Dillard’s, Inc. (NYSE: DDS) — Best House in a Condemned Neighborhood, Priced for Zero Melt

Issuer: Dillard’s, Inc. · Ticker: NYSE: DDS (Class A common) · CIK: 0000028917 · CUSIP: 254067101 Sector: Consumer Cyclical — Department Stores · HQ: Little Rock, Arkansas (founded 1938; ~29,100 employees; Texas-incorporated since 2025-08-31) Price: ~$550.93 (close, 2026-07-17) · Market cap: ~$8.6B · Net cash: ~$0.9B · Enterprise value: ~$7.7B · Shares outstanding: ~15.62M (11.63M Class A + 3.99M Class B) Fiscal year-end: Saturday nearest January 31 (company convention used throughout: FY2025 = 52 weeks ended 2026-01-31; Q1-FY2026 = 13 weeks ended 2026-05-02) Latest primary sources: FY2025 Form 10-K (filed 2026-03-27); Q1-FY2026 Form 10-Q (filed 2026-06-05); DEFM14A (filed 2026-04-06, doubling as the 2026 annual-meeting proxy); trailing 60-month EDGAR corpus (1,091 filings) reviewed in full. Coverage tag: INITIATION — first published coverage of DDS by the author.

Independent fundamental research · Date: 2026-07-18. The analysis below (Sections 1–15) carries no BUY/SELL recommendation and no price target; Facts, Interpretations, Assumptions, and Open Questions are labeled throughout. The single labeled exception is the Claude’s Take block immediately following.

⚡ Claude’s Take

The author’s own subjective opinion, offered as general information and not investment advice. The analysis in Sections 1–15 below deliberately takes no position.

Verdict: AVOID at ~$550 — revisit in the mid-$300s-to-low-$400s — and explicitly NOT-A-SHORT. This is a “great operator, wrong price” call, and both halves are load-bearing. The operator half is real: the best-run department store in America, family-controlled, net cash, ~93% owned real estate, a genuine ~10.5% retail operating margin in a format where peers earn 0–4%, and a capital-allocation record — $1.5B of buybacks at a blended ~$236 per share — that belongs in a textbook. The wrong-price half is arithmetic: at $550.93 the market underwrites zero melt. A reverse-DCF on normalized free cash flow of $600–650M requires perpetual FCF growth of roughly +0.6% to +2.3% a year — in a channel that has shrunk ~4% a year for a decade, where Dillard’s own revenue has been flat in nominal terms for fifteen years. Every melt-path DCF lands at ~$328–436 per share. The stock sits at the 92.8th percentile of its own ten-year valuation history (P/E 90.4th, P/S 92.3rd) — richer than at any point outside the 2021 squeeze and the COVID-boom earnings spike. And the single most eloquent signal in the whole file: the Dillard family, the best-timed buyer of this stock for three decades, bought ~$500M a year at $175–255, tapered to token amounts at $360–367, and has bought zero since roughly August 2025 — the family’s own bid stopped above ~$500. The zone where the risk/reward turns constructive is the SOTP midpoint (~$430) down through the floor of the base scenario zone (~$400) and into the melt-path DCF band below it — call it the mid-$300s to low-$400s, where a buyer gets the real-estate floor and the capital-return machine without paying for a stabilization that hasn’t been proven.

Why not a short: because the same microstructure that makes the stock expensive makes it dangerous to bet against. 35.25% of the float is already short — 9.26 days to cover, at or above 2021 squeeze-era levels — in a ~7.3M-share float the family and the ESOP don’t sell into. Five consecutive special dividends ($15/$15/$20/$25/$30) make carry expensive; a sixth could land in November. The family could resume the buyback at any lower price and melt the float further. A net-cash company buying back 5–8% of its float per year can melt up for years while the business melts down; shorting a squeeze-prone cannibal with a fortress balance sheet is how you get carried out. The 2021 squeeze (~$58 to ~$181 in six months) and the H2-2025 melt-up (+58% in five months) are the receipts.

Framing: value/contrarian-in-reverse. The consensus — “great operator, full price” — is correct on both halves; the variant perception is that the market prices the melt at zero, and one quarter of positive comps (Q1-FY2026, +3%) is carrying the entire multiple. The tape agrees with caution: this is a post-melt-up momentum unwind in a thin-float value name — down 21% from the December 2025 all-time high, below falling 50- and 200-day EMAs, negative Momentum loading (−0.26) with mildly positive Quality (+0.22) and strong DividendYield (+0.69) tilts, idiosyncratic vol ~36%/yr. Conviction: medium-high that the price is full; low on timing — squeeze dynamics make the path violently non-linear in both directions. The single piece of evidence that would flip the call bullish: two more quarters of positive comps with retail gross margin holding at or above ~40%, proving the +650bps merchandise-margin step-up is structural rather than a long COVID-era lag (next test: the Q2 print, ~August 2026). The single piece that would confirm the bear: comps back negative, or retail gross margin down 100bps+ year-over-year.

“The family that bought back a third of the company at $236 won’t touch it at $550.”

📈 Stock Price Action — Five-Year Event Map

Five-year arc (FACT): the adjusted close ran from $127.78 (2021-07-19, the window edge — the stock had already tripled off its ~$45 January-2021 squeeze base) to an all-time closing high of $699.51 on 2025-12-10 (intraday $741.98 on 2025-12-11), and has since fallen ~21% to $550.93 (2026-07-17). 52-week range $440.51–$741.98 (intraday). Five-year price return ≈ +331%; trailing 12 months +30.5%; trailing 6 months −15.0%. Price sits just above the 21-day EMA ($543) but below the 50-day ($559) and 200-day ($566) EMAs, with the 50 below the 200 — a bearish cross state.

* Unadjusted closing prices (as printed on the tape); % moves computed on dividend-adjusted closes. The adjusted series differs mainly around the five $15–$30 special dividends (2021–2025).

Cycle narratives (cause = INTERPRETATION unless noted):

1. 2021 squeeze. FACT — ~$58 (2021-01-04) → ~$181 (2021-06-30) → ~$237 by 2021-11-01. INTERP — ~30%+ of a buyback-shrunk float short, a meme-era bid, and a genuine earnings recovery forcing covers; squeeze and fundamentals reinforced each other.
2. Nov 2021 whipsaw. FACT — Q3-FY21 beat (11/18) lifted the stock ~14% to $410; it lost 19.4% on 11/30. INTERP — squeeze exhaustion meeting the omicron/taper risk-off; high-short-interest names de-risked first.
3. May 2022 sympathy crash. FACT — −17.2% on 5/18/22, the day Target fell ~25% on a margin miss, six days after DDS’s own Q1 beat. INTERP — pure sector read-through, not company news.
4. Aug 2022 beat. FACT — +17.8% on the Q2-FY22 print (EPS $9.30, another >40% gross-margin quarter). INTERP — estimate beat plus mechanical short covering.
5. Feb 2023 miss. FACT — −17.1% on the Q4-FY22 print. INTERP — the market began pricing normalization off the FY2022 peak; the drawdown bottomed near $285 two months later.
6. Nov 2024 beat + election. FACT — +11.5% on 11/14/24 after Q3-FY24 (EPS $7.73, revenue beat). INTERP — print plus the post-election risk bid in domestic retail.
7. Apr 2025 tariff whipsaw. FACT — −19% peak-to-trough around “Liberation Day” (low $294.77 on 4/8), then +11.3% on the 4/9 pause. INTERP — DDS traded as a high-beta tariff proxy despite no discrete company exposure.
8. H2-2025 melt-up. FACT — +58% from 7/17/25 to the 12/10/25 all-time closing high, through two beats and the record $30 special. INTERP — earnings delivery + capital returns + short covering fed on themselves in the thin float.
9. 2026 drawdown. FACT — −21% over seven months, including −7.8% on the Q4-FY25 print (2/24/26) and −7.3% on 6/15/26 with no company news. INTERP — earnings-normalization repricing plus post-special/post-melt-up unwind; 90-day volume fell ~150k → ~111k shares/day — sellers without a crowd of buyers.

Positioning context (FACT, settlement 2026-06-30): short interest 1,084,593 shares = 35.25% of the ~7.33M-share float, 9.26 days to cover, and rising month-over-month — at or above the levels that fueled the 2021 squeeze era. Five special dividends in five years ($105/share cumulatively) sit on top of shorts’ carry cost.

1. Executive Summary

Dillard’s is the best operator in a dying format, and the market is paying full price for both facts at once. That is the whole story; everything below is evidence.

The business (FACT). ~272 stores across 30 states — Sunbelt-weighted (Texas ~57, Florida ~42) — plus dillards.com; ~46.0M square feet of retail space, ~43.0M (~93%) owned outright; a middle-to-upper-middle customer; exclusive and private brands (Antonio Melani, Gianni Bini, Daniel Cremieux, Roundtree & Yorke) at 22.3% of sales; a ~$6.6B revenue base nominally flat for fifteen years while the US department-store channel shrank from ~$232B (2000) to ~$154B (2025). The family controls the company (Class B elects two-thirds of the board; ~51% of the economics), and Dillard’s is the only major department store whose merchandising itself earns real money: a ~10.5% retail operating margin versus 0–4% at peers, because it pays almost no rent in a format where rent is the difference between profit and loss.

The industry (verdict: structurally bad, argued in Section 3). A ~35% nominal contraction over 25 years, profit pools migrated to off-price (TJX, Ross, Burlington), Amazon, discounters, specialty, and brand-direct; a dying B/C-mall substrate; no defensible middle. In Marathon’s terms the industry is in a deep capital-exit phase — but e-commerce broke the cycle’s normal completion, so the endgame is managed decline with reprieves, not recovery. The winner is whoever melts slowest with the lowest fixed costs and returns the cash. That is, almost word for word, Dillard’s playbook.

The financials (FACT). FY2025 revenue $6,563M, gross margin 40.3%, operating margin 10.5%, net income $570M, diluted EPS $36.42, FCF ~$624M, ROIC 23.4%, net cash ~$0.9B, an undrawn $800M revolver to 2030. Earnings are normalizing off an artificially high COVID-era plateau (FY2021–22 EPS $41.87/$50.95), but the gross-margin step-up is partially holding — ~650bps of the ~950bps boom-era gain retained, all of it merchandise margin, with SG&A fully round-tripped to ~27%. Q1-FY2026 printed comps +3% — the first positive comp after 0%/−3%/−4% — with retail gross margin up 30bps: the first hard evidence of a floor. Quality-of-earnings is handled explicitly: TTM reported EPS of $42.11 includes a $104.1M pretax ($5.10/share) one-time interchange-litigation settlement (ex-gain ≈ $37.0); core FY2025 EPS ≈ $33.2 after property-sale gains and special-dividend tax benefits.

The capital allocation (verdict: exemplary, argued in Section 7). Share count down 86% from the FY1996 peak (~113.8M → 15.62M); $1.508B of buybacks over five fiscal years at a blended ~$236/share versus $550.93 today; price-sensitive discipline that tapered buying to zero above ~$500; five consecutive special dividends totaling $105/share; 83% of five-year free cash flow returned, unlevered; no sale-leasebacks, no empire M&A, modest profit-linked pay. The 2026 WDC merger and Voting & Exchange Agreement extend family control to the next generation — the governance discount is the price of admission.

The tension (INTERPRETATION). The market prices zero melt: at ~$550.93 (~8.8x EV/EBITDA, 14.9x ex-gain P/E, 92.8th percentile of its own ten-year history), embedded expectations are flat-to-growing free cash flow in perpetuity — while the channel declines ~4%/yr, the last control transaction in the asset class (Nordstrom) cleared at ~4–5x EBITDA, and the family’s own buyback stopped above ~$500. The debate is not whether Dillard’s is good — it is. The debate is melt-rate versus multiple, and at this price the multiple has already conceded the melt-rate argument to the bulls. One quarter of positive comps is carrying the entire premium; the Q2 print (~August 2026) is the next test.

