Dollar General FY2026 Q2 Earnings: $550M Profit on a 127 bp Margin Bridge

Dollar General just showed what a discount-retail P&L looks like when the margin bridge does more work than the top line. Net sales rose 5.2% to $11.29 billion, but net income jumped 33.8% to $550 million. Gross margin expanded 127 basis points to 32.6% — and management says roughly 81 of those basis points came from tariff refunds after related reinvestments. That is the quarter: traffic-led comps, a cost-of-goods story you can check in the ledger, and a non-recurring refund sitting inside the same Expenses:CostOfRevenue line as ordinary COGS.

The Headline Numbers​

Fiscal 2026 second quarter is the 13 weeks ended July 31, 2026, compared with the 13 weeks ended August 1, 2025 (Form 8-K exhibit 99, filed 2026-08-27):

Sales added about $563 million. Net income added about $139 million. Operating profit rose 29.2% while the SG&A rate stayed flat at 25.8% — so the leverage is not in overhead discipline as a rate; it is in gross margin and a lighter interest load.

Diluted EPS of $2.48 includes "an estimated benefit from tariff refunds, after related reinvestments, of approximately $0.25," per the release. Strip that estimate and you are looking at a still-up quarter, but not a 33% EPS print. That is the number the Bear Case will keep quoting.

Revenue Deep Dive: Traffic First, Ticket Second​

Dollar General does not publish a software-style segment P&L. The useful split is same-store sales into traffic and ticket, then category breadth.

Same-store sales rose 3.5%, with traffic +2.0% and average transaction +1.5%. That mix matters. A ticket-led print can be price or mix; a traffic-led print is more customers walking in. The release frames it as broad-based: growth "in each of the consumables, seasonal, home products, and apparel categories."

Management signal scan on demand language (verbatim from exhibit 99):

"Our results reflect continued momentum across the business, including our fifth consecutive quarter of customer traffic growth and the sixth consecutive quarter of positive comparable sales growth across all four merchandising categories."

That is the clean robust demand hit. CEO Todd Vasos also said results "exceeded our expectations even before considering the benefit from tariff refunds after related reinvestments." The company is separating operating momentum from the refund — useful when you read the margin bridge next.

What the release does not say: it does not claim demand exceeds supply, tight supply, an industry upcycle, market expansion ahead of plan, or new-product ramps outperforming expectations. For a 20,000-store value retailer, absence of supply-constraint language is expected; the demand signal that matters here is the fifth straight traffic quarter, not a capacity story.

The thesis the mix supports: Dollar General is still a traffic-and-convenience franchise. New stores contributed, store closures were a partial offset, and the same-store engine is running on customers first. That is the discount-retail counterpart to a thin-margin grocer print — volume in the door, margin in the COGS line.

The Margin Story: Bridging 127 Basis Points​

Gross profit was $3,681 million, or 32.6% of sales, versus 31.3% a year earlier — +127 bp. The release's own bridge, quoted verbatim:

"This gross profit rate increase was primarily attributable to tariff refunds, a lower LIFO provision, and lower distribution costs; partially offset by increased markdowns and increased transportation costs. The Company estimates the gross margin benefit of tariff refunds, after related reinvestments, was approximately 81 basis points."

Three constructive legs, two offsets, and a quantified refund:

The ledger does not invent separate chart accounts for "tariff refund" or "shrink." Expenses:CostOfRevenue is 7,609 MUSD, equal to filed cost of goods sold ($7,609 million). The refund lives in the source comment and in this bridge — the accountant move is reconciling COGS to the filing, then explaining the rate in prose.

Operating margin also got a refund boost: "The Company estimates the operating margin benefit of tariff refunds, after related reinvestments, was approximately 66 basis points." SG&A as a percent of sales was "essentially flat year over year, at 25.8%." So the P&L shape is: gross-rate expansion, flat SG&A rate, lower net interest ($42.9 million vs $57.7 million), higher tax dollars on higher pretax income.

Pricing language is mild, not a "sustained ASP increases" theme. Ticket +1.5% sits beside traffic +2.0%; the company is not narrating list-price power. The margin story is COGS and distribution, not selling-price harvest.

The One Big Question: How Much of This Margin Is Durable?​

The quarter's defining issue is not whether Dollar General can grow mid-single-digit sales. It did. The question is how much of the 127 bp gross-margin expansion survives once the tariff-refund window closes.

Back-of-the-envelope from the company's own estimate: 81 bp of 127 bp ≈ 64% of the gross-margin expansion is the refund after reinvestments. The residual ~46 bp is the mix of lower LIFO, lower distribution costs, and the markdown/transportation offsets. That residual is the part you can argue is operational. The 81 bp is the part you cannot annualize without a new refund.

