CHWY: Higher Profit, Less Free Cash — The Cost of Expansion

The 60-second read

Five things to know

1. Operations improved. Gross margin stayed 30.4% while operating profit increased.
2. Cash did not keep pace. Capex rose and quarterly FCF fell to $89.5M.
3. The settlement is separated. $24M legal proceeds benefit GAAP income and are removed from adjusted EBITDA.
4. The balance sheet changed. Acquisitions and buybacks accompany $600M new principal debt.
5. Price requires conversion. Owner-cash DCF needs 23.4% five-year growth to justify $20.75 under our discount assumptions.

Three scenarios, one probability-weighted number

Analyst scenarios for the coming 12 months; rounded values, not predictions.

Signposts: what would change our mind

Analyst monitoring dates, not confirmed earnings dates. Reassess the call on ownercash plus growth; do not upgrade on adjusted EBITDA alone.

The selloff asks a cash question

Chewy raised its full-year outlook and the stock still fell. The September 9 close was $20.75, down 10.83% from $23.27. That combination is not proof that the market misread the release. It tells us the quarter must be judged against the expectations embedded in the price, the cash investors can ultimately keep, and the cost of the expansion now under way. We cannot assign the day's decline to a single cause from a price series. Our cash-quality explanation is an analytical interpretation, not a statement about every shareholder's motive.

The business has improved across the longer record. Trailing revenue is $13,068.7 million, trailing adjusted EBITDA is $823.0 million, and reported free cash flow is $568.1 million. Yet the economic claim belongs to common shareholders after compensation, capital needs and financing. Deducting trailing stock compensation from reported free cash flow leaves an illustrative owner-cash proxy of $270.2 million. That is a deliberately demanding lens, and the distinction explains why a low-looking EBITDA multiple can coexist with a modest cash yield.

The saved FMP quote places the stock inside a $17.40–$40.55 trailing-year range with beta 1.424. Those are observations about market behavior, not valuation anchors. A previous high is not a promise of recovery. The daily price endpoint reports different opening data from the quote endpoint, so this packet uses the matching closes and explicitly excludes the conflicting opens. Nine earnings-day moves are computed from the prior trading close to the filing-day close; they illustrate volatility rather than forecast the next reaction. Historical prices are not adjusted here for a hypothetical investor's taxes or trading costs.

Primary source

Growth improved; conversion became harder

Revenue increased 7.28% to $3,330.2 million. Management separately reports 5.7% growth excluding SmartPak and Modern Animal. The reported increase is therefore a combination of the established platform and acquired activity. It would be misleading to call all of it organic, and equally misleading to remove the acquired sales while leaving their costs in a margin comparison. Our charts use consolidated reported revenue; the acquisition-excluded growth rate is labeled separately because the release does not provide a complete standalone adjusted income statement for the acquired businesses.

The prior quarter's investor deck guided second-quarter sales to $3.30–$3.33 billion and adjusted diluted earnings of approximately $0.36. Actual revenue reached the upper end after rounding and adjusted diluted EPS was $0.36. Management's current full-year revenue range is $13.46–$13.57 billion, against $13.40–$13.55 billion in that prior deck. The EBITDA-margin range moved from 6.6%–6.8% to 6.7%–6.8%. Those changes lift the floor more than the ceiling. They support improved execution without suggesting an unrestricted acceleration in demand.

Gross profit rose to $1,011.2 million while the gross margin stayed at the rounded 30.4% level. Operating income increased from $69.7 million to $92.0 million. Adjusted EBITDA grew considerably faster than sales, but adjusted net income grew only 5.46%. Quarterly free cash flow then moved in the other direction. The central question is where that apparent momentum goes between the operating result and cash left for shareholders. We separate revenue, operating profit, adjusted earnings and cash because each measures a different economic step. None should be used as a shortcut for the others.

Primary source

A repeat-customer engine, not a subscription annuity

Chewy reports one operating and reportable segment. The chief operating decision maker uses consolidated gross profit and net income, so this packet does not manufacture pharmacy, veterinary or Autoship segment EBITDA. Product categories provide useful sales detail, but their economics cannot be isolated from the available disclosures. Consumables remain the biggest category. Pet health and specialty products, hardgoods, and other activity expand the range of customer needs the platform serves; that variety is useful only if customer acquisition and service costs leave an acceptable return.