2. Business Overview

What the company is (FACT). Dillard’s, Inc., founded 1938, headquartered in Little Rock, Arkansas, is one of the largest surviving US fashion department-store operators: ~272 stores in 30 states (272 at Q1-FY2026, including 28 clearance centers), plus dillards.com and an owned 850,000-sq-ft internet fulfillment center. The footprint is deliberately Sunbelt-weighted — Texas (~57 stores) and Florida (~42) alone are ~36% of the fleet — placing most stores in demographically growing metros rather than the Rust Belt corridors that killed the regional chains. ~29,100 employees. Fiscal year ends the Saturday nearest January 31.

The footprint’s defining feature: ownership (FACT). Dillard’s operates ~46.0M square feet and owns ~43.0M of it (~93%) — 248 of 271 stores owned outright at FYE2025, 15 leased, 7 owned-on-leased-land, 1 partial; all six distribution centers, the fulfillment center, and HQ are owned. There are no sale-leasebacks and never have been — the anti-Macy’s. Gross PP&E of $3,827M is carried at $948M net, ~75% depreciated, so book value massively understates the real estate (Sections 6 and 10). This single fact drives the cost structure, the survival economics, and much of the valuation debate; it recurs in every section.

How it makes money (FACT). Two segments, wildly unequal:

• Retail Operations (~96% of net sales; $6,232M of $6,563M FY2025 revenue): the department stores and e-commerce. Category mix (stable for three years): ladies’ apparel 20%, men’s 19%, cosmetics 16%, ladies’ accessories/lingerie 14%, shoes 14%, juniors’/children’s 9%, home 4%. The model is full-price, national-brand-plus-exclusive-brand merchandising for a middle-to-upper-middle customer — deliberately less promotional than Macy’s or Kohl’s, with in-store presentation and service levels closer to the old-line department-store standard. Exclusive/private brands (Antonio Melani, Gianni Bini, Daniel Cremieux, Roundtree & Yorke, and others) were 22.3% of net sales in FY2025 (23.5% in FY2023 — slipping, worth watching); they carry higher margin and, because they cannot be price-compared on Amazon, blunt direct price transparency (Section 4).
• Construction — CDI Contractors (~4% of sales, ~$242M FY2025): a general-contracting business descended from the in-house team that built the stores. It is strategically sensible vertical integration (it builds and remodels Dillard’s own boxes) and economically immaterial: FY2025 pretax income $3.5M on a 5.1% gross margin, carried on cost-to-cost accounting. It exists in the financials; it does not exist in the thesis.

A third, non-segment profit stream matters more than its size: credit-card alliance income — Dillard’s share of the economics of its private-label and co-brand card portfolio, now issued by Citibank after the long Wells Fargo relationship ended in 2024. This was $39.6M in FY2025, down from $67.2M in FY2023 (−41%), at effectively 100% margin — ~5.7% of FY2025 pretax income (Section 6). There is no subscription or contract base; the recurring elements are the card income and an aging but loyal repeat-customer base.

The ownership and control structure (FACT). Two share classes with identical economics (same dividends; B convertible 1:1 into A): publicly traded Class A (11.63M shares, NYSE: DDS) and Class B (3.99M), which elects two-thirds of the board (Class A elects one-third). The Dillard family — CEO William T. Dillard II, President Alex Dillard, SVP Mike Dillard, and the wider family — holds ~51% of total shares economically and effective voting control; DDS is a NYSE “controlled company.” In June 2026 the family’s holding company (W.D. Company, 99.99% of Class B) was merged into Dillard’s share-for-share — economically neutral — with a Voting and Exchange Agreement locking recipients’ Class B into bloc voting and descendant-only transfers, expressly to preserve controlled-company status into the next generation (Sections 7 and 8). CEO William T. Dillard II is approximately 80; succession is locked in, not open.

Management temperament (FACT/INTERPRETATION). Dillard’s holds no earnings calls, issues no guidance, and gives no investor presentations (FACT; no transcripts exist for it in any public transcript corpus). The bonus formula is 1.5% of pretax income plus 3.5% of the YoY pretax increase (FACT, DEFM14A) — profitability, not growth. INTERPRETATION: an owner-operated company in the literal sense; the silence is a transparency cost for analysts and a discipline benefit for operators, and it means every claim in this report had to be built from filings, price data, and third-party sources rather than management narrative.

Verdict — Business Overview. A simple, understandable, deliberately unglamorous business: sell branded apparel, cosmetics, and home goods at full price through owned Sunbelt stores, keep costs famously lean, send the cash to the owners. The model’s simplicity is a feature for analysis (few moving parts, clean GAAP, >100% FCF conversion) and a constraint for the thesis (no adjacency, no platform, no growth vector — what you see is all there is). The one genuinely unusual asset — ~93% owned real estate — is what separates Dillard’s economics from every dead or dying peer, and it is an asset, not a strategy.

3. Industry Dynamics

Verdict up front: structurally bad — unambiguously, and it is not a close call. This is the same verdict a prior published analysis of Macy’s reached, and nothing in this fresh pass argues against it. What follows is the evidence chain.

A channel in 25-year liquidation (FACT). US department-store channel sales peaked at roughly $232B in 2000 and fell to roughly $154B by 2025 — a ~35% nominal contraction (far worse in real terms), running at ~−4.1%/yr over 2018–23 (US Census retail data / Statista-derived channel series). Third-party aggregators with different channel definitions land on different levels (~$187B, ~2.7% CAGR decline 2020–25) but the same direction: ~3–5%/yr decline is the one point every source agrees on. Critically, the decline is ongoing, not stabilized: March 2025 Census retail data showed department-store sales still falling year-over-year while total retail rose.

Where the profit pool went (FACT). The share did not vanish; it migrated to structurally advantaged formats:

1. Off-price: TJX (~$60.4B sales), Ross (~$22.8B), Burlington (~$11.5B) — all per the companies’ own FY2025 filings; global off-price ~$370B, growing high-single-digits, share gains accelerating as anchors close. Off-price passes both Greenwald moat tests (share stability, sustained high ROIC) on supply-side buying scale and supplier captivity; department stores pass neither.
2. Amazon/e-commerce — hitting home categories hardest (Macy’s home category down ~15% over two years, per Macy’s public disclosures).
3. Mass discounters and clubs (Walmart, Target, Costco).
4. Specialty category killers — Ulta and Sephora dismembering beauty, historically the department store’s best category (16% of DDS sales).
5. Brand-direct/DTC: Nike, Ralph Lauren, and most national brands are cutting wholesale dependence on department stores, as documented across the brands’ own investor communications. INTERPRETATION: the most underappreciated long-term wound for Dillard’s specifically — every brand that goes direct dilutes the reason to visit.

The mall substrate (FACT/INTERPRETATION). Enclosed-mall traffic is in secular decline; anchor closures (Sears, JCPenney bankruptcies; Nordstrom pruning; Macy’s −150 stores) create a negative feedback loop in which each closure reduces traffic for the survivors. Class A malls survive; Class B/C malls die. Dillard’s sits mostly as the anchor of Sunbelt Class B malls — the dying cohort — with two mitigants that are unique to DDS: (a) Sunbelt demographic growth versus the Rust Belt decay that killed the regionals, and (b) ownership — a dying mall imposes no rent burden on Dillard’s, and the owned pad retains alternative-use value. Both mitigants are about survival economics, not about arresting the traffic decline.

The survivor set tells the story (FACT). Sears/Kmart, JCPenney, Bon-Ton, Lord & Taylor, the May Co. regionals: gone or absorbed. Belk: bankrupt 2021. Neiman/Saks: merged into Saks Global (Dec 2024), distressed. Nordstrom: taken private May 2025 at $24.25/share (~4–5x EBITDA) — insiders judging the assets worth more off-market. Kohl’s: three CEOs in three years, dividend cut ~75%, FY2025 profit $272M on declining sales — the live bear case. Macy’s: merchandising ~break-even, the Citi card profit share (~$669M, ~76% of operating income, per Macy’s public filings) carrying the P&L. Against that field, Dillard’s ~$6.6B revenue versus $6.4B in FY2012 — a fifteen-year nominal flat line while the channel shrank ~35% — means Dillard’s share of the shrinking pie steadily rose [Forbes, “How Dillard’s Survived the Department Store Bloodbath,” 2026-06-22]. Flat nominal revenue is still a ~30% real decline: relative victory, absolute erosion.

Marathon capital-cycle read (INTERPRETATION). On Marathon’s framework the industry sits in a deep bust / capital-exit phase, and the checklist of inflection signs is superficially complete: consolidation ✓, capacity withdrawing ✓ (Sears, JCP, Nordstrom, Kohl’s, Macy’s closures), zero new entrants ✓, maximum pessimism ✓ (the Street nickname is “Dullard’s”), disciplined management at the best player ✓. In a normal cycle, that combination precedes supply-side stabilization and recovering returns for survivors. The catch is Marathon’s own breakdown condition: technology disruption. E-commerce broke the cycle’s completion mechanism — when a store closes, its customers go to Amazon and off-price, not to the surviving anchor — so capacity exit does not restore pricing power. Demand is leaving the channel faster than capacity. The correct read is endgame consolidation: “managed decline with intermittent reprieves,” not cyclical recovery. The endgame winner is whoever melts slowest with the lowest fixed costs and returns the cash to shareholders rather than reinvesting it into the decline. Note that this is, almost verbatim, Dillard’s actual strategy — no growth capex (capex ~0.5x D&A), harvest the boxes, buy back stock, special-dividend the rest. The industry verdict and the company verdict are different questions; confusing them is the classic mistake with DDS.

Regulatory and structural factors. Nothing exotic: no reimbursement, licensing, or rate regulation. The live exogenous factors are (a) tariffs — apparel is import-heavy and the April 2025 tariff shock whipsawed the whole complex (DDS −19% peak-to-trough around “Liberation Day,” then +11.3% on the pause); management cited “unpredictable costs” on the Q4-FY25 print; (b) consumer cyclicality in the middle-income cohort; © the slow attrition of the store-card profit stream as card economics and partners shift (Wells Fargo → Citi, Section 6).

Barriers to entry: high and irrelevant. No rational actor would build a department store today, so incumbents are protected from new entrants while being dismembered by existing disruptors — the barrier protects against the wrong enemy.

Verdict — Industry Dynamics. Structurally bad, full stop: a ~35% nominal contraction over 25 years and still falling, profit pools captured by formats with genuine advantages, a dying mall substrate, multi-front competition with no defensible middle, and a capital cycle that cannot complete because the disruption is technological. The disconfirming evidence weighed: capacity is exiting fast (which flattens survivors’ comps at the margin — plausibly one contributor to DDS’s +3% Q1 comp), Sunbelt demographics are genuinely better than the national mall average, and the channel is not going to zero — apparel remains a touch-and-feel category and ~$154B is still an enormous pie. None of that changes the direction of the tide; it only argues the tide may slow. Being the best swimmer matters, but this report prices the water, not just the swimmer (Section 10).

4. Competitive Position

Verdict up front, stated directly — an absent moat must be named: Dillard’s has NO durable competitive advantage of any Greenwald type. It has one narrow, genuine supply-side cost advantage — owned real estate plus lean operations — which is sufficient to keep free cash flow positive through the industry’s decline but not to arrest the decline. What looks like moat-like performance is, in Greenwald’s taxonomy, operational effectiveness in a barrier-free industry, executed better than anyone else in the format. The tests, one by one:

Demand-side advantages (captivity): none (FACT/INTERPRETATION). Switching costs are ~zero — the Dillard’s customer is one tap from Amazon and one exit ramp from TJX or Ross. Search costs are nil; apparel is a considered but simple purchase. The only shred of captivity is habit: an older, loyal, Sunbelt customer with decades of Dillard’s routine and the store card. Greenwald’s warning applies precisely — habit fades, and captive customers age without generational replacement. The traffic data corroborates: Dillard’s 28 clearance centers saw visits rise +7.5% YoY (Jan–Aug 2025) while total store visits were flat — the loyal base is holding and trading down within the four walls; the incremental customer is not arriving [dcfmodeling.com, accessed 2026-07-18 — secondary source, directional].

