in 2025 and another large tranche in the first half of 2026 at prices far above the prior year average, and the shrinking share count is the reason earnings per share climbed even though the number of customers barely moved. If retention slips below the high seventies, the buyback math stops covering the gap.
The timing trigger is the annual impairment test that runs each October, alongside the fall pricing cycle that sets renewal prices for the following year. A clean pass, with the large goodwill balance from the 2-10 HBW deal left intact, would close the loop on the biggest accounting question left over from the acquisition. A write down would reset the valuation debate all at once.
Frontdoor is the revenue leader in the United States home warranty category, operating under the American Home Shield, HSA, OneGuard, Landmark, and 2-10 HBW brands from Memphis, Tennessee. The core product is an annual service plan covering breakdowns of major home systems and appliances, from HVAC and plumbing to refrigerators and dishwashers, and the company settles several million service requests a year through a nationwide network of independent contractor firms. The business is deliberately not an insurance company in most states. The contracts are service agreements, which keeps the product outside insurance regulation in many jurisdictions while the home warranty brands book claims as a service cost, and only the builder warranty book and the Texas captive carry insurance accounting.
The strategic structure is a two-sided platform. On the demand side, customers pay roughly $600 a year on average and file almost two claims per year, so the product is used far more often than homeowners insurance and the relationship stays active. On the supply side, independent contractors get a steady flow of work orders and Frontdoor sets quality standards, monitors performance, and routes roughly 84 percent of requests to a preferred tier of about 4,200 firms. That routing matters because preferred contractors are cheaper and correlate with higher retention, so the network is both a cost lever and a quality moat at the same time.
The growth strategy of the past two years was to buy an adjacent book of business. Two years ago, Frontdoor bought the builder warranty leader 2-10 HBW, and the settled price sat in the high hundreds of millions, well below a billion. The deal added a meaningful block of homeowner contracts plus a builder warranty business in which, through most of the following year, roughly one in six new homes in the country carried its protection. The strategic logic was to own the point at which a new home first enters the home warranty funnel, before the builder hands the homeowner over to the resale market. It also added a genuinely different revenue line. Builder warranties are underwritten and reinsured and tied to new home starts, which diversifies the top line away from the resale channel that has been weak.
The balance sheet was refinanced on the same day. The new credit agreement stacks a short tenor term loan, a longer tenor term loan, and a quarter billion dollar revolver, for total debt just over $1.1 billion. Against roughly three quarters of a billion in cash and securities, net debt sits under one times the trailing adjusted EBITDA. The capital allocation is a buyback rather than a dividend. A two year authorization approved in 2024 still has several hundred million left at the end of 2025, and the company is funding repurchases out of free cash flow rather than new issuance.
The product shelf is wider than the average investor pictures. The base home warranty covers repair or replacement of major systems and appliances under normal wear and tear, with optional add ons for pools and spas, and customers can choose service levels that change the service fee and the quality tier of contractor assigned. Around that base sit the non warranty offerings. The New HVAC upgrade program discounts full system replacements and routes the work to the same contractor network. The Moen Flow partnership has licensed plumbing contractors install smart water shut off devices, a category where homeowner insurance discounts and insurer support drive adoption. Home maintenance contracts, rekeying, and a video based Virtual Experts service in the American Home Shield app round out the engagement layer, and the builder warranty book from 2-10 HBW adds structural, systems, and workmanship protection sold to builders rather than homeowners.
The real moat is not any single product but the claim clearinghouse. Several million work orders a year, spread across every trade and every state, give Frontdoor a dataset on failure rates, repair costs, and contractor performance that no regional competitor can match. That data feeds dynamic pricing algorithms that set plan prices by risk, which is why renewal revenue can grow through price even when the customer count is flat. The contractor network is the second half of the moat. Roughly 17,000 independent firms rely on the platform for volume, and the preferred tier of a few thousand top performers handles the large majority of requests at lower cost. A new entrant would have to build both the data and the network before its pricing engine or its service quality could compete, and neither is buildable in a single year.