The second-quarter category table reports consumables of $2,233.4 million, hardgoods of $395.8 million, pet health and specialty products of $578.9 million, and other sales of $122.1 million. The corresponding year-earlier amounts were $2,149.4 million, $346.1 million, $507.6 million and $101.1 million. Category sales add exactly to consolidated revenue. That is a useful reconciliation check. It is not permission to apply a high software multiple to health-related revenue: the company sells products and delivers services with fulfillment, staffing and inventory requirements.

Active customers reached 21.705 million, compared with 20.906 million a year earlier. The current definition includes approximately 43,000 SmartPak customers but excludes additions from Modern Animal. Sequential adds were 208,000. Net sales per active customer reached $602 and is a trailing-four-quarter measure divided by the ending customer count. It is not the quarter's average ticket. Autoship customer sales reached $2,817.2 million, or 84.6% of total sales. This counts sales to customers who used Autoship during the preceding 364 days, including their other orders. It does not mean that 84.6% of revenue is contractually recurring or generated solely by scheduled subscriptions. The behavior is attractive; the contractual certainty is much lower than that phrase would imply.

Primary source

The settlement and the financing bill

The quarter's $80.5 million GAAP net income includes $24.0 million of net legal settlement proceeds. That is a material positive item, but it is not concealed inside management's adjusted EBITDA: the company deducts the settlement in its reconciliation. Our concern is not that the entire EBITDA improvement comes from the settlement. It does not. The concern is that a viewer comparing only net income growth can mistake a non-operating receipt for repeatable earning power. Operating income is the cleaner starting point for identifying progress in the retail and services platform.

The bridge starts with $92.0 million of operating income, adds $19.9 million of combined interest and other income, deducts $31.4 million of income-tax provision and arrives at $80.5 million. The detailed discussion splits the combined line into $4.7 million of net interest expense and $24.6 million of other income. The new debt has therefore changed financing from a small benefit into a cost. A settlement can outweigh that cost in one quarter without making the financing burden disappear next quarter. The tax rate was 28.1%, computed from provision divided by pretax income; we do not assume the settlement's exact after-tax effect without a disclosed allocation.

Management's adjusted EBITDA reconciliation adds depreciation and amortization of $39.7 million, share compensation and related taxes of $85.9 million, the tax provision, net interest expense, transaction costs of $6.4 million and other adjustments of $2.1 million, then removes the legal proceeds. It reaches $226.7 million. The full bridge balances in the calculation file. Adjusted net income uses a different definition and excludes transaction costs prospectively starting in fiscal 2026. Consequently, a year-over-year adjusted EPS comparison contains a definition change as well as operating performance. This packet flags it instead of silently presenting the adjusted series as perfectly comparable.

Primary source

What is left after paying people?

Share compensation is the largest recurring difference between headline adjusted profitability and the result common owners can keep without dilution. Quarterly expense excluding payroll taxes was $83.5 million, derived from the $150.4 million first-half amount less $66.9 million in the first quarter. Including related taxes, management's EBITDA add-back is $85.9 million, or 2.58% of sales. These are different definitions and must not be interchanged. Deducting compensation from free cash flow is an analytical replacement-cost proxy: it asks what remains if employees were compensated without issuing additional equity. It is not a second GAAP cash-flow statement.

Trailing reported free cash flow was $568.1 million. Trailing stock compensation was $297.9 million. The resulting owner-cash proxy was $270.2 million, which yields 3.18% on the diluted equity value at $20.75. Reported free cash flow yields 6.68%. Both numbers are reproducible; the second is not automatically superior just because it is larger. A shareholder who accepts SBC without an economic deduction must instead model future dilution or the cash cost of buybacks that offset it. We choose the explicit deduction and keep the share count stable in the DCF.

Receivables and inventory do not show an obvious accounting emergency. Quarter-end receivables divided by quarterly revenue implies 6.3 days, while inventory divided by quarterly cost of goods sold implies 36.3 days, using a 91-day quarter. These are endpoint approximations rather than average-balance turnover metrics. Acquired balance sheets and seasonal stocking can distort them. The unusually large fiscal 2024 tax benefit also makes a raw multi-year net-income chart look better than the underlying operating trajectory. A normalized assessment must distinguish taxes, working capital, compensation and recurring operating costs instead of awarding a single quality score merely because earnings are positive.