Network effects: none. Not claimed, none exist.

Economies of scale plus captivity: fails the captivity requirement. Dillard’s has real regional density (TX/FL concentration → local distribution and advertising efficiency), but density without captivity is an efficiency, not a barrier. Scale fails the test.

Supply-side / cost advantage: REAL — the one genuine structural edge (FACT/INTERPRETATION).

• Owned real estate (~93% of square footage). Peers pay occupancy of several percent of sales to mall landlords; Dillard’s pays almost none. In a format where the industry’s operating margin is 0–5%, avoiding rent is the difference between profit and loss — the metric that would deteriorate if this advantage vanished is occupancy-adjusted operating margin: Dillard’s ~10–11% retail operating margin versus peers’ 0–4%. Greenwald’s caveat applies honestly: this is a relative cost advantage against a defined peer set in a declining market. It extends survival and protects FCF; it cannot restore growth.
• Lean operations. Famously tight SG&A, tight inventory control (retail turns 2.6x versus ~2.2x pre-COVID), no gimmicks, no earnings calls, no investor-relations theater [Forbes, 2026-06-22]. SG&A at ~27.1% of sales in FY2025 — identical to FY2019 — says the cost discipline is structural, not a COVID artifact.
• Exclusive/private brands (22.3% of sales). INTERPRETATION: primarily a margin lever and mild differentiation, not captivity. Nobody abandons Amazon because of Gianni Bini; but exclusivity defeats direct price-comparison and supports the ~40% gross margin on a far less promotional model than Macy’s uses to reach a similar headline number. Margin defense, not a moat. (Worth flagging: penetration slipped from 23.5% FY2023 to 22.3% FY2025 — a mild counter-trend to watch.)

Greenwald’s two empirical tests.

• Share stability: PASSES — in the strangest possible way (FACT). Revenue $6.4B (FY2012) → $6.6B (FY2025): a fifteen-year nominal flat line while the channel shrank ~35%. Dillard’s share of the pie rose steadily. But the test says “durable relative position” while the absolute trend says the market is evaporating underneath — a passing grade on a curve where the school is closing.
• ROIC: strong now, anomalous at the peak (FACT/INTERPRETATION). FY2025 ROIC 23.4% (normalizing: 32.5% → 25.2% → 23.4%; FY2021–22 spike ~41%). Pre-COVID it ran 6–9% — adequate, not moat-grade. The honest read: current returns are the best in the format, explained by cost structure plus the partially-retained margin step-up plus a buyback-shrunk capital base — not by any barrier a rival cannot cross. Greenwald’s own resolution applies: absent an identifiable advantage, sustained outperformance is operational effectiveness — real, valuable, and in principle emulable.

Head-to-head versus the format (FACT):

(Sources: the companies’ public filings; Kohl’s per Idaho Business Review 2026-03-10; DDS per FY2025 10-K.)

The one-line difference (INTERPRETATION): Dillard’s is the only major department store whose merchandising operation itself earns real money. Every other survivor is a finance company with stores attached (Macy’s card), an LBO (Nordstrom), a restructuring (Kohl’s), or a corpse. That distinction is real and underappreciated — it earns the quality premium (Section 10) — but it is a distinction of competence and cost structure, not of franchise. Greenwald: in industries without barriers, only operational efficiency matters — and DDS is the canonical case. What keeps it profitable is (1) the structural cost advantage (no rent, lean opex), (2) conservative merchandising (full-price discipline, exclusive brands), and (3) capital allocation (Section 7). None is a moat; the investment question they leave behind is melt-rate versus return-rate, not moat width.

Verdict — Competitive Position. No durable advantage of any Greenwald type; one narrow and genuine supply-side cost advantage (owned real estate + lean operations) that keeps FCF positive through the decline; operational effectiveness at a best-in-format level as the true explanation of outperformance. Disconfirming evidence weighed: the share-stability test technically passes, the 22% exclusive-brand book provides real margin insulation, and the aging-but-loyal customer base is decaying slowly enough that clearance-center traffic is growing. All true — and all consistent with “the best house in a condemned neighborhood,” not with a franchise. If a claimed advantage cannot be tied to a financial outcome that would deteriorate without it, it is not a moat; here, the only claim that survives that test is occupancy-adjusted cost.

5. Growth History and Forward Opportunities

Growth history (FACT): there isn’t any, at the top line, by design. Revenue: $6.42B (FY2016) → $6.34B (FY2019) → $4.43B (FY2020, COVID) → $7.00B (FY2022 peak) → $6.56B (FY2025) → $6.60B (TTM). Fifteen-year nominal CAGR ≈ 0; ~−30% in real terms. All revenue is organic — Dillard’s has made no acquisitions of consequence in decades (the last major deal was the 1998 Mercantile Stores purchase, outside this report’s window). Store count has been essentially static for three years (273 → 272 → 271 → 272) after a long, slow rationalization — the shrink phase is largely complete. Unit growth is nil; volume/price decomposition shows the FY2021–22 revenue bulge was price/mix (margin), not traffic.

Where “growth” actually happened: per-share (FACT). The genuine growth vector of the last decade was denominator destruction: shares outstanding −51% in nine years (32.16M FY2016 → 15.62M FY2025). Pre-COVID EPS of $4–7 became FY2025 EPS of $36.42 on the same revenue — a margin step-up plus a share-count collapse. That is real value creation, but it is arithmetic, not franchise growth, and it cannot repeat at the same rate without either another margin miracle or much lower prices for the buyback.

Forward opportunities — each real, each small (FACT/INTERPRETATION):

1. The Coterie Shop (2024): a curated, younger-skewing shop-in-store format; no disclosed economics; immaterial at $6.6B scale. Merchandising evolution, not a growth engine.
2. Opportunistic real-estate moves: a new Dayton, OH store replacing a Macy’s anchor; the Longview Mall JV (~$34M, anchor-protection at distressed pricing); Willow Bend closure realizing pad value. Share-of-a-shrinking-pie tactics, individually worth tens of millions — Longview is 0.4% of market cap.
3. Card-income stabilization: service-charge income +12% YoY in Q1-FY2026, the first positive print after −41% cumulative decline through the Wells Fargo → Citi transition. One data point on a ~$40M line.
4. Comp inflection: Q1-FY2026 comps +3% after −4%/−3%/0%. If capacity exit in DDS’s Sunbelt trade areas has finally outpaced demand exit, flat-to-positive comps are the only credible “growth” scenario the equity story has. One quarter is evidence, not a trend (Section 10).
5. E-commerce: real infrastructure (dillards.com, owned fulfillment), but margin-dilutive fulfillment for a store-centric model — defensive, not offense.

What is explicitly NOT on the table (FACT/INTERPRETATION): no international expansion, no M&A appetite, no new-format rollout, no guidance. The stated capital priority is “strategic investments to enhance the value of existing properties, stock repurchases and dividend payments” — harvest language, confirmed by capex at $93M (~1.4% of sales, ~0.5x D&A).

Verdict — Growth. Low/no-growth, and the low growth is low-quality in form but honest in kind: there is no fake growth (no acquisition-stacked revenue, no promotional volume bought at negative margin), and the per-share growth that occurred was bought cheaply. The forward list is a collection of small, rational, non-transformative levers. The bull’s only real growth claim — sustained positive comps as the last man standing — is an industry-structure bet, not a company initiative, and one quarter of evidence is all that supports it. Any valuation that requires Dillard’s to grow is, on the record, betting against fifteen years of flat revenue and management’s own stated priorities.

6. Financial Quality

The COVID whiplash normalization — the full arc (FACT; FY2025 10-K filed 2026-03-27, Q1-FY2026 10-Q filed 2026-06-05, EDGAR XBRL cross-verified; company FY convention):

* DDS’s statement of operations runs to “income before income taxes” with no operating-income line; operating income here = gross profit − SG&A − D&A − rentals − other expense, tied to filing components. FY2021 OCF/FCF flattered by a +$242.5M working-capital release. TTM includes the $104.1M pretax litigation gain — see below.

Three eras, three different companies: pre-COVID (FY2016–19) a thin-margin ~breakeven-through-the-cycle retailer earning $111–221M on ~$6.4B; the boom (FY2021–22) when lean inventories, full-price selling, and stimulus demand lifted gross margin ~+950bps to 43%+ and net income to $862–892M; and the normalization (FY2023–25) — three consecutive down years (revenue −1.8%, −4.1%, −0.4%; operating margin 13.5% → 11.2% → 10.5%; ROIC 32.5% → 25.2% → 23.4%).

The central financial fact: the margin step-up is PARTIALLY holding, and it is all merchandise margin (FACT/INTERPRETATION). Gross margin has given back ~310bps of the ~950bps boom-era gain but retains ~+650bps versus the pre-COVID 33–34% base — entirely a gross-margin phenomenon: SG&A as a percent of sales has fully round-tripped to the pre-COVID ~27.1% (23.5% at the FY2021 trough). The structural gain sits in merchandise margin: inventory discipline (turns 2.6x vs ~2.2x pre-COVID; cash conversion cycle ~59 days vs ~71–75), less clearance selling, exclusive brands at 22.3% of sales. The erosion is decelerating — operating-margin give-back of −265bps (FY2023), −222bps (FY2024), −74bps (FY2025) — and Q1-FY2026 printed comps +3% (after −4%/−3%/0%) with retail gross margin up 30bps YoY to 45.8%: the first hard evidence of a floor forming well above the pre-COVID base. If the +650bps retention is structural, current earnings (~$570M net income, ~10.5% operating margin — ~2.5x the pre-COVID base on the same revenue) are near a defensible mid-cycle; if gross margin drifts back toward 35%, earnings power falls toward ~$300–350M. This single question dominates the valuation debate (Section 10).

Quality of earnings — the run-rate is cleaner than the tape (FACT):

• Q1-FY2026 includes a $104.1M pretax one-time gain ($79.6M after tax, $5.10/share) from the settlement of long-standing payment-card interchange-fee litigation (10-Q Note 9) — ~14% of FY2025 pretax income. Ex-gain, Q1 pretax was ~$223M (+4.3% YoY), net income ~$171M, EPS ~$10.94. TTM reported EPS of $42.11 is ≈ $37.0 ex-gain. Every TTM multiple in this report uses the ex-gain basis.
• FY2025 EPS of $36.42 includes ~$1.00/share of property-sale gains ($20.4M pretax on five mostly-closed properties sold for $25.7M) and a $2.24/share ESOP special-dividend tax benefit ($35.0M) → core FY2025 EPS ≈ $33.2. The ESOP benefit recurs whenever a special is paid (+$16.3M to +$35.0M/yr over five years) — arguably run-rate for this company, but it scales with the special, not with operations.
• The tax rate steps up ~500bps: FY2025 ETR 17.9% (FY2024 18.7%, FY2023 19.4%) guides to ~23% for FY2026 now that special-dividend ESOP deductions are lapped — a structural headwind worth ~$2/share of EPS against flat pretax income.
• Otherwise clean: FCF conversion >100% of net income in every recent year, no SBC distortion of consequence, no capitalized-interest games, no channel-stuffing tells (inventory +2.5% on flat sales; AP days ~54).

