STZ: Shipments Rebound, Consumer Demand Still Trails

The 60-second read

Five things to know

1. Shipment growth outran demand. Beer shipments +5.5%, depletions -0.6%; the 6.1-point gap needs an inventory explanation.
2. EPS growth leaned on buybacks. Comparable net income $637.3M versus $637.8M; diluted shares declined about 3%.
3. Margin paid for investment. Beer margin fell 160 basis points to 39.0% despite better gross margin.
4. Cash improved through lower capex. H1 FCF rose 4.2% as operating cash flow fell 0.7%.
5. The outlook stayed steady. Comparable EPS $11.20–$11.90 and FCF $1.6–$1.7B were reaffirmed; our base value is $130.

Three scenarios, one probability-weighted number

Analyst scenarios; probabilities are judgments, not forecasts.

Signposts: what would change our mind

Thresholds are Charged Alpha decision rules. The next packet should grade them against this dated baseline.

The price and the hurdle

A dated market reference; no invented post-release reaction.

Our reference price is the October 6 pre-release quote of $115.91, captured at 3:59:59 p.m. Eastern, just before the regular session ended and before the 4:05 p.m. release. The separately saved daily-price endpoint reports a $115.67 close; we retain that series as supplied. That price is the denominator throughout this packet. We do not describe its daily gain as an earnings reaction. A separate extended-hours ranker snapshot is useful for discovery, but it is not interchangeable with this dated quote. This distinction matters because a seemingly precise upside percentage becomes misleading when the starting price moves while the fair-value estimate stays fixed.

The proposed call is HOLD, with moderate conviction and high uncertainty. A $130 base value suggests some potential appreciation, but our downside case is $85 and the bull case is $165. These are analyst scenarios, not management forecasts or market price targets. We require a twenty percent discount to the base value before upgrading on price alone, equivalent to $104. The operating evidence could justify a change before that threshold, but the change must be an improvement in consumer demand and cash durability, not simply another quarter of distributor shipments.

The stock has spent much of the saved fifteen-month price history repricing a slower beer franchise. The low multiple can therefore be either an opportunity or an accurate reflection of diminished growth. A mature consumer brand can compound value without dramatic sales growth if margins hold and cash reaches shareholders. It cannot do so indefinitely by increasing the amount sold into distribution while retailer demand remains weak. We therefore separate the attractive price of current earnings from the evidence required to treat those earnings as sustainable.

The quarter

Sales rebounded; comparable net income did not.

Constellation generated $2,633.0 million of net sales, compared with $2,481.0 million a year earlier. The approximately six percent increase is a real improvement against a difficult comparison period, and both business divisions contributed. Comparable diluted earnings per share of $3.74 exceeded the saved FMP estimate of $3.55; revenue exceeded the $2,538.8 million estimate. Those are provider consensus comparisons, not proof that every analyst used the same adjustment policy. We retain the raw forecast snapshot so future estimate revisions cannot quietly alter the apparent surprise.

The most useful distinction is between dollars earned and earnings per share. Comparable attributable net income was $637.3 million versus $637.8 million, essentially unchanged. Diluted weighted-average shares declined from 175.938 million to 170.601 million. A smaller denominator lifted comparable EPS three percent despite the flat numerator. Buybacks can create value when shares are cheap, and this is not an accusation of improper accounting. It does mean that the headline per-share improvement overstates the improvement in the underlying quarterly profit pool.

Reported EPS rose twenty-five percent to $3.32 from $2.65. That more dramatic increase is even less useful as a measure of operating momentum because the prior-year tax burden was unusually heavy. Reported pretax income actually declined. We keep GAAP and comparable results side by side, rather than choosing whichever version produces the most exciting headline. Net sales refer to sales after excise taxes; the larger gross sales figure in the statement is not an alternative revenue denominator for the margin calculations.

What customers actually bought

Shipments are not the same measure as depletions.

Beer contributed $2,473.6 million of sales and $964.2 million of segment operating profit. Its dominance means that the investment case still turns primarily on the beer franchise even when wine and spirits produces a large percentage growth rate. Beer shipments rose 5.5 percent to 123.9 million case equivalents, while depletions declined 0.6 percent. Depletions measure distributor shipments to retail customers, not direct consumer consumption. They are nevertheless closer to end demand than the company shipment measure and therefore provide an important check on the revenue improvement.