Primary source

The cash cushion has almost disappeared

Chewy ended the quarter with $611.0 million of cash and $1.2 million of marketable securities. Against that sits a new $600.0 million term-loan principal. Net cash on a principal basis is therefore only $12.2 million, or about $0.03 per diluted share. The balance sheet carries the loan at $591.7 million after unamortized issuance costs; using that carrying number as the repayment obligation would overstate the cash cushion. Operating lease liabilities are a separate fixed claim and are not netted into this EBITDA valuation because the earnings measure already includes lease expense. The distinction should remain consistent in any peer comparison.

The first half used $552.8 million for acquisitions, including $378.0 million for Modern Animal and $174.8 million for SmartPak, net of acquired cash. It also used $400.0 million for common-stock repurchases. First-half free cash flow was $160.3 million. This does not mean the company is unable to fund its operations, but it does show that the pace of discretionary capital allocation exceeds internally generated free cash flow over that period. Financing and the opening cash reserve absorb the difference. A buyback reduces the share count; it does not by itself prove value creation if acquisition returns and financing costs are unfavorable.

The new term loan matures in June 2033. The filing reports a weighted-average interest rate of 5.41% and an effective rate of 5.66% at quarter end. There is no near-term principal wall, but a floating-rate obligation creates an ongoing claim on future cash. Management still had $350.1 million available under the repurchase authorization. We would prefer evidence of durable cash conversion and acquisition returns before treating the remaining authorization as automatically accretive. Quarter-end equity was $370.5 million, a thin book cushion relative to the scale of sales and total liabilities. Liquidity is adequate to analyze; it is no longer a large part of the stock's fair value.

Primary source

Three routes; a modest margin for error

Our fair value is $18 per share, with a $12–$28 scenario range and High uncertainty. This is a HOLD at $20.75, conviction three out of five. The current price is above the central estimate, but the upside from successful execution and the positive recurring operating trajectory argue against a categorical sell thesis. We require a 25% margin below central fair value before a valuation-led BUY, implying $13.50 under unchanged assumptions. That threshold is an analytical discipline, not a prediction that the stock will trade there. A material improvement in owner cash can also lift the value rather than requiring a lower share price.

The first route applies 10 times the $912.3 million EBITDA implied by management's revenue and margin midpoints, then adds $12.2 million net cash and divides by 410.1 million diluted shares. It yields $22.27. This is the most generous route because it places a multiple on earnings before recurring compensation. The second route charges $320 million of assumed annual SBC and related taxes, $160 million depreciation and amortization, $25 million net interest and a 25% tax rate against the same EBITDA midpoint. Applying 23 times that normalized earnings stream yields $17.13. These forward costs and multiples are our assumptions, explicitly separated from management guidance.

The third route starts with $270.2 million of trailing free cash flow after SBC, grows that proxy 10% annually for five years, discounts at 10%, and uses 2.5% terminal growth. It produces $12.33. The terminal calculation assumes the compensation deduction adequately offsets future dilution and that acquisition spending is not continuously required to sustain the forecast. We blend 50% normalized earnings, 30% EBITDA and 20% owner DCF, giving $17.71, rounded to $18. At today's price, the same owner-cash DCF needs approximately 23.4% annual growth for five years. That is much more demanding than our 10% base assumption. The market is already paying for substantial conversion progress.

Primary source

Fresh views deserve more weight than stale targets

The saved FMP news feed includes an Evercore ISI downgrade to In Line from Outperform on September 9 with a $25 target. That is a fresh response on the print date. By contrast, Mizuho's $32 target dated September 4 and Morgan Stanley's $37 target dated August 26 predate these results. Goldman Sachs' $34 target dated June 12 belongs to the prior earnings cycle. We show the dates because a collection of targets is not automatically a current consensus. Our $18 central value is below these examples because we place an explicit economic cost on recurring share compensation and require more cash evidence from the acquired businesses.