The credit-card income decline — a quiet drag (FACT). Card-alliance income: $67.2M (FY2023) → $54.1M (FY2024) → $39.6M (FY2025), −41% in two years, at ~100% margin, now ~5.7% of pretax income. The Wells Fargo alliance ended in 2024 (program agreement disclosed January 2024; Citi launch mid-2024) and was replaced by a new long-term Citibank alliance for the private-label and co-brand cards. Portfolio attrition through the transition is the driver. Q1-FY2026 service-charge and other income rose +12% YoY ($20.2M vs $18.1M) — one quarter of possible stabilization, not yet a trend. Note this is distinct from the one-time $104.1M interchange litigation settlement, which is unrelated to alliance economics. INTERPRETATION: unlike Macy’s — where the card profit share (~$669M) IS the business (~76% of operating income, per Macy’s public filings) — Dillard’s card stream is a nice annuity, not the P&L; its erosion is a headwind of a few percent of pretax, not an existential one.

Balance sheet: a fortress, run for survivability (FACT). Cash + short-term investments $1,073M at FYE2025 → $1,417M at Q1-FY2026 (after paying $485M of FY2025 dividends and receiving settlement cash). Debt ~$558M: $321.7M of unsecured notes at 7.00–7.75% (maturities $96.0M FY2026 / $80.0M FY2027 / $145.8M FY2028), $200M of 7.5% subordinated debentures due 2038 (interest deferrable up to 20 quarters), ~$36M finance leases. Net cash ≈ $515M at FYE2025, ≈ $862M at Q1-FY2026. Interest income ($47.3M) exceeds interest expense ($41.0M). An $800M revolver sits undrawn (no borrowings FY2023–25), extended to 2030-03-12, covenants sprung only below $80M availability. Unfunded nonqualified pension ~$314M — the one real off-balance-sheet-flavored obligation, modest against $755M of TTM FCF. Liquidity: current 2.65x, Altman Z 8.9, debt/EBITDA 0.64x. The owned real estate is carried at $948M net of 75% depreciation and is unencumbered: no mortgages, no sale-leasebacks.

Working capital and KPIs (FACT). Inventory $1,201M FYE2025 (+2.5%); turns 2.6x; CCC ~59 days — structurally ~12–16 days tighter than pre-COVID, a working-capital release already in the FCF run-rate, not a future source. Sales/sq ft $138 (FY2025) vs $143 (FY2023) — gentle productivity decline. Capex $93M (~1.4% of sales, ~0.5x D&A of $181M): harvest mode, with the corollary that PP&E is aging (75% depreciated). Construction segment: declining and immaterial.

Verdict — Financial Quality. Do economics improve with scale? No — but they degrade more slowly than peers’, which in this industry is the same trophy. Dillard’s shows no operating leverage to growth (it has had none to test), but it has demonstrated remarkable downside operating resilience: revenue −6% from FY2022 peak while operating margin held ~2.5x the pre-COVID base, because the fixed-cost floor (no rent, owned DCs, lean SG&A) is genuinely low. Financial quality is high: net cash, unencumbered owned real estate, >100% FCF conversion, a 51% nine-year share shrink, and honest accounting. The honest counters: the earnings base carries three quiet drags (card-income erosion, the ~500bps tax step-up, and a buyback that has stopped adding per-share leverage at this price), ~$3.24 of FY2025’s $36.42 EPS was one-time or special-linked, and PP&E aging means maintenance capex cannot stay at 0.5x D&A forever. The central open question the numbers leave: whether ~10.5% operating margin is the new mid-cycle floor or a waypoint lower — Q1-FY2026 says floor, three prior years said keep falling.

7. Capital Allocation

Verdict up front: intelligent, owner-grade capital allocation — among the best in retail — with a permanent family-control discount as the price of admission. This is the strongest section of the Dillard’s file, and the one the market narrative actually gets right. The evidence:

The cannibal playbook, quantified (FACT). Shares outstanding (Class A + B): ~113.8M at the FY1996 peak → 15.62M today — an 86% reduction over thirty years; −54% over ten years; −31% over five. Treasury stock carried at $5.46B cost against $1.78B of book equity — the cumulative repurchase spend dwarfs the remaining book. The program long predates the squeeze era ($300–500M/yr back in FY2014–16). The recent five fiscal years, with disclosed average prices:

Blended: $1,508M for ~6.4M shares ≈ $236/share versus $550.93 today — ~2.3x on repurchased capital, retiring ~29% of the then-outstanding count. The pattern is price-sensitive, not mechanical: ~$500M/yr deployed at $175–255, token ~$110–120M/yr at $360–367, nothing at $500+. $165.2M remains on the May-2023 $500M authorization. INTERPRETATION: this is the single most eloquent valuation datapoint in the entire file (Section 10) — the best-informed buyer of DDS stock, with a thirty-year timing record, has a strike price somewhere below ~$500.

Special dividends: the substitute return channel (FACT). As buybacks stopped clearing the family’s hurdle, returns pivoted to specials: $15 (Dec-2021), $15 (Dec-2022), $20 (Jan-2024), $25 (Jan-2025), $30 (Jan-2026, ~$468M — the largest in company history) — $105/share, ~$1.73B total, five in five years, plus a regular quarterly raised $0.20 → $0.25 → $0.30 (announced 2025-11-20; regular cost ~$19M/yr, token). The specials are ESOP-tax-advantaged (the company’s 401(k)/ESOP holds 38.8% of Class A, generating the +$16–35M/yr tax benefits noted in Section 6). Total returns versus FCF: FY2025 95%; five-year cumulative $3.32B returned of $3.99B FCF = 83% — entirely self-funded, no leverage: the cash pile rose to $1.07B while this happened. TTM shareholder yield ~5.8%, nearly all special dividend; buyback yield now ~0. INTERPRETATION: specials are pro-rata (tax-inefficient for some holders but scrupulously fair — unlike Class-A-only buybacks they do not concentrate family control) and they are board-discretion, not policy — their recurrence (sixth special, ~Nov-2026?) is an open question that matters to the total-return floor (Section 13).

What they did NOT do (FACT): no sale-leasebacks, ever (FY2025 monetized only five already-mostly-closed properties for $25.7M). No transformational M&A. No leveraged recap — returns were funded from FCF while keeping net cash. No guidance-chasing capex ($93M, ~0.5x D&A). The one new, non-core use of cash: $34.3M contributed to “mall joint ventures” (the Longview purchase — anchor-protection at ~$53/sq ft; JV earnings $0.4M, 0.4% of market cap). Cheap and defensible now; it is the first deviation from the pure harvest playbook and bears watching for creep.

Family control and governance — the discount side (FACT/INTERPRETATION).

• Structure: Class B elects 2/3 of the board; family ~51% of economics; NYSE controlled-company exemptions apply. Buybacks retire Class A only, so every repurchase mechanically tightens the family’s grip — accretive per share, concentrating in control.
• The 2026 WDC merger (closed 2026-06-04): the family holdco W.D. Company (99.99% of Class B) merged into DDS share-for-share (41,496 A + 3,985,776 B issued; WDC’s identical shares cancelled) — economically and voting neutral for public holders, tax-free, special-committee approved (vote 14,199,181 for / 28,127 against). Purpose: administrative simplification and estate planning — and its companion Voting and Exchange Agreement locks recipients’ Class B into bloc voting (majority of WTD II / Alex / Mike), grants an irrevocable proxy, and restricts transfers to lineal descendants, expressly to preserve controlled-company status. INTERPRETATION: economically neutral today; strategically it entrenches the next generation and forecloses any outsider activism.
• Texas reincorporation (effective 2025-08-31): part of the post-Tesla migration to a management-friendlier jurisdiction; weakens minority recourse at the margin. The poison-pill remnant was formally eliminated 2025-07-18 — the dual-class structure makes a pill redundant anyway.
• Compensation is modest and profit-linked (FACT, DEFM14A): cash bonus pool = 1.5% of pretax income + 3.5% of the YoY pretax increase (FY2025 pool $10.42M on $694.5M pretax); CEO total $9.10M, but $4.43M of that is a pension-value swing — direct comp ~$4.7M, 62.5% bonus. Family base salaries frozen ≥3 years. No options, no severance/CIC agreements. Say-on-pay ~98–99% — non-informative, since the family votes ~half the shares.
• Related parties (FACT): ~6 additional family members on payroll at ~$5–6M/yr aggregate (<1% of SG&A — real “jobs program” evidence, economically trivial); Stephens Insurance ~$1.07M; The Connor Group $3.42M. No family store leases, no family real-estate JVs, no self-dealing M&A — the classic family-retailer conflicts are absent.
• Insider behavior (FACT, full 60-month Form 3/4/5 sweep — 989 filings, 1,050 transactions): ZERO open-market purchases and ZERO open-market sales by any Dillard family member or entity; zero 10b5-1 plans anywhere in the corpus. Nobody sold the 2021–22 squeeze. The only conviction trade in five years: director Warren Stephens bought 20,000 shares (~$4.6M) at ~$228–231 in February 2022, after the stock had already tripled. All 15 open-market sales in the corpus were non-family and tiny (10,236 shares / $4.1M combined). INTERPRETATION: a maximally aligned insider tape — the family neither sells nor needs to buy; buybacks and specials are their only liquidity valve, and they chose specials over selling.

Disconfirming evidence, weighed honestly: (1) They missed the 2020 COVID bottom (~$103M of FY2020 buybacks at $40–60 would have been the trade of the decade; they chose liquidity preservation in an existential moment — defensible for a family whose wealth IS the company, but it dents the “perfect timer” mythology). (2) Buybacks retire Class A only, tightening the dual-class grip. (3) The mall JVs are a new, non-core cash use — trivial at $34M, a yellow flag only if they grow. (4) Capex at 0.5x D&A is a managed-decline posture, not renewal. (5) Governance is naked entrenchment — Class B control, the Voting & Exchange Agreement, Texas law — and the discount for permanent minority passenger status is structural and forever.

Verdict — Capital Allocation. Among the best in retail: counter-cyclical cannibal repurchases at a blended ~$236, disciplined tapering to zero as the price ran, special dividends substituted when buybacks stopped clearing the hurdle, 83% of five-year FCF returned unlevered, a fortress balance sheet kept for survivability, no financial engineering, modest profit-linked pay, and a spotless insider tape. Minorities are permanent passengers — but passengers on a well-driven bus, and the driver’s returns have been the passengers’ returns. The control discount is the admission price, not a capital-allocation failure.

8. Changes and Headwinds — Last Two Years

The last ~24 months were unusually eventful for a company this quiet. Timeline of the material changes (FACT unless noted; sources: 8-Ks, DEFM14A, 10-K/10-Q, company releases — full event table in Appendix B):

• Credit-card partner transition (2024): the Wells Fargo alliance was non-renewed; a new long-term Citibank program launched mid-2024. Card-alliance income fell $67.2M → $54.1M → $39.6M (FY2023–25). Headwind, possibly stabilizing (+12% YoY in Q1-FY2026 service-charge income).
• Comps trajectory turned (FACT): FY2023 −4% → FY2024 −3% → FY2025 0% → Q1-FY2026 +3%, with retail gross margin +30bps. The first positive comp since the boom and the strongest fundamental data point in two years. INTERPRETATION: one quarter; the entire current multiple is levered to whether it holds (Section 10).
• Earnings normalization continued: FY2025 net income $570M / EPS $36.42 (vs $593M / $36.82); the Q4-FY25 print drew a −7.8% single-day reaction on flat comps and “unpredictable costs” language. Then Q1-FY2026: EPS $16.04 including the $104.1M interchange-settlement gain — a solid underlying quarter (+4.3% ex-gain pretax) that got no traction in the tape.
• The $30 special and the quarterly raise (2025-11-20): largest special dividend in company history (~$468M, paid 2026-01-05), quarterly raised $0.25 → $0.30. The return-of-capital engine is now dividend-first.
• Buybacks stopped (FACT): zero repurchases since ~August 2025; $9.8M TTM; $165.2M left on the authorization. The family’s own valuation signal, discussed in Sections 7 and 10.
• WDC holdco merger (signed 2026-03-20, approved 2026-05-28, closed 2026-06-04): economically neutral consolidation of the family’s Class B into direct ownership, with the Voting & Exchange Agreement locking next-generation bloc voting and descendant-only transfers. Post-merger Schedule 13D: WTD II at 30.0% of Class A. Governance entrenchment extended a generation (Section 7).
• Reincorporation Delaware → Texas (special meeting 2025-08-19; effective 2025-08-31) and elimination of the poison-pill remnant (2025-07-18, a housekeeping formality).
• Real-estate moves: Longview Mall JV with Trademark Property (~$34M for a 646K-sq-ft Class B mall where DDS is anchor, ~Aug-2025) — the first mall-ownership experiment; Willow Bend (Plano) closure January 2026 with pad value realized in redevelopment; a new Dayton, OH store replacing a Macy’s anchor; five mostly-closed properties sold for $25.7M (gain $20.4M).
• Board: Warren Stephens resigned 2025-04-29 (confirmed US Ambassador to the UK) — the board’s most prominent outside business figure, and the author of the only conviction insider buy in the corpus.
• Tariff whipsaw (Apr-2025): −19% peak-to-trough around “Liberation Day,” +11.3% on the 90-day pause; DDS bought ~$98M of stock into that selloff — its last material purchase.