The technology layer is improving but is not a standalone franchise. The customer app handles claims, video diagnosis, and account management, and the contractor platform manages dispatch and quality scoring. These tools cut the cost to serve and support the non warranty upsell, but they are not a product other companies would pay to license. The honest read is that the platform is an efficiency engine inside a service business, not a software business in its own right. The durable advantages are scale in claims, scale in contractors, and brand trust that has been built over decades, and those three things compound slowly but they compound.
The 2-10 HBW add on changes the moat profile in one specific way. Builder warranties are underwritten with real insurance accounting, and a large portion of the structural risk is ceded to reinsurers, so the book behaves more like a reinsurance platform than a service contract book. That means Frontdoor now holds two distinct risk books. One is priced on actuarial claims experience and the other on engineering review of builder members plus reinsurance capacity. The underwriting discipline has to be different for each, and the integration of the two books is an operating test that has no clean precedent in the company history.
Full year 2025 was the first full year with the new business in the group. Revenue reached $2,093 million, up 14 percent. Net income hit a record level near $255 million. Adjusted EBITDA reached $553 million at a 26 percent margin. The mix shift is the story inside those numbers. Renewals, which are the recurring core, contributed the large majority of revenue, and the renewal line itself grew double digits on pricing while the count of renewed contracts actually fell. First year sales from the real estate and direct to consumer channels were roughly flat, and the non warranty and other line, which now includes the builder warranty business, nearly doubled off a small base. The practical consequence is that revenue growth is coming from price and from a new business, not from a growing installed base.
Cost discipline held. Cost of services rendered rose more slowly than revenue, so gross margin expanded from about 54 percent to 55 percent, helped by favorable weather that cut HVAC claim frequency by a few million of cost, a lower number of service requests per customer, and process improvements in the contractor network. Contract claims costs were essentially flat in absolute terms even as the book grew, which is the single most important operating statistic in the report. Selling and administrative expenses grew slower than revenue as well, absorbing the integration costs of the acquisition and the 2024 brand relaunch without dragging the operating line. Depreciation and amortization roughly doubled on the acquired intangibles, and interest expense nearly doubled on the acquisition debt, which is why net income grew more slowly than EBITDA.
The first half of 2026 shows the model holding at a lower growth rate. Revenue came in at $1,096 million, up 5 percent year over year. Adjusted EBITDA reached $324 million at a 30 percent margin. Net income rose double digits to $167 million. The second quarter alone was the strongest print of the year, with revenue of $645 million and an EBITDA margin in the mid thirties. The installed base, the metric that had been shrinking for two years, finally turned positive at roughly 2.1 million warranties, with retention steady in the high seventies. Renewal revenue grew on price, the real estate channel added a small number of first year policies, and the non warranty line, led by the HVAC upgrade program, grew in the high teens to twenties.
Cash generation is the quiet strength of the balance sheet. Operating cash flow of $245 million in the first half of 2026, against modest capital spending, produced free cash flow that comfortably covered the buyback and the scheduled debt amortization. Cash on the balance sheet rose to $627 million from $566 million a quarter earlier, and the revolver was untouched. Debt is being paid down slowly through scheduled amortization, with the bulk of principal maturing in the 2029 to 2031 window. The company is, in other words, generating cash at a level that supports the buyback, the debt service, and a stable operating base at the same time, and the margin between those three uses is what funds the valuation.
The forward picture hinges on three named events already in motion rather than on new announcements. The first is the August 31, 2026 departure of the Chief Accounting Officer, announced in a current report and effective mid September, with the Chief Financial Officer absorbing the principal accounting officer duties until a successor lands. The mechanism is straightforward. The controller role sits at the center of the claims accrual process, the goodwill testing, and the regulatory capital reporting for the insurance subsidiaries, so a transition in that seat during the impairment-test season concentrates risk in one function. The consequence for shareholders is mostly about continuity. If a successor is named quickly and the fourth quarter close runs clean, the event fades. If the vacancy stretches into year end, the quality-of-earnings question becomes a standing item on every earnings call.
The second event is the June 29, 2026 election of a new independent director, a former chief financial officer of a homebuilder, to the Audit Committee. The mechanism is a governance hedge against the same transition. Adding a director with both CFO experience and homebuilding industry fluency to the committee that oversees the claims reserves and the builder warranty underwriting signals that the board is strengthening the exact control point where the 2-10 HBW book and the warranty book meet. For shareholders the value is indirect but real. Audit committee depth is the first line of defense against a claims reserve error or an over optimistic builder underwriting book, and both of those are the two risks most likely to produce a surprise in this business.