The separate ratings endpoint stops before the current print in the saved snapshot. It therefore cannot be described as the distribution of post-earnings recommendations. We present the latest available rating by firm as an on-record snapshot and keep the fresh Evercore news separate. The feeds have different refresh schedules and taxonomies; they should not be merged by silently labeling every historical Outperform a new Buy. A high target can coexist with a reduced recommendation if the analyst sees increased risk or fewer near-term catalysts.

Annual consensus estimates in the saved dataset carry fiscal-end dates and analyst counts. Their revenue expectations can be compared with management's revenue range, but their EBITDA field is not directly comparable with Chewy's adjusted EBITDA guidance: the normalization basis is not supplied. We therefore avoid presenting a spurious beat based on dissimilar definitions. The normalized earnings route in this packet is our model rather than the Street's EPS estimate. Similarly, price-target upside is arithmetic from a chosen target and current price, not a probability-weighted return. The investor's job is to compare the assumptions and the freshness of the information, not count how many firms prefer a positive adjective.

Primary source

Execution earns credit; allocation still needs proof

Management deserves credit for reaching the upper end of the sales guide and improving operating profitability without needing a gross-margin breakout. The public language emphasizes customer durability, continued customer growth and disciplined execution. Those claims have evidence in the operating results, but they remain claims about a developing business. A higher full-year outlook does not settle the return on capital from the two acquired businesses. We separate the operating score from the allocation score so that good retail execution does not automatically validate every purchase or repurchase.

The most consequential capital-allocation decision is the combination of acquisitions, repurchases and a new term loan. The quarter's buybacks retired approximately 9.9 million Class A shares for about $199.9 million excluding commissions and excise taxes. That price may prove attractive if the acquired capabilities compound owner earnings; the share reduction alone cannot answer the question. Modern Animal expands the service footprint and SmartPak adds an equine business. Their strategic fit is plausible. Their standalone margins, incremental returns and integration milestones are insufficiently disclosed to award a premium in this packet.

Governance also belongs in the investment case. Class A has one vote per share and Class B has ten. The company identifies BC Partners affiliates as controlling stockholders and uses controlled-company governance exemptions. Both common classes share economic distributions on a per-share basis, but an ordinary public holder's influence does not track economic ownership. The vendor float estimate is not a substitute for voting-control analysis. Deloitte & Touche has audited the company since 2017, and these are US financial statements rather than an ADS or VIE structure. None of those features removes execution risk; they explain the rights and accounting perimeter an investor is actually buying.

Primary source

1. What are the acquisition-excluded customer acquisition costs and repeat-purchase economics?
2. When will Modern Animal generate positive incremental cash after clinic investment?
3. How much of annual capital spending is maintenance versus expansion?
4. What recurring SBC dilution or cash replacement cost should owners model?
5. Which owner-return threshold governs the next repurchase versus debt repayment?

The bear case must survive a real bull case

The highest-ranked risk is paying for adjusted profit that takes too long to become owner cash. Compensation remains an economic expense even when it does not consume cash in the period it is recognized. Rising investment needs can absorb the rest. This risk has high impact because valuation is sensitive to the level and growth rate of the cash proxy. It has medium-to-high likelihood because the current gap is observable, although the company can improve it through growth and operating leverage. The risk is not that all adjusted measures are worthless; it is that investors capitalize them without paying attention to the claims they exclude.

Acquisition integration comes next. Veterinary services, pharmacy capabilities and an equine business broaden customer relevance but add operating complexity, staff requirements and potential investment demands. A third risk is competition: price transparency and fulfillment expectations constrain how much margin can be earned from a repeat-purchase relationship. Consumables' importance helps demand resilience but does not guarantee pricing power. The annual report also says products from the three largest vendors represent approximately 39% of sales in each of the last three fiscal years. That creates dependence on supplier relationships even though no single pet parent drives revenue.

The strongest bull case is that recurring customer behavior and growing health activity allow revenue to compound while infrastructure costs grow more slowly. If share compensation falls as a percentage of sales, acquisitions mature and owner cash rises, the business may deserve the upper valuation route or better. Our model does not dismiss that outcome: it gives the bull scenario a meaningful probability. The strongest bear case is a service expansion that requires ongoing capital while compensation and financing absorb the profit improvement. We land between them because the operating record is improving, but the current share price already requires a faster owner-cash trajectory than our conservative DCF. Price alone does not resolve that tension.