Verdict — Changes and Headwinds. Net-net, the two-year changes strengthen the operating thesis and weaken the governance/entry-price case. The comps inflection, margin-floor evidence, and possible card stabilization are genuine improvements in the fundamental picture; the capital-return machine accelerated ($30 special); the real-estate strategy showed intelligent opportunism. Against that: the Citi transition cost two years of card income, the tax rate steps up ~500bps, the buyback — the best per-share value lever — is dormant because of the price, and the governance moves (Voting & Exchange Agreement, Texas) make the family-control discount permanent and hereditary. None of the changes alters the industry verdict (Section 3); they sharpen the question of what the stabilization is worth (Section 10).

9. Risk Analysis

Structured risk matrix. Likelihood/impact are INTERPRETATION; evidence basis is cited per row. The distinguishing feature of DDS’s risk set: existential/financing risk is about as close to zero as retail gets — the risks that matter are earnings-power, valuation, and governance risks.

Verdict — Risk Analysis. The risk stack is unusual: business-model risk (melt, margin) is real and material; financial risk is negligible; governance risk is structural but economically bounded (the family’s returns are the minorities’ returns); and the largest risk at today’s price is valuation risk — paying the top of the base zone for a stabilization that has one quarter of evidence. The disconfirming read: the two risks that could make the equity cheap fast (melt acceleration, GM reversion) both have genuine, recent counter-evidence (comps +3%, GM +30bps, decelerating erosion) — this is a debated risk set, not a one-sided one.

10. Valuation Discussion

Embedded-expectations analysis only — no price target, no recommendation. Scenario zones are fundamental-clearing ranges, not targets. Sell-side figures cited below are third-party color, not adopted.

Anchors (FACT, 2026-07-17). Price $550.93; 15.617M shares; market cap $8.60B; net cash $862M (Q1-FY2026); EV ~$7.74B. TTM EBITDA ~$870–880M → EV/EBITDA ~8.8x; EV/EBIT ~11.1x. P/E 13.1x on reported TTM EPS ($42.11) — which includes the $104.1M settlement — 14.9x on ex-gain TTM EPS (~$37.0); 16.5x on core FY2025 EPS (~$33.2). FCF yield ~8.8% reported / ~7.6% ex-settlement. Shareholder yield ~5.8% TTM (nearly all special dividend).

Own-history percentiles — the single highest-signal datum (FACT, azitrading.com valuation-percentile data, 2026-07-17). DDS trades at the 92.8th composite percentile of its own ~10-year valuation history: P/E 15.13x = 90.4th, P/S 1.315x = 92.3rd, P/B 4.83x = 95.7th. (P/B is structurally broken for DDS — $5.46B of treasury stock against $1.78B book equity makes book a buyback artifact, not asset value; cited for completeness.) INTERPRETATION: the history is contaminated in both directions — the 2021 squeeze (price decoupled from fundamentals) and the pre-COVID depressed-EPS years (optically high P/Es on $4–7 EPS) — which makes a 90th-percentile read a stronger statement, not a weaker one: only the squeeze/boom extremes were richer. The EV/EBITDA series makes the same point without the distortions: FY-end marks of 4.1–5.0x (FY2017–20) → 3.9–5.4x (FY2022–24) → 7.7x (FY2025) → ~8.8x spot — a ~2x re-rating versus pre-COVID on revenue that is the same $6.4–6.6B it was in 2012. What changed: the +650bps retained margin (real earnings power), the 51% share shrink with 83% of FCF returned (real per-share value), and float scarcity (positioning). The current multiple prices the first two as permanent.

The comp set (FACT, 2026-07-17, third-party color):

M and KSS are the only listed same-format reads — the right anchors for “what a declining department store trades at” (5–7x EV/EBITDA) — though neither matches DDS on quality (real merchandise margin, net cash, owned boxes, no restructuring risk), so a premium is deserved. The question is its size: DDS at ~8.8x versus M at 5–7x is the widest absolute gap in the series, and the most honest mark — informed insiders plus a sophisticated partner paying ~4–5x EBITDA in a negotiated 2025 control transaction for a better-positioned department store — sits far below. INTERPRETATION: the premium to peers is partly justified; the premium to the transactions is not, unless the melt is truly over.

Reverse-DCF — what $550.93 underwrites (FACT math, INTERPRETATION read; EV $7.74B, normalized FCF $600–650M central $625M, WACC 9–10% — deliberately franchise-penalizing for a no-moat, high-idio-vol small-cap):

• Gordon frame: implied perpetual FCF growth of +0.6% to +2.3%/yr. The market is not pricing a melting ice cube; it is pricing a stabilized annuity with slight growth.
• Two-stage melt paths (10 years, exit multiple on year-10 FCF): channel-rate melt (−4%/yr, 6x exit) → ~$328/sh; slow melt (−2%/yr, 7x) → ~$376; flat decade (0%/yr, 8x) → ~$436; modest growth (+2%/yr, 8x) → ~$489. To reach the current EV under any melt-framed DCF requires ~+3–5%/yr decade growth or exit multiples of 10–15x FCF — against Nordstrom’s actual ~4–5x EBITDA takeout and DDS’s own pre-COVID 4–5x EV/EBITDA.

Every melt path lands 25–40% below the current price. The load-bearing embedded expectations at $550.93 are: (1) the +650bps gross-margin retention is permanent; (2) Q1-FY2026’s +3% comp marks the end of the decline — one quarter of evidence; (3) the ~2x premium to where these assets actually transact persists.

Sum-of-the-parts (per share on 15.617M; the RE band is reasoned, not measured — no appraisal exists):

INTERPRETATION: the $8.60B market cap sits at the top of a deliberately generous SOTP — the market is simultaneously paying full price for the real-estate optionality AND for near-peak merchandise earnings power. Historically it paid for one or the other, never both. And the SOTP high end requires ~$3B for the real estate — defensible only on a pad-by-pad redevelopment basis the family has shown no appetite to execute at scale (no sale-leasebacks ever; monetizes dead boxes only after closure; Longview is 0.4% of market cap). Treat the high end as real optionality controlled by a family that monetizes glacially.

Scenario zones (2–3-year view to FY2028 — RANGES, not targets; exit multiples anchored to DDS own history, M/KSS, the Nordstrom takeout, and BKE):

• BEAR ~$220–350 — the melt resumes: GM slides to 37–38%, comps back to −3 to −5%, operating margin 6–7.5%, card income fades, 23% ETR → EPS ~$22–28; exit 4.5–5.5x EV/EBITDA as the stabilization premium collapses (Nordstrom/Macy’s zone). P/E cross-check $22–28 × 8–11 ✓. Load-bearing assumption: the GM step-up is cyclical lag, not structure.
• BASE ~$400–560 — stabilized annuity, premium narrows: comps flat-to-+2%, GM ~40%, operating margin 9.5–10.5%, EPS $32–37 (23% ETR offsets comp gains); specials continue; exit 6–8x. Today’s $550.93 sits at the very top of the base zone — the market is already paying for the stabilization case in full.
• BULL ~$600–850 — last-man-standing re-rate: comps stay positive through FY2027–28 as capacity exit outpaces demand exit in DDS’s Sunbelt trade areas; GM holds ≥40.5%; buyback resumes and shrinks the count toward ~14M (~5%/yr EPS leverage); EPS $42–48; exit 8–10x. Squeeze spikes can overshoot beyond the zone (2021, H2-2025). Load-bearing assumption: demand exit slows to ≤ capacity exit in DDS’s specific trade areas — the industry-level evidence (Section 3) is against it; the bull case needs DDS-specific share consolidation to dominate.

What the market is pricing correctly vs incorrectly (INTERPRETATION). Correctly: (1) best-in-format operations deserve a premium over M/KSS; (2) survivability — no rent, net cash, undrawn revolver — means no bankruptcy discount; (3) the capital-return stream is real (~5.8% TTM shareholder yield, 83% of FCF returned) — the special-dividend cadence (~$25–30/yr ≈ ~5% of price) is a genuine cash floor under total return if continued. Incorrectly, or at least unproven: (1) zero melt — flat-to-growing FCF in perpetuity against a ~−4%/yr channel and 15 years of flat revenue; (2) permanence of the margin step-up on one quarter of floor evidence; (3) persistence of a ~2x premium to the transaction benchmark. For $550.93 to work, all three must hold simultaneously.

The squeeze caveat (FACT/INTERPRETATION). With 35.25% of the float short (9.26 days-to-cover), the “market’s embedded view” is partly a positioning artifact, not a clean expectations signal: shorts are a structural bid; the 2021 squeeze and H2-2025 melt-up both overshot any plausible fundamental path; the current −21% drawdown is the unwind of that overshoot. Two implications: the ~8.8x premium partly reflects float-scarcity economics rather than a considered DCF; and zone paths will be violently non-linear — the bear zone can gap there in days (10–19% single-day moves are routine), and the bull zone can overshoot beyond any DCF. Zones describe where fundamentals would clear, not where the tape must travel.

Verdict — Valuation. At $550.93 the market prices Dillard’s as a stabilized annuity at the 93rd percentile of its own history, above every melt-path DCF, at the top of a generous SOTP, and at ~2x the asset class’s last control-transaction multiple. The quality premium over peers is earned; the premium over reality (zero melt, permanent margins, persistent multiple) is not yet evidenced. The strongest single informed-money datapoint on what DDS is worth remains the family’s own tape: $1.5B deployed at a blended ~$236, token buying at $360–367, zero above ~$500. The market is bidding what the family declined to.

11. Variant Perception

The consensus belief (INTERPRETATION, triangulated from sell-side/press color). “Great operator, full price.” Sell-side is split buy/sell; Zacks had DDS as a #1 momentum rank in June 2026 while GuruFocus flagged it overvalued the same month; the Seeking Alpha bear case (“Strength Is Temporary,” 2026-05-15) argues tariff-inflated comps and e-commerce erosion; the Forbes long-read (2026-06-22) admires the survival story. This report’s read: consensus is correct on both halves — and is not actually a view. “Great operator, full price” is a description, not a variant perception; the money is made or lost on what “full” means.

The strongest bull case (INTERPRETATION): melt-rate ≤ capacity-exit, plus cannibal math. Dillard’s doesn’t need the industry to recover — it needs demand exit in its own Sunbelt trade areas to slow to the pace of capacity exit, showing up as flat-to-positive comps while ~35% of the float is short. The evidence: comps inflected −4 → −3 → 0 → +3; capacity is exiting DDS’s markets fast (Macy’s −150 stores, JCP/Sears/Belk carcasses); the margin floor held at ~10.5% through three normalization years; the Citi book is stabilizing; and if the buyback resumes at lower prices, the float math turns reflexive — every retired share concentrates earnings and squeezes a third-of-float short base. In that world EPS grinds to $42–48 on ~14M shares, specials continue at $25–30 (~5% yield), and the stock compounds double digits with squeeze spikes on top. 2021 and H2-2025 prove the overshoot machinery exists.