The third event is the ongoing integration of 2-10 HBW, now a full year and a half in. The mechanism is an operating test. The company has already sold the acquired office building in Aurora, Colorado, added a $4 million purchase price true up to goodwill, and is amortizing the acquired intangibles through the income statement. The consequence is that the acquisition has moved from a headline event to a cost and margin line item. The builder warranty book is tied to new home starts, which have been flat to soft, so the growth in that line depends more on builder partner retention and pricing than on housing construction volumes. The execution risk is not that the deal was a mistake but that the combined underwriting and pricing discipline across two different risk books has to be proven over several more annual cycles before the integration can be called complete.
The thesis variables that the next two to three earnings reports have to move are the installed base, retention, renewal pricing, and the builder warranty loss ratio. The installed base has to keep growing, even if slowly, for the renewal revenue engine to compound. Retention has to hold in the high seventies, because every point of retention lost is a point of revenue that the buyback cannot replace. Renewal pricing has to keep delivering the price realization that has been driving the top line while the count is flat, and the builder warranty book has to show a stable loss ratio that confirms the underwriting and reinsurance program is working as designed.
The largest risk is the customer base itself. The count of active home warranties shrank for two straight years before finally turning positive in 2026, and the real estate channel that feeds first year policies is directly tied to home resale volumes, which remain depressed by high mortgage rates and low inventory. If the housing market stays frozen and the direct to consumer channel cannot replace the lost real estate flow, the installed base reverts to negative growth and the entire renewal revenue base erodes. The mechanism is a compounding decline. Fewer renewals mean fewer second year customers, which means fewer third year customers, and the buyback shrinks the share count but cannot add a single new contract. A retention slip below the high seventies accelerates the bleed, and the stock would reprice toward a shrinking annuity.
The second major risk is claims cost inflation, which has been the quiet margin threat through 2025 and 2026. Contractor labor, parts, and appliance replacement costs have all risen, and the company has been absorbing the inflation while holding claims costs roughly flat through process improvements and favorable weather. The favorable weather headwind in 2025 and the first half of 2026, worth a few million of cost in each period, is not a permanent tailwind. A hot summer or a harsh winter spikes HVAC claims, and the lag in claims settlement, roughly three months from incurrence, means the cost shows up in the quarter after the weather, not the quarter it hits. If inflation outpaces the process gains and the weather turns, gross margin, which has been the engine of the earnings growth, compresses.
The third risk is the 2-10 HBW builder warranty book and its underwriting discipline. The book is underwritten and reinsured, which means it carries genuine insurance risk that the home warranty service contracts do not. A structural claims development, a builder member failure, or a reinsurance capacity constraint would hit the loss ratio directly, and the goodwill balance from the deal, now nearly $1 billion, is the accounting expression of that risk. If the book underperforms its underwriting assumptions, the annual impairment test, which runs each October, becomes the mechanism through which the market discovers the overpayment, and a write down would reset the multiple the market assigns to the whole company.
The counterargument that deserves a full airing is that the stock has already priced in a lot of the good news, and the debt structure reinforces that view. Total debt of just over $1.1 billion is manageable at current cash flow levels, but it is fixed, and the buyback is the one offset to the flat customer count. The company has locked a large portion of the floating rate debt into a fixed rate through a swap arrangement, which protects against rising rates but caps the benefit if rates fall. The risk is not a refinancing event, with the next major maturities in the 2029 to 2031 window, but a sustained period in which free cash flow is consumed by debt service and the buyback, leaving no buffer for a claims cost shock. The share price has nearly doubled from its level two years ago, and the market cap now sits at roughly $5.5 billion, which values the company at close to 11 times trailing adjusted EBITDA. At that multiple, the installed base turning positive, the margin expansion, and the buyback are all in the number. The bear case is that the housing market does not recover, the installed base stops growing again, retention drifts, and the stock mean reverts toward a multiple more consistent with a zero growth service business. That is the scenario in which the buyback is not enough, and the multiple, not the earnings, does the damage.