Inventory is the balance-sheet check on the shrink/distribution narrative. Merchandise inventories were $6.55 billion at July 31, 2026, versus $6.61 billion a year earlier — down 2.7% on an average per-store basis, per the release. In the ledger that is Assets:Current:Inventories at 6553 MUSD. A retailer talking about distribution costs and damages while holding inventory roughly flat in dollars (down per store) is consistent with a throughput story, not a restocking binge.

Raised guidance reinforces management's confidence but does not make the refund recurring. The company raised fiscal 2026 net-sales growth to a 4.0%–4.3% range and same-store sales to approximately 2.5%–2.9%. Those are top-line guides. They do not promise another 81 bp refund.

Tracking a $32B Retailer in Plain Text​

Modeling Dollar General in Beancount forces every million to reconcile. Income credits are negative; expenses are positive; Equity:Adjustments absorbs net income so the income-statement transaction sums to zero — the same house conventions we use for every Open Ledger company.

Here is the FY2026Q2 income statement as booked (MUSD = millions of USD):

; Check: −11290 + 7609 + 0 + 2912 + 43 + 176 + 550 = 0 ✓
2026-07-31 * "Dollar General Corporation" "FY2026Q2 Income Statement"
  Income:Revenue                         -11290 MUSD
  Expenses:CostOfRevenue                   7609 MUSD
  Expenses:ResearchAndDevelopment          0 MUSD
  Expenses:SellingGeneralAdministrative    2912 MUSD
  Expenses:OtherNet                          43 MUSD
  Expenses:IncomeTax                        176 MUSD
  Equity:Adjustments                        550 MUSD  ; net income offset

And the inventory assertion that carries the shrink/distribution story:

2026-07-30 pad Assets:Current:Inventories                       Equity:Adjustments
2026-07-31 balance Assets:Current:Inventories                        6553 MUSD

Cost of goods sold is one number in the chart. The 81 bp tariff-refund component is named in the period-file banner, not split into a fake sub-account — because the filing does not present a separate tariff-refund P&L line you can pad to.

The balance-sheet number that tells the narrative is inventories staying near $6.6 billion while sales grow mid-single digits and gross margin expands. Cash rose to $1.59 billion; total assets were $32.17 billion. Long-term obligations were $4.56 billion against $9.29 billion of shareholders' equity.

The Multi-Year Arc​

Five fiscal years from the 10-K companyfacts series, plus this quarter:

FY2021–FY2024 compressed net margin from 7.0% to 2.8% while inventories peaked near $7.0 billion in FY2023. FY2025 started the repair (net margin 3.5%, inventories down to $6.3 billion). Q2 FY2026's 4.9% net margin and 32.6% gross margin sit above the recent annual run-rate — with the caveat that a large slice of the gross-rate jump is the refund.

The compounding story is not software-like revenue acceleration. It is a value retailer rebuilding margin after a multi-year squeeze while keeping traffic positive for five straight quarters.

The Verdict: Bull vs. Bear​

Bull Case

• Fifth consecutive quarter of customer traffic growth, with same-store sales +3.5% and all four merchandising categories positive on comps for a sixth straight quarter (release).
• Operating profit +29.2% with SG&A rate flat — operating leverage without SG&A rate cuts.
• Net interest down 25.7% to $42.9 million, adding a clean below-the-line assist.
• Inventories roughly flat in dollars and down 2.7% per store — not restocking into a margin illusion.
• Guidance raised on net sales (4.0%–4.3%) and same-store sales (~2.5%–2.9%) after the beat.

Bear Case

• Of the 127 bp gross-margin expansion, ~81 bp is tariff refunds after reinvestments — non-recurring by construction.
• Diluted EPS includes an estimated $0.25 refund benefit; without it the EPS growth story is much quieter.
• Markdowns and transportation costs are already listed as offsets; if those worsen, the non-refund residual shrinks.
• Ticket growth is only +1.5% — not evidence of sustained pricing power.
• FY2023–FY2024 showed how fast net margin can fall when the COGS line turns against you; one good refund quarter does not rewrite that history.

Our Take

Treat Q2 as two stacked stories. The first is real: traffic is compounding, comps are broad, SG&A is in control as a rate, and the inventory account is not ballooning. The second is a one-time COGS credit large enough to dominate the gross-margin headline. Own the stock or the narrative for the traffic streak; do not annualize the 81 bp. The Open Ledger entry keeps that honest — Expenses:CostOfRevenue matches filed COGS, inventories sit on their own account, and the refund is a comment you can read next to the numbers instead of a hidden chart account.