The strongest bear case (INTERPRETATION): the step-up is a lag, and the bid is gone. The +650bps of retained gross margin is not structure — it is the long tail of the COVID inventory-discipline regime decaying on a multi-year lag, as department-store margins have always mean-reverted. SG&A has already round-tripped, so there is no cushion: when GM breaks, operating margin goes back toward 6–7%, EPS toward $22–28, and the multiple reverts to the 4.5–5.5x zone where these assets actually clear (Nordstrom ~4–5x) — bear-zone math ~$220–350, i.e., −40% to −60%. And the tell is the family’s own tape: the best-timed buyer in the stock’s history went to zero above ~$500. If DDS were cheap, the Dillards would be buying it. They aren’t.

The factor-positioning read (FACT/INTERPRETATION, FactorsToday 2026-07-17/18). The tape says the market is already debating the bear case: negative Momentum loading (−0.26) with mildly positive Quality (+0.22) — not a momentum trade carrying a low-quality name, but a value/yield/small-cap profile (DividendYield +0.69, Value +0.29) whose own tape is in a seven-month downtrend below falling 50/200-EMAs, with 3m/6m Sharpe of −0.86/−0.88. The dominant tilts are in a favorable regime (Value z +1.70; DividendYield z +1.66 over 252 days) while the sector basket is out of favor, and the recent unwind is idiosyncratic (specific vol ~36%/yr ≈ two-thirds of total variance). The crowd is not offsides in either direction: ~35% of the float is short (not an abandoned name), yet the momentum crowd has left. INTERPRETATION: an honest standoff — shorts pay ~5%/yr in specials to bet on melt, value buyers bet on stabilization, and the microstructure guarantees the resolution is violent.

The 3–5 assumptions that matter most, with falsifiers:

1. The +650bps gross-margin retention is structural. Falsifier: retail GM −100bps+ YoY for two consecutive quarters (watch Q2-FY2026, ~Aug-2026). Confirmer: GM ≥ ~40% with comps ≥ flat through FY2026.
2. Q1’s +3% comp is a floor, not a blip. Falsifier: comps negative in either of the next two prints. Confirmer: two more positive quarters — at which point the base zone migrates up and the “one quarter of evidence” objection dies.
3. The family’s bid returns at lower prices. Falsifier: no buyback resumption even on a material drawdown AND no sixth special in Nov-2026 — that combination would say the family’s own valuation read has shifted down structurally. Confirmer: buyback resumption disclosures in Q2/Q3 10-Qs, or a sixth special.
4. Special dividends recur. Falsifier: Nov-2026 passes without a declaration — removes the ~5% cash floor under total return and the carry cost that keeps shorts honest. Confirmer: declaration of special #6.
5. The exit-multiple premium persists. Falsifier: any melt-path quarter that prints DDS’s EV/EBITDA toward the 5–7x peer band on still-stable earnings — i.e., de-rating without fundamental deterioration. Confirmer: none available quickly; multiples are confirmed only in hindsight.

Where this report lands relative to consensus (INTERPRETATION). The variant perception is narrow and specific: the market is pricing the melt at zero. Consensus admires the operator (correct), concedes the price is full (correct), and stops there. The incremental claim here is that “full” is quantifiable — top of the base zone, top of the SOTP, 93rd percentile of own history, +0.6–2.3% perpetual growth embedded — and that the resolution trigger is dated and observable: the Q2-FY2026 print, the buyback line in the next 10-Q, and the November dividend decision.

12. Fact vs. Interpretation

13. Open Questions

1. Does the +3% comp hold? Q2-FY2026 prints ~August 2026 — the single datum the current multiple is levered to. Two more positive quarters with retail GM ≥ ~40% substantially proves stabilization; a negative print breaks it.
2. Will the board declare a sixth consecutive special (~Nov-2026)? Board discretion; no filing commits. At ~$25–30 it is ~5% of the price — the cash floor under total return and a meaningful part of shorts’ carry cost. Its absence would itself signal the family’s own valuation read.
3. Does the buyback resume, and at what price? $165.2M remains authorized. The 10-Q repurchase tables will reveal the family’s current strike price — the best informed-money valuation datapoint available.
4. Interchange settlement — are further installments coming? Which case/network, and whether additional payments remain, is not fully clear from the 10-Q note; a second gain would distort TTM multiples again.
5. Longview Mall JV economics: DDS’s ownership %, capital account, and whether the $34.3M “mall joint ventures” line covers assets beyond Longview — the 10-K does not itemize. Forward question: does mall ownership stay a $34M anchor-protection trade or creep into a capital sink — the first deviation from the harvest playbook?
6. Durability of exclusive-brand penetration: 23.5% → 22.3% (FY2023→FY2025). If the margin-defense lever is itself slipping, the GM-floor case weakens.
7. Card-income trajectory under Citi: one quarter of +12% service-charge income. Stabilization or transition noise?
8. Class B Form 3 paper trail: recipient family members’ Form 3s for the post-merger Class B distribution have not yet appeared in the parsed corpus (issuance sits in 8-K/Form D/13D) — a mechanical item, but the corpus should be checked.
9. Real-estate value: no appraisal exists; the $1–3B band is reasoned from book, the Longview print, and pad-level redevelopment analogs. Any large JV expansion or pad sale would provide the first real marks.

14. What Must Be True

BULL case — “last man standing gets paid” (zone ~$600–850). Must be true: (1) comps stay ≥ flat through FY2026–27 — demand exit in DDS’s Sunbelt trade areas slows to ≤ capacity exit as Macy’s/JCP/Sears/Belk carcasses consolidate share to the surviving anchor; (2) the +650bps GM retention is structural — retail GM holds ≥ ~40%; (3) the Citi card book stabilizes (~$40M floor, +12% Q1 print extends); (4) the buyback resumes on any weakness, shrinking the count toward ~14M and re-adding ~5%/yr of EPS leverage; (5) specials continue at ~$25–30/yr. Falsification test: two consecutive quarters of negative comps or retail GM −100bps+ YoY, starting with the Q2-FY2026 print (~Aug-2026); or a passed November with no special AND no buyback resumption on weakness. Any one of those breaks the bull’s load-bearing assumption that the melt is over.

BEAR case — “the step-up is a lag, the bid is gone” (zone ~$220–350). Must be true: (1) the retained GM step-up is the decaying tail of the COVID inventory-discipline regime, not structure — GM slides toward 37–38%; (2) SG&A, already round-tripped, offers no cushion — operating margin compresses to 6–7.5%, EPS to $22–28; (3) the multiple de-rates to the 4.5–5.5x transaction zone once stabilization is falsified; (4) tariff/clearance pressure or a middle-income consumer downturn accelerates the slide; (5) the family stays on the sidelines — confirming, by inaction, that even ~$400 doesn’t clear their hurdle. Falsification test: the mirror — comps hold positive through the holiday quarter (Q4-FY2026 print, ~Feb-2027) with GM ≥ 40%, and/or buyback resumption north of ~$400 (the family re-underwriting the price). The bear must also explain three years of decelerating margin erosion and the Q1 inflection — genuine evidence against it.

What must be true for the CURRENT price (~$551) to be roughly right: the stabilization case in full — permanent GM retention, comps ≥ flat, premium persistence — with no margin of safety for any of the three failing. That is the whole valuation debate in one sentence.

15. Source Appendix

Full source documentation for this report — every SEC filing relied upon, every URL cited, the third-party data pulls and their reconciliation to filings, and the insider-transaction sweep — is set out in Appendix B below. Primary sources of record: FY2025 Form 10-K (filed 2026-03-27, acc. 0000028917-26-000006), Q1-FY2026 Form 10-Q (filed 2026-06-05, acc. 0000028917-26-000019), DEFM14A (2026-04-06), and the 8-K series 2021–2026, all publicly available on SEC EDGAR.

The analysis above (Sections 1–15) carries no BUY/SELL recommendation and no price target; the Claude’s Take block is the sole, labeled exception and reflects the author’s own independent opinion, offered as general information and not investment advice. Facts, Interpretations, Assumptions, and Open Questions are labeled throughout.

APPENDIX A — Standard Diligence Questionnaire

Dillard’s, Inc. (NYSE: DDS) — Standard Diligence Questionnaire

Supplemental appendix to the analysis above, dated 2026-07-18. Every answer is grounded in that analysis and in the public sources listed in Appendix B. This appendix contains no BUY/SELL recommendation and no price target.

Label key: (F) = Fact, sourced to a filing, data pull, or cited press item · (I) = Interpretation, an analytical judgment on the facts · (A) = Assumption, a working premise not directly evidenced · (OQ) = Open Question, unresolved as of the report date.

Fiscal-year convention (F): Company labeling is used throughout — FY2025 = 52 weeks ended 2026-01-31; Q1-FY2026 = 13 weeks ended 2026-05-02. (ROIC.ai labels these one year later; filing labels win.)

1. General — What thoughtful questions have other investors asked about this company?

The questions below are the ones the evidence shows sophisticated investors actually litigate in this name — several map directly onto observable behavior (the family’s own buyback tape, the short base, sell-side framing split “buy/sell” per Zacks/MarketBeat color (F)).

1. Is the post-COVID gross-margin step-up durable, or a long cyclical lag? This is the load-bearing question. (F) Gross margin jumped ~+950bps in FY2021–22 (to 43%+) and has since given back ~310bps, retaining ~+650bps vs the 33–34% pre-COVID base — and the retention is entirely a merchandise-margin phenomenon: SG&A as % of sales has fully round-tripped to the pre-COVID ~27.1%. (F) Q1-FY2026 printed comps +3% with retail GM +30bps YoY to 45.8% — the first hard stabilization evidence after three down years. (I) If retention is structural, current earnings are near a defensible mid-cycle; if GM drifts back toward 37–38% or lower, EPS falls toward the mid-$20s and the multiple premium has no anchor.
2. What is the owned real estate really worth? (F) DDS owns ~43.0M of ~46.0M sq ft (~93%), plus all 6 distribution centers and HQ; net PP&E is $948M against $3,827M gross — 75% depreciated, so book massively understates value. (I) Value is highly unequally distributed: Sunbelt anchor pads carry redevelopment value (Willow Bend pad realized through redevelopment; third parties like Trademark transacting with DDS), while the B/C-mall box tail is worth only its store FCF — the Longview datapoint (~$34M for an entire 646K-sq-ft Class B mall, ≈$53/sq ft distressed) tempers aggressive marks. The reasoned band is $1–3B; no appraisal exists (OQ) — the number is reasoned, not measured.
3. Why is the short interest ~35%, and what does it do to the stock? (F) 1,084,593 shares short = 35.25% of the ~7.33M-share float, 9.26 days-to-cover (settlement 2026-06-30), at/above 2021 squeeze-era levels; shorts are rising (1.02M prior month). (I) The short base is the secular-decline thesis expressed against a family-controlled, buyback-shrunk float — it is simultaneously a structural bid (forced covering on any good print) and evidence that a large, informed constituency is betting on the melt. It also means quoted multiples and “embedded expectations” are partly a positioning artifact, not a clean expectations signal.
4. Will the family ever sell, or take it private? (F) Zero open-market purchases or sales by any Dillard family member or entity in the entire 60-month Form 3/4/5 corpus (1,050 transactions parsed); the 2026 WDC merger + Voting & Exchange Agreement locks Class B into bloc voting with transfers restricted to lineal descendants, expressly to preserve “controlled company” status. (I) Every signal points the other way: the family chose escalating special dividends over selling shares even at $400–600+ prices. The Nordstrom take-private (~4–5x EBITDA, May 2025) shows what the control path looks like for this asset class — but DDS’s family is entrenching, not exiting. (A) A take-private is not impossible post-succession, but nothing in the public record supports expecting one.
5. Is the current multiple a squeeze artifact? (F) EV/EBITDA ~8.8–9.1x at spot vs DDS’s own pre-COVID 4.1–5.0x, Macy’s 5–7x, and the Nordstrom control transaction at ~4–5x; composite own-history valuation percentile 92.8th. (I) Partially — the H2-2025 melt-up (+58% in five months to $731) and the 2021 squeeze both overshot any fundamental path, and ~35% short interest props the price structurally. But the re-rating also capitalizes real changes: +650bps retained gross margin, a 51% nine-year share shrink, and an 83%-of-FCF return record. The honest decomposition: the multiple prices genuine quality plus a scarcity/positioning premium that would not survive a fundamental break.
6. Why did the best-informed buyer stop buying? (F) Buybacks: $544.9M at $175 avg (FY2022) → $452.9M at $255 → $281.4M at $307 → $121.0M at $367 → $107.8M at $359 — and zero since ~August 2025 ($9.8M TTM), with $165.2M left on the authorization. Blended FY22–26: $1.508B for ~6.4M shares ≈ $236/sh vs $550.93 today. (I) The family’s revealed hurdle sits somewhere below ~$500; they substituted special dividends rather than chase the stock. This is the single cleanest informed-money datapoint in the file.
7. What happens at succession? (F) CEO William T. Dillard II is ~80 years old (proxy biography); the WDC merger (closed 2026-06-04) is explicitly estate-planning-driven; the Voting & Exchange Agreement names WTD II, Alex Dillard, and Mike Dillard as the initial “Authorized Representatives” controlling the Class B bloc. (I) Governance continuity to the next generation is already engineered — succession is an estate event, not a control event.

2. Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? (F) EPS arc: $4.37 (FY2019) → −$3.16 (FY2020 COVID) → $41.87 / $50.95 (FY2022 peak) → $44.78 → $36.82 → $36.42 (FY2025) → reported TTM $42.11 (includes the $104.1M interchange settlement gain) → ~$37.0 ex-gain TTM; core FY2025 EPS ≈ $33.2 after stripping $1.00/sh property-sale gains and the $2.24/sh ESOP special-dividend tax benefit. Op margin: ~2.6–5.2% pre-COVID → 16–17% boom → 10.5% FY2025. (I) Earnings are normalizing from the COVID-era peak and sit between peak and mid-cycle: current op margin is ~2.5x the pre-COVID ~4% base and ~640bps below the FY2021 peak; net income ~$570M is ~4x the pre-COVID $111–221M norm on essentially the same revenue. (I) Whether ~10.5% op margin is the new mid-cycle floor or a waypoint lower depends entirely on the +650bps retained gross margin (see Question 1.1) — the margin erosion is decelerating (−265bps FY23, −222bps FY24, −74bps FY25) and Q1-FY2026 is the first quarter of positive comps since the boom. (I) Judgment: current earnings are above true mid-cycle if mid-cycle means the pre-COVID base, but plausibly at mid-cycle if the GM step-up is structural — that is an unresolved empirical question (OQ), with one quarter of favorable evidence.

Driven by the external environment or by management’s own actions? Both, separably. (F) External: stimulus-era demand + industry-wide inventory scarcity + a no-promotion environment drove the FY2021–22 GM explosion (the whole retail sector spiked); tariffs and consumer cyclicality drive the swings around the trend. (F) Company-specific: the retained portion is attributable to company-specific discipline — inventory turns 2.6x vs ~2.2x pre-COVID, cash conversion cycle ~59 days vs ~71–75 pre-COVID, exclusive-brand penetration 22.3%, full-price selling discipline. (I) The boom was macro; the hold is management.

How stable are revenues? (F) Remarkably stable in nominal terms: $6.4B (FY2012) → $6.4–6.5B (FY2016–19) → $6.6–7.0B (FY2021–22) → $6.56B (FY2025) — a 15-year nominal flat line excluding the COVID crater. Comps: −4% (FY23), −3% (FY24), 0% (FY25), +3% (Q1-FY26). (I) Nominal flatness over 15 years is ~30% real decline; “stable revenue” and “melting business” are both true — unit volumes are shrinking while price/mix holds the top line.

Outlook for products/services? How big will this market be — growing, shrinking, domestic or international? (F) US department-store channel sales peaked ~$232B (2000) → ~$154B (2025), declining ~4.1%/yr over 2018–23; March 2025 Census data showed department-store sales still falling while total retail rose. (F) The business is ~100% domestic: 272 stores in 30 states (TX ~57, FL ~42) plus dillards.com; no international operations. (I) The market is shrinking secularly, not cyclically — share is migrating to off-price (TJX/ROST/BURL, growing HSD), Amazon/e-commerce, discounters, specialty (Ulta/Sephora), and brand-direct/DTC. The industry verdict reached in Section 3 of the main report: “managed decline with intermittent reprieves,” not cyclical recovery. DDS’s share of the shrinking pie is rising, but the pie shrinks faster than DDS gains.

3. Business Quality & Competitive Moat

Is the industry getting more or less competitive? (I) More competitive, on every front simultaneously. (F) Off-price continues to gain share with an accelerating double tailwind (displaced anchor shoppers + freed closeout inventory); Amazon keeps compressing home categories; national brands are deliberately reducing wholesale dependence via DTC — arguably the most underappreciated long-term wound for Dillard’s specifically, whose model is national brands at full price. Barriers to entry are high but irrelevant: no rational actor would build a department store today; the barrier protects incumbents from new entrants while doing nothing against existing disruptors. (F) Marathon capital-cycle read: deep capital-exit phase (Sears, JCP, Bon-Ton, L&T gone; Nordstrom private; Macy’s/Kohl’s shrinking) — but demand is leaving faster than capacity, and e-commerce is the textbook technology disruption that breaks the normal capital-cycle recovery.

How profitable is the business (ROIC, ROE)? (F) ROIC (per ROIC.ai aggregated data): 6.0–9.1% pre-COVID (FY2016–19) → 41.5%/41.4% (FY2021–22) → 32.5% → 25.2% → 23.4% (FY2025). ROE recomputed on average equity: ~32% FY2025 (note: ROE is optically inflated — treasury stock of $5.46B at cost shrinks book equity to $1.78B; ROIC.ai’s displayed ROE of ~9% is inconsistent with the filings-implied figure and is not quoted). (I) 23.4% ROIC is the best in the format by a wide margin (peers single digits) but is explained, not moat-derived: it is the product of a cost structure (no rent, lean SG&A), the retained margin step-up, and a shrunken capital base — not of a barrier competitors cannot cross. Pre-COVID ROIC of 6–9% is what this business earned without the COVID anomaly — the honest no-moat base rate.

How profitable is the industry — how many competitors, what barriers to entry? (F) Traditional department stores earn 0–5% operating margins at best: Macy’s ~3.9% core (merchandising ~break-even standalone; the Citi card profit share ~$669M is ~76% of op income), Kohl’s ~2–3%, Saks Global distressed. (I) Dillard’s ~10.5% op margin is the outlier, and the occupancy-adjusted gap is the metric that would deteriorate if its one advantage vanished: DDS ~10–11% vs peers 0–4%, i.e., most of the gap is the owned-real-estate cost advantage. Direct same-format listed competitors have dwindled to Macy’s and Kohl’s — but the true competition is the formats taking the profit pool (off-price at ~12% op margins).

Can the business be easily understood? (F) Yes — an unusually simple model: ~272 mall-anchor stores selling national brands + exclusive labels at full price, ~96% of sales from retail merchandising, a tiny (~4% of sales, ~breakeven) in-house construction arm (CDI Contractors), card-alliance income ~5.7% of pretax, no international, no meaningful M&A, no financial engineering. (F) The company holds no earnings calls — press releases only — which is itself a governance/transparency data point.

Can it be undermined by foreign low-cost labor? (I) Not directly — retail labor is domestic and non-offshorable — but indirectly yes: the merchandise it sells competes with import-priced goods at Amazon, Shein/Temu-style channels, and off-price. (F) The tariff whipsaw of April 2025 (−11.9% then +11.3% in days) showed the market treats DDS as a high-beta tariff proxy; the 10-K flags cost/unpredictable-cost risk. (I) Tariff exposure cuts both ways: import-cost pressure on national brands vs. potential pricing-cover for domestic inventory — net effect unquantified (OQ).

Do brands matter? (F) Exclusive/private brands (Antonio Melani, Gianni Bini, Daniel Cremieux, Roundtree & Yorke) are 22.3% of sales (FY2025, declining from 23.5% FY2023). (I) They are a margin lever, not captivity: exclusivity reduces direct price-comparison and supports the ~40%+ gross margin, but no customer declines Amazon because of Gianni Bini. National brands matter as traffic-pull — which is precisely why brand DTC migration hurts.

What is the nature of competition? Customers’ switching costs? (F) Switching costs ~zero: the customer is one tap from Amazon, one exit ramp from TJX/Ross; no search costs, no lock-in of any kind. The only demand-side “advantage” is habit in an aging Sunbelt customer with a store card — and Greenwald’s caveat applies: habit fades and is not replaced generationally. (F) Traffic data corroborates: clearance-center visits +7.5% YoY (Jan–Aug 2025) while total visits were flat — the loyal base holds, the incremental customer does not arrive. (I) Competition is pure price/assortment/convenience against structurally advantaged formats; Dillard’s competes by being cheaper to operate, not harder to leave.

Moat verdict (Greenwald). (I) No durable competitive advantage of any Greenwald type — no captivity, no network effects, scale without captivity (fails the captivity requirement). One narrow, genuine supply-side cost advantage: ~93% owned square footage (no rent in a 0–5%-margin industry) plus famously lean opex — sufficient to keep FCF positive through the decline, not to arrest it. Share-stability test passes only because the pie shrinks faster than DDS does. On the standard test — a claimed advantage must be tied to a metric that would deteriorate without it — the only claim that survives is occupancy-adjusted margin, and that is a cost advantage, not a franchise. What keeps DDS profitable is operational effectiveness executed better than anyone in a barrier-free industry — real, valuable, and emulable in principle.

4. Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? (F) Owned real estate, massively. 271 stores / ~46.0M sq ft, ~43.0M sq ft owned (~93%), plus 6 distribution centers, an 850K-sq-ft internet fulfillment center, and HQ — all owned. PP&E gross $3,827M, accumulated depreciation $2,879M, net $948M — 75% depreciated. Book value is a depreciated-cost artifact, not a value measure. (F) Corroborating transactions: Willow Bend pad realized through third-party redevelopment; Longview Mall bought with Trademark for ~$34M (~$53/sq ft for a whole distressed mall); 5 properties sold FY2025 for $25.7M at a $20.4M gain. (I) The reasoned band is $1–3B ($64–192/sh) vs $948M book — floor + free optionality, but no appraisal exists (OQ) and the family monetizes glacially ($34M JV = 0.4% of market cap; no sale-leasebacks ever). Also: $1,417M cash + short-term investments at Q1-FY2026, of which the $211.5M short-term-investment component is sometimes overlooked (ROIC.ai’s net-debt line misses it (F)).

Off-balance-sheet liabilities? (F) Modest. (a) Unfunded nonqualified defined-benefit pension ~$314M (FYE25, 5.4% discount rate) — real, on-balance-sheet as a liability but easy to miss; ~$5M/qtr cost flows through other expense. (b) Operating leases are small — only 15 leased stores + 7 owned-on-leased-land out of 271; ~$36M finance leases. There is no large capitalized-lease distortion of the kind that inflates peers’ adjusted debt. © $25.3M letters of credit against the revolver. (d) Purchase obligations $1,254.8M (merchandise + store construction) due FY2026 — routine retail commitments. (e) Mall JV interests $34.3M (equity method, other assets); JV terms/ownership % undisclosed (OQ).