The framework starts from the cash flow the business actually produces and works back to what the multiple implies. Frontdoor generated adjusted EBITDA of $553 million in 2025. The 2026 run rate points toward roughly three quarters of a billion. First half free cash flow alone covered the buyback and the debt amortization with room to spare. Against a market cap near $5.5 billion and net debt of roughly half a billion, the enterprise value is just above six billion, trading at a little under 11 times trailing adjusted EBITDA. The earnings multiple is in the low twenties on trailing net income and closer to 18 on the forward estimate, which is high for a zero to low growth service business but defensible for one with a 30 percent EBITDA margin, a renewal revenue base, and a buyback that is actively shrinking the share count.
The bear case prices the company as a shrinking annuity. If the installed base stops growing, retention drifts toward the mid seventies, and the housing market stays frozen, the renewal revenue base erodes and the buyback, which is the only offset, gets cut. In that scenario, adjusted EBITDA growth goes to zero or negative, and the multiple reverts toward 7 or 8 times, the level at which flat growth home services businesses typically trade. The enterprise value compresses toward the mid four billions, a drawdown in the high twenties to low thirties from the current level, and the earnings per share support from the buyback disappears along with it. This is the scenario the counterargument in the risk section describes, and it is the one the stock is most exposed to if the housing data keeps pointing down.
The base case holds the current trajectory. The installed base grows at a low single digit rate, retention stays in the high seventies, renewal pricing delivers mid single digit revenue growth, and the builder warranty book stabilizes at a consistent loss ratio. Adjusted EBITDA grows at a low single digit clip, the buyback continues at a reduced pace, and the multiple holds near 10 times. The enterprise value drifts toward $6.5 billion, a modest gain from here, with the total return coming almost entirely from the shrinking share count rather than from a multiple expansion. This is the scenario the current price is priced for, and it requires nothing dramatic, just continuity.
The bull case needs two things to happen at once. The housing market eases, mortgage rates fall, and the real estate channel starts adding meaningful first year policies again, and the non warranty line, the HVAC upgrade program and the Moen partnership, grows into a real second revenue stream rather than a rounding error. In that scenario, the installed base grows at a mid single digit rate, renewal pricing keeps delivering, and the builder warranty book contributes a steady, growing, and well underwritten line. Adjusted EBITDA growth accelerates toward the high single digits, the multiple re expands into the low teens, and the enterprise value pushes toward $8 billion, a gain in the high twenties from here. The bull case is not a turnaround story, it is a re acceleration of a business that is already profitable, and that distinction is what makes it credible.
Frontdoor is a profitable, cash generative, renewal weighted service business that bought an adjacent book of business at a reasonable price and has integrated it without a visible stumble. The earnings growth is real but it is coming from price and from the acquisition, not from a growing customer base, and that distinction is the entire thesis in one sentence. The installed base finally turned positive in 2026, which removes the most immediate threat, but the growth is thin enough that any slip back into negative territory would re open the question.
The accounting question from the 2-10 HBW deal is the largest single overhang, and the annual impairment test that runs each October is the event that resolves it. A clean pass, with the nearly billion dollar goodwill balance left intact, confirms that the price paid is supported by the cash flow the book is producing. A write down would not change the cash flow but would reset the multiple and signal that the underwriting assumptions were too optimistic. The departure of the controller in the middle of the test season adds a layer of execution risk to an event that is already the most important disclosure of the year.
The counterargument is that the stock has already done the work. The share price has nearly doubled in two years, the multiple is in the low to mid teens on forward earnings, and the good news, the margin expansion, the base turning positive, the buyback, is all reflected in the number. The honest read is that the current price pays for a base case that requires nothing dramatic but also offers no outsized upside. The risk reward is roughly symmetric, with the downside protected by the cash flow and the buyback, and the upside capped by the multiple until the housing data actually turns.
The judgment is that Frontdoor is a hold quality business at a fair to slightly full price. The renewal machine is working, the cash flow is real, and the acquisition has not broken anything. The missing ingredient is the customer growth that would justify a higher multiple, and that ingredient depends on a housing market the company does not control. The events to watch are the installed base in each quarter, the retention rate, the builder warranty loss ratio, and the October impairment test. If the base keeps growing and the test passes, the buyback does the rest. If the base slips again, the multiple does the damage.