The resulting 6.1 percentage-point gap does not prove channel stuffing. Inventory positions, order timing and distributor restocking can produce legitimate differences in a single quarter. The company says an additional sell day and the Labor Day timing shift largely offset each other. Without a complete channel inventory bridge, we cannot specify how many dollars of revenue were borrowed from a future quarter. We can specify the test: shipment growth should converge toward depletion growth while retail demand turns positive. Persistent divergence would weaken the revenue-quality assessment.

Brand performance was uneven. Modelo Especial depletions declined approximately two percent and Corona Extra approximately five percent, while Pacifico, Victoria and Modelo Chelada grew approximately nineteen, fifteen and five percent. Share gains against a shrinking category are valuable competitive evidence, but they are not equivalent to absolute category growth. Marketing can support the faster brands and protect distribution, yet it also consumes current margin. We need to see whether the growing brands become large enough to offset weakness in the flagship franchises.

Wine and spirits sales rose seventeen percent to $159.4 million, and segment profit turned positive at $6.1 million. Depletions grew 10.2 percent, supporting the direction of the recovery. Its 3.8 percent margin still leaves limited earnings contribution relative to beer. The year-to-date reported decline reflects divestitures: organic half-year sales grew after removing $142 million from the comparison base. Do not mix that organic half-year calculation with the current quarter, whose comparison contains no divestiture adjustment.

The profit bridge

Tax comparisons and share count explain the headline.

Gross profit increased to $1,393.0 million, but selling, general and administrative expense rose to $538.2 million and the company recognized $49.8 million of asset impairment and related expense. Subtracting those amounts produces $805.0 million of reported operating income. Adding $7.8 million from unconsolidated investments and subtracting $83.4 million of net interest expense produces $729.4 million of pretax income. Taxes of $147.1 million and noncontrolling interest of $16.5 million then leave $565.8 million attributable to Constellation shareholders. Each step reconciles to the saved release.

The comparable operating figure adds back $91.6 million, reaching $896.6 million. That bridge includes impairment, restructuring, business reconfiguration, transition services and commodity derivative effects. Some are noncash valuation items; others are cash costs associated with reshaping the portfolio. Calling an expense noncomparable does not establish that its economic cost is zero or that another restructuring charge will never recur. We use comparable earnings as a normalized reference and then cross-check the cash available after investment, rather than automatically treating every add-back as owner income.

Beer gross margin improved, helped by lower tariff expenses and fixed-cost absorption, but operating margin fell to 39.0 percent from 40.6 percent as marketing and other selling costs absorbed the benefit. This is a deliberate reinvestment story with an unproven payoff, not an immediate operating-leverage windfall. Corporate and other costs also rose to $73.7 million from $45.6 million. Wine improvement therefore did not translate dollar for dollar into enterprise profit growth. Our valuation gives credit for durable branded economics while requiring proof that incremental spending improves depletion trends.

Earnings quality

Good cash generation, with denominator and channel caveats.

The quality scorecard is deliberately mixed. Half-year operating cash flow of $1,478.6 million exceeds attributable net income of $1,219.6 million, and stock compensation of $34.9 million is modest relative to sales. Neither observation eliminates the need to inspect working capital. The statement shows cash absorbed by receivables and other liabilities, partly offset by accounts payable. Company inventory balances declined from fiscal year-end, but these are company inventories; they do not measure the inventory held by independent beer distributors.

Our approximate receivable days divide the quarter-end receivable balance by quarterly sales and multiply by the ninety-two days in the reported quarter. Inventory days use quarter-end inventory divided by quarterly cost of product sold. These are transparent approximations, not average-balance turnover statistics, and seasonality makes a single point insufficient for declaring efficiency improved or deteriorated. We publish the inputs so readers can distinguish an accounting snapshot from a claim about collection quality. The latest formal quarter filing will provide a fuller working-capital discussion.

The most material comparability issue is the difference between a flat comparable profit pool and rising EPS. A second issue is the unusually high prior-year reported tax rate: the release shows 37.9 percent in the prior quarter versus 20.2 percent currently. We do not extrapolate the resulting GAAP earnings growth. Guidance assumes approximately twenty percent comparable tax for the full year, while reported guidance benefits from separate items. Those different tax assumptions explain why the reported annual EPS outlook can sit above the comparable outlook without implying better core operations.