Primary source

A scorecard for the next filing

The next quarter should be judged against a small number of numeric tests rather than whether management repeats encouraging language. The current third-quarter guide calls for $3.323–$3.358 billion of sales and a 6.6%–6.7% adjusted EBITDA margin. We will compare results with that range, then inspect the acquisition-excluded growth rate and the owner-cash bridge. A sales beat accompanied by weaker cash conversion is a mixed result. A modest sales result with a sustained reduction in compensation intensity and healthy repeat purchasing can be more valuable than it first appears.

The research checkpoints below use December 2026 and the fiscal year-end review in March 2027 as monitoring windows, not company-confirmed future earnings dates. The calendar will be updated when Chewy announces its release schedule. The latest loan, acquisition and repurchase figures are already available in the second-quarter filing, so there is no need to wait for a transcript to acknowledge them. This packet is based on saved releases, investor decks and statutory filings. It does not attribute additional statements to a call transcript that was not part of the evidence set.

Our key upgrade condition is trailing owner-cash proxy above $325 million alongside acquisition-excluded growth of at least 6%, with the cash cushion stable. Those are analyst thresholds. They would indicate the platform is creating more cash per share, not simply reporting a larger adjusted metric. A downgrade becomes more likely if net debt exceeds $200 million while that cash proxy falls below $250 million, or if gross margin stays below 30%. The thresholds are not automatic trading orders; they identify what would force a valuation refresh. There was no prior research packet with a comparable dated scorecard in the supplied episode evidence, so we start a track record here and will grade these rows next time.

Primary source

Sources, arithmetic and limits

All financial amounts in the tables are US dollars in millions except per-share figures, customer counts and expressly labeled units. Older fiscal 2024 earnings releases presented dollars and weighted shares in thousands; the extraction normalizes those to millions. Subsequent comparisons round historical amounts, so a small rounding difference between releases is not evidence of a restatement. Fiscal 2024 had 53 weeks and its fourth quarter had 14 weeks; reported historical figures are not silently normalized for that extra week. The original source tables remain saved alongside the normalized dataset. Every quarterly record carries its primary URL and raw table rows, and the calculation scripts assert revenue less cost equals gross profit, gross profit less operating expenses equals operating income, pretax less tax equals net income, and operating cash less capital expenditures equals free cash flow.

The quarterly cash-flow statement is year-to-date. We use the explicitly quarterly free-cash-flow reconciliation for quarterly operating cash and capex. Standalone SBC is derived from cumulative cash-flow adjustments within the same fiscal year. EBITDA reconciliations are taken from the current-quarter columns and include all reported adjustment lines. Chewy does not disclose separate category profit, so the segment charts correctly use the single consolidated reportable segment. The company has no reported minority-income leakage in this record. Missing or inapplicable disclosures are identified as such; they are never converted into favorable zero expenses.

The FMP quarterly income series duplicated the prior fourth quarter under fiscal 2026's first quarter. Primary documents show $3,357.2 million revenue and $94.8 million net income for that first quarter, and those replace the erroneous vendor pair. FMP's quote and historical-price endpoints also disagree on opening prices, which are not used. The quote's market capitalization implies a different share denominator from the filing, so enterprise value in our model uses 410.1 million diluted weighted shares consistently. Quarter-end issued Class A and B shares total 401.471524 million; period-end shares and average diluted shares answer different questions. The vendor float feed is retained for completeness with a conflict warning and does not determine governance conclusions.

Owner cash means free cash flow less stock compensation in this analysis. EBITDA means earnings before interest, taxes, depreciation and amortization; Chewy's adjusted version removes additional items. Enterprise value means modeled equity value plus principal debt less cash and securities. The DCF terminal value is a continuing-business assumption, not cash sitting in a bank. Our probability weights, discount rate, growth assumptions, multiples and margin-of-safety requirement are analytical judgments and can be wrong. Financial statements establish the starting facts; they do not establish our fair value. Source snapshots, calculation files and disclosure make the distinction reviewable.

Primary source