Cash and financial flexibility

Capital expenditure timing supports free cash flow.

Half-year operating cash flow declined slightly to $1,478.6 million from $1,489.3 million. Free cash flow nevertheless increased to $1,124.5 million from $1,079.2 million because capital expenditure fell to $354.1 million from $410.1 million. That is a useful cash improvement, but it is driven by lower investment spending rather than faster operating cash generation. The distinction changes how confidently we extrapolate the result: deferred project spending can return in the second half, whereas a durable improvement in unit economics would be more valuable.

The company reaffirmed annual operating cash flow of $2.4–$2.5 billion, capital expenditure of approximately $800 million, and free cash flow of $1.6–$1.7 billion. Subtracting the first-half result leaves only $475.5–$575.5 million of free cash flow implied for the second half, with $445.9 million of remaining capital expenditure. Seasonality and project timing make that plausible; it is not a forecast that the first-half cash rate will simply repeat. Our valuation uses the annual midpoint, not twice the strong first-half figure.

Debt comprises $116.5 million of short-term borrowings, $1,103.0 million of current long-term maturities and $9,096.2 million of other long-term debt. Cash is only $98.5 million. We calculate net debt of $10,217.2 million and treat it as a material constraint on equity flexibility. Management reports comparable net leverage near its three-times target, but that adjusted company measure is not identical to our balance-sheet net debt divided by an independently chosen EBITDA denominator. Refinancing access and cash generation matter more than the small cash balance viewed in isolation.

First-half cash dividends of $353.8 million and treasury-share purchases of $454.7 million used $808.5 million. The headline repurchase update of $530 million runs through September and therefore covers a different period. SpikedAde adds a post-quarter $75 million payment and up to $278 million of contingent consideration over five years. The contingent amount is a maximum performance-linked obligation, not cash already spent or certain debt. It belongs in the risk assessment without being mechanically deducted as an immediate payment in every scenario.

Three valuation checks

Every route is an assumption, not an observed fact.

The first route capitalizes the $11.55 midpoint of comparable EPS guidance at twelve times earnings, producing $138.60 per share. Twelve times is our judgment for a mature branded business facing weak absolute beer demand and investment pressure. It is not presented as a mechanically verified historical average. A sustained return to depletion growth would support a higher multiple; prolonged category weakness or recurring restructurings would support less. EPS capitalization is easy to understand, but it can hide the capital required to sustain those earnings.

The second route discounts equity cash generation. We begin with the $1,650 million free-cash-flow guidance midpoint, grow it three percent annually for five years, discount at nine percent, then use a 2.5 percent terminal growth rate. The resulting equity value divided by 170.601 million diluted shares is $155.85. This cash measure is after interest, so we do not subtract net debt a second time. It is also before all distributions to noncontrolling owners and other financing uses, making the result a cross-check rather than a precise distributable-cash promise.

The third route values operations at eleven times the $2,910 million midpoint of comparable operating-income guidance. That gives $32,010 million of enterprise value. Subtracting $10,217.2 million of net debt and dividing by 170.601 million shares gives $127.74. We do not add the wine business again because its operating income is already in the consolidated denominator. Nor do we add goodwill as if it were cash. This route makes leverage explicit and exposes how a small multiple change can have a larger effect on equity value.

The routes place our central estimate near $130 after rounding and acknowledging model uncertainty. At the $115.91 reference, the stock trades at 10.04 times the comparable EPS midpoint and offers an indicative 8.3 percent FCF yield using diluted shares. The implied enterprise value is equity capitalization plus net debt; investors are paying for continued cash generation, not a cash-rich balance sheet. A low multiple alone does not guarantee mispricing. Our HOLD reflects the modest base upside relative to the bear case and the missing evidence of sustained retail-demand growth.

The sensitivity grid changes the operating-income multiple and the earnings base independently. This is more useful than reporting a fair value to the cent, because the uncertainty is economic rather than arithmetic. In the bear case, weaker beer demand and greater spending pressure reduce the earnings franchise and the market multiple. In the bull case, brand investment works, depletions turn positive and stronger cash flow supports both debt discipline and repurchases. The probability-weighted value is $127.50, below the $130 base because the downside case remains material.

The Street comparison

These targets were published before this release.

The saved FMP target-news feed shows a wide range of sell-side valuations before the current earnings announcement. RBC Capital had a $185 target dated October 2, Bernstein $180 on September 28, UBS $145 on September 23 and Barclays $122 on September 22. These are dated individual analyst targets supplied by FMP, not a fresh post-results consensus. Listing the date is essential: a target published before the shipment/depletion divergence and current expense details were disclosed has not yet incorporated a published response to those facts.

Our $130 central value sits much nearer the lower end of those selected targets than the upper end. The disagreement is not that premium beer brands lack value. It is about how much improvement in volume, investment productivity and valuation multiple should be capitalized today. An optimistic target can be reasonable if the analyst assumes a successful recovery over a longer horizon. Our approach demands evidence of that recovery before awarding the higher multiple. This allows a reader to identify the operative assumption rather than choose a target by reputation.

The saved annual estimate feed has fiscal 2027 revenue of approximately $9,087 million and EPS of $11.71, based on thirteen EPS analysts. Those figures are close to the upper end of revenue guidance and above the $11.55 company EPS midpoint. Consensus for fiscal 2028 is higher, but the feed is a current snapshot rather than a controlled before-and-after revision series. We do not call it an upgrade cycle. The next useful comparison is how estimates change after the conference call and whether they remain consistent with the stated margin and cash-flow outlook.

Management and capital allocation

Reinvestment must eventually earn its return.

Management is making a recognizable choice: spend behind brands and capabilities while preserving approximately three-times comparable net leverage and returning cash to owners. That may be sensible when a company has strong brands in a weak category. The burden of proof is that the spending improves depletion trends or strengthens future pricing power. Marketing is an expense when incurred, but its investment value can emerge over time. We therefore avoid treating every additional selling dollar as waste and avoid treating every announced investment as guaranteed growth.

The new SpikedAde acquisition expands exposure to spirit-based ready-to-drink products. The upfront payment is comparatively limited, but the maximum contingent consideration is substantially larger. Performance-linked consideration can align payment with success, yet it also makes the eventual purchase cost uncertain. The release says the transaction does not affect the fiscal 2027 outlook. We therefore include no incremental earnings benefit in our base model. Future disclosures should clarify distribution expansion, contribution margins and how success payments interact with the company’s stated capital-return priorities.

The annual report identifies Class A shares with one vote each and an immaterial Class 1 nonvoting class with conditional conversion rights. The historical Class B restructuring is complete and is not a current delisting of STZ Class A. KPMG is the auditor. Employment was approximately 9,400 at fiscal year-end, including roughly 1,100 through the equally owned glass joint venture. These facts establish the corporate structure and operating footprint, but they do not guarantee alignment or eliminate execution risk. Sands-family interests and nomination rights remain relevant governance context.

Five questions should shape the conference call: what explains the beer shipment/depletion gap; how much of the current marketing increase is structural; what evidence would justify changing the full-year demand outlook; how much brewery spending remains discretionary; and what return hurdle governs SpikedAde’s contingent payments. We will judge the answers by measurable follow-through rather than tone. The call is scheduled for October 7, so this packet does not attribute answers to management before they are given.

What can break the thesis

The strongest arguments on both sides.

The central downside risk is that premium beer demand remains weak for longer than investors expect. Market-share gains can conceal absolute volume pressure when the whole category contracts. If Modelo and Corona declines persist, growth in smaller brands may not fully compensate, and more marketing may be required just to defend the revenue base. A second risk is a reversal of the favorable shipment timing. Distributor replenishment can make one quarter look healthier without changing the final customer’s willingness to buy.

Leverage magnifies these operating risks. The company has substantial net debt, current maturities and continuing capital requirements in Mexico. Tariffs, agricultural inputs, packaging, water availability and brewery concentration can affect cash costs in ways that brand strength cannot immediately offset. The current quarter benefited from lower tariff expense, but investors should not treat that benefit as permanent. Portfolio impairments and transition costs also remind us that divesting weaker assets does not instantly eliminate their economic consequences.

The bull case deserves equal consideration. Constellation continues to gain share, owns valuable imported beer franchises, and generates meaningful cash after investment. Wine and spirits is recovering from a low base, and a lower capital-spending phase could leave more cash for owners. If depletions turn positive while marketing becomes more productive, the same share reduction that currently masks flat profit could amplify genuine profit growth. A market valuation near ten times guided earnings might then prove too pessimistic.

Our position is conditional rather than permanent. We do not need every risk to disappear before becoming constructive, but we need either a larger valuation cushion or stronger evidence that the earnings base is stabilizing. Conversely, a HOLD becomes less defensible if the company lowers cash guidance, lets leverage climb materially or continues to rely on shipments that outrun demand. The ranked table is a monitoring framework, not a statistical probability model. Impact and likelihood are qualitative judgments grounded in the business exposures disclosed in the filings.

Dated tests

The next report should grade these signposts.

The October 7 conference call is the first immediate catalyst because the release leaves the distribution and spending bridges incomplete. A persuasive explanation would quantify channel inventory and distinguish planned marketing investment from cost inflation. General statements about category leadership are useful competitive context but will not by themselves settle the investment question. We will preserve this packet’s signposts so a later review can measure whether management’s explanations were followed by better demand, margins and cash generation.

The next quarterly earnings release, whose date is not yet verified here, is the key operating checkpoint. Our thresholds are analytical decision rules, not company guidance. Positive beer depletion growth, a shipment/depletion gap narrower than three percentage points and beer margin near or above the guided full-year upper bound would improve confidence. A renewed depletion decline below negative three percent, a persistent wide gap or a cut to the $1.6 billion free-cash-flow floor would move the evidence the other way.

The dividend record date of October 30 and payment date of November 13 are verified in the release. They are corporate-calendar events, not standalone reasons to buy the shares, because a dividend transfers value out of the company. The more consequential capital-allocation milestones are progress through the remaining capital budget, debt maturity management and disclosure of acquisition-related payments. We distinguish those observable events from a prediction that the stock must rerate on any particular date.

Data, sources and method

Source identity and units are part of the analysis.

All monetary operating values in the financial tables are US dollars in millions unless explicitly marked per share. The current release’s column order is three months ended August 31, 2026, then August 31, 2025, followed by the corresponding six-month periods. The balance sheet compares August 31, 2026 with February 28, 2026. We verified those headers before transcribing values. Quarterly diluted share counts are weighted averages used for EPS; they are not the same as period-end shares outstanding or the share count embedded in a market-data provider’s capitalization.

The saved dataset contains nine quarterly provider income statements, balance sheets and cash-flow statements, plus three fiscal years of cash-flow history. Current-quarter provider classifications contain missing or inconsistently signed fields, so the filing controls the current profit bridge. Fiscal 2026 stock compensation is corrected to $67.7 million from the annual filing/company facts because the provider returned zero. Historical provider rows are retained as supplied for auditability. Nine primary releases also retain full segment, operating KPI and comparable reconciliations. Two older releases required OCR; their selected segment and EPS figures were cross-checked against following-year comparative tables. Original and subsequently divestiture-adjusted historical comparisons differ, so we identify that basis rather than splice them silently.

The price history is sourced from FMP and stored locally with the quote, analyst estimates, target news, ratings and earnings calendar. Historical reaction calculations use the next available trading close after each recorded report date. Because original announcement timing can differ, those returns should not be treated as a uniform event study or proof that earnings caused the move. The current release has no completed post-announcement regular session at this writing. Our valuation assumptions are stored separately from reported results and can be recalculated from the supplied inputs.

Free cash flow means operating cash flow less purchases of property, plant and equipment. Comparable earnings are the company’s non-GAAP definition; organic revenue adjusts the comparison perimeter for specified acquisitions or divestitures. Depletions are distributor sales to retail customers. Net debt is our sum of short-term borrowings and current and long-term debt, less cash. Fair values and scenario probabilities are analytical judgments, not guarantees. Rounding can cause displayed totals to differ slightly from source tables; calculations use the retained underlying values.

• October 6, 2026 8-K Exhibit 99.1: Q2 FY2027 results, guidance and reconciliations
• Fiscal 2026 annual report: structure, risks and annual cash flow
• Q1 FY2027 Form 10-Q
• FMP: saved quote, prices, financials, earnings estimates and target news, October 6, 2026