How conservative is the accounting? (I) Conservative on every axis tested. (F) Net cash ~$515M FYE25 → ~$862M Q1-FY2026 ($1,417M cash+STI vs ~$558M total debt: $321.7M unsecured 7.00–7.75% notes + $200M 7.5% subordinated debentures due 2038 + ~$36M finance leases); $800M revolver undrawn (no borrowings FY2023–25, matures 2030-03-12, covenant-lite); interest income $47.3M > interest expense $41.0M; FCF conversion >100% of net income every recent year; no SBC distortion; Altman Z 8.9; current ratio 2.65x. (I) Net interest income, 75%-depreciated PP&E, and no capitalized games means reported numbers understate financial strength. (F) Items to strip for a clean run-rate: (1) Q1-FY2026 $104.1M pretax / $5.10-per-share interchange-litigation settlement gain (~14% of FY2025 pretax — genuinely one-time); (2) recurring ESOP special-dividend tax benefits (+$16.3M to +$35.0M/yr; $2.24/sh in FY2025) that flatter the ETR (17.9% FY25; company guides ~23% FY2026 — a ~500bps structural step-up); (3) recurring-but-lumpy property-sale gains ($20.4M FY2025, $1.00/sh). Clean basis: core FY2025 EPS ≈ $33.2; ex-gain TTM EPS ≈ $37.0.

How CapEx-hungry is the business? (F) Not at all: capex $93.4M FY2025 (~1.4% of sales) ≈ ~0.5× D&A (~$181M), trending down ($137M FY2018 → $93M); zero new stores opened FY2025; store count stable (271–272). (I) This is deliberate harvest mode — remodel selectively, build nothing — rational for the format; it also means reported earnings are not propped up by underinvestment beyond what the format’s decline justifies, but equally that no capital is being deployed toward a turnaround that isn’t coming.

5. Capital Allocation & Management

How much FCF does the business generate? (F) FY2021 $1,176M (flattered by a $242.5M working-capital release) → $828M → $751M → $610M → $624M (FY2025); TTM $755M reported, ~$650M ex-settlement. Normalized band used in the valuation work: $600–650M, central $625M. FCF conversion >100% of net income in every recent year.

How does management use it — what is the philosophy? (F) The stated priority order (10-K MD&A): “strategic investments to enhance the value of existing properties, stock repurchases and dividend payments.” The observed record (F): 83% of 5-year FCF returned ($3.32B of $3.99B), entirely self-funded — cash+STI rose to $1,073M while debt stayed ~$522M. (I) The philosophy is a textbook declining-industry cannibal playbook: harvest, return everything, empire nothing — run for family survivability through any retail winter, not for ROE maximization.

Significant acquisitions recently? (F) None of consequence. The only new use of cash is mall joint ventures: $34.3M contributed FY2025 (the Longview Mall, TX purchase with Trademark Property Co., ~646K sq ft for ~$34M, DDS as anchor) — 0.4% of market cap, $0.4M equity earnings. (I) Rationale is anchor-protection and control of the mall’s fate at distressed pricing; the watch-item is creep — a growing mall-ownership hobby would be the first deviation from the harvest playbook (and simultaneously a signal the family sees RE value > retail value). No transformational M&A, ever.

Buying back shares? (F) This is the core strength. Share count (A+B): ~113.8M peak (FY1996) → 15.617M today — −86% over ~30 years; −54% over 10 years; −31% over 5. Cash deployed: $1.508B in FY2022–26 for ~6.4M shares at a blended ~$236/sh vs $550.93 today (~2.3× on repurchased capital). The pattern is price-sensitive, not mechanical: ~$500M/yr at $175–255 (2021–22), tapering at $307–367, zero since ~August 2025; $165.2M remains on the May-2023 $500M authorization. The return engine has shifted to special dividends as the price rose. (I) The one miss: they conserved cash at the 2020 COVID bottom ($40–60) rather than buying — defensible solvency-first behavior for a family whose wealth is the company. Note (F): buybacks retire Class A only, mechanically concentrating family control with every repurchase.

Issuing large amounts of new shares to insiders? (F) No. No options exist at all; the Stock Bonus Plan grants 6% of cash comp in stock with no vesting (tiny, non-dilutive in practice); total share count falls every year. The June 2026 Form 4 wave (~35 filings) was routine annual stock-bonus grants plus WDC-merger bookkeeping — zero open-market transactions, zero 10b5-1 plans in the entire 60-month corpus, and zero open-market sales by any family member or entity. The only conviction trade in five years was a non-family director (Warren Stephens) buying ~$4.6M at ~$228–231 in Feb-2022.

Compensation policy of directors/management? (F) Formula-based and profit-linked: cash bonus pool = 1.5% of pre-tax income + 3.5% of the YoY pre-tax income increase (individual cap 1% of pretax; paid only if pretax > 0). FY2025 pool $10.42M on $694.5M pretax. CEO WTD II total comp $9.10M (of which $4.43M is a pension-value swing; direct comp ~$4.7M, 62.5% bonus); family base salaries frozen ≥3 years ($1.17M CEO/President). No severance/CIC agreements. (I) Comp is small and tied to profitability, not vanity growth — owner-grade. (F) The disclosed friction: ~6 additional family members on payroll at ~$5–6M/yr aggregate (Mahaffy $1.56M, WTD III $1.64M, Lucie $1.10M, Jazic $1.10M, Hobbs $0.59M, Banks $0.16M) plus related-party vendor payments (Connor Group $3.42M to a director’s firm; Stephens Insurance ~$1.1M) — real “jobs program” evidence but <1% of SG&A; no family store leases, no self-dealing M&A.

Motivations of management? (I) Family compounder, not empire-builder. (F) The family holds ~51% of total shares economically (directors+officers 34.7% of Class A; Class B ~entirely family) with Class B electing 2/3 of the board; they never sell; they chose special dividends over personal liquidity; the balance sheet is run for survivability; the WDC merger + Voting & Exchange Agreement (bloc voting, lineal-descendant-only transfers) exists to preserve controlled-company status across generations; the Delaware→Texas reincorporation (effective 2025-08-31) weakens minority recourse. (I) Alignment on economics is excellent — the family eats its own cooking and its returns have been minority holders’ returns. Alignment on control/exit is nil — minorities are permanent passengers on a well-driven bus; the dual-class discount is structural and permanent. Say-on-pay (98–99% approval) is non-informative because the family votes half the shares.

6. Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? (F) No — ordinary common stock. DDS is the Class A common of Dillard’s, Inc. (NYSE-listed, US domestic issuer, Texas corporation since 2025-08-31). (F) There is a dual-class structure: Class A (11,630,838 sh, public) elects 1/3 of the board; Class B (3,986,233 sh, ~entirely Dillard family) elects 2/3; one vote per share each, identical economics, B convertible 1:1 into A. NYSE “controlled company.” (I) Practical consequence: the investable float is ~7.3M shares ex-ESOP/insiders — the mechanical root of the squeeze dynamics. The $200M Dillard’s Capital Trust I 7.5% subordinated debentures (2038) are a separate trust-preferred instrument, not the common.

Dividend policy? (F) Small regular quarterly + large board-discretion specials. Regular: $0.30/q (raised from $0.25 on 2025-11-20; ~$19M/yr, ~0.2% yield). Specials: $15 (Dec-2021), $15 (Dec-2022), $20 (Jan-2024), $25 (Jan-2025), $30 (Jan-2026, ~$473M, largest ever) — five in five years, ~$1.73B total (per the company’s primary earnings releases). TTM shareholder yield ~5.8%, nearly all special. (I) Specials are ESOP-tax-advantaged (the ESOP-dividend deduction generates $16–35M/yr of tax benefits) and function as the return-of-capital valve now that buybacks have stopped. No filing commits to recurrence (OQ) — a sixth special (~Nov-2026) is board discretion; discontinuation would remove the ~5% yield floor and signal a shift in the family’s own valuation read.

How profitable is the business? (F) FY2025: revenue $6,563M, gross margin 40.3%, op margin 10.5%, net income $570M, EPS $36.42 (core ≈ $33.2), ROIC 23.4%, ROE ~32% (buyback-inflated), FCF $624M. TTM: revenue $6,605M, reported EPS $42.11, ex-gain EPS ~$37.0, EBITDA ~$870–880M. At $550.93 (2026-07-17): market cap $8.60B, net cash $862M, EV ~$7.74B; P/E ~14.9x ex-gain (13.1x reported), EV/EBITDA ~8.8–9.1x, EV/EBIT ~11.1x, FCF yield ~7.6–8.8% (ex-settlement vs reported). (F) Own-history: composite 92.8th percentile (P/E 90.4th, P/S 92.3rd, P/B 95.7th — the last structurally broken by $5.46B of treasury stock against $1.78B book equity); EV/EBITDA ~2x its own pre-COVID 4.1–5.0x. (I) A premium to Macy’s (5–7x) and Kohl’s is deserved on quality (net cash, owned RE, real merchandise margin, survivability); the open question is whether the premium — wider than it has ever been, ~1.8x the last control transaction (Nordstrom ~4–5x EBITDA) — is also a float-scarcity artifact (see Question 1.5).

Is net income diverging from cash from operations? (F) No — conversion is >100% and has been every recent year. FY2025: NI $570M vs OCF $717M; TTM: NI $657M (incl. the settlement) vs OCF $848M; 5-year cumulative FCF $3.99B comfortably exceeds cumulative NI. Working capital is tight (CCC ~59 days, inventory turns 2.6x) with no large one-time WC inflation in the current run-rate. (I) The only historical distortion ran the other way (FY2021 OCF flattered by a +$242.5M payables rebuild — not repeatable). The divergence question that matters here is not NI-vs-CFO but reported-vs-core earnings: the $104.1M settlement gain, ESOP tax benefits, and property-sale gains (Section 4 of this questionnaire) — strip those, not the cash flow.

7. Risks & Downside

What factors would cause the stock to decline? (Ordered roughly by weight of evidence; each labeled.)

1. Melt acceleration (I, the fundamental base case for shorts): comps revert to the channel rate (−3 to −5%/yr) as demand exit continues to outpace capacity exit; one quarter of +3% comps is currently carrying the entire multiple. Falsifier/confirmation: the Q2-FY2026 print (~Aug-2026).
2. Gross-margin reversion (I): the +650bps retained GM proves cyclical; GM slides toward 37–38% (or the pre-COVID 33–34%), and with SG&A already round-tripped, op margin compresses to ~6–7.5% (or lower) — EPS toward the mid-$20s or below, and the premium multiple loses its anchor. This is the bear’s load-bearing assumption; three years of decelerating erosion plus Q1-FY2026 are genuine evidence against it so far.
3. Multiple/positioning reset (F/I): the stock trades at the 92.8th percentile of its own history and ~1.8x the Nordstrom control price; the H2-2025 melt-up was partly short-covering. A normalization toward peer multiples on stable earnings alone implies a large drawdown — and with ~35% short interest, 10–19% single-day gaps are routine in both directions (F: −19.4% 2021-11-30, −17.2% 2022-05-18, −17.1% 2023-02-21, −7.8% on the FY2025 print). Squeeze-driven dislocation cuts both ways: the same microstructure can overshoot upward beyond any fundamental path.
4. Credit-card income decline (F): card-alliance income $67.2M → $39.6M in two years (−41%), ~5.7% of pretax at ~100% margin; the Wells Fargo→Citibank transition (Sep-2024) reset the program. Q1-FY2026 service-charge income +12% YoY is one hopeful data point, not a trend (OQ: Citi alliance economics/trajectory).
5. Mall-real-estate value impairment (I): the RE band ($1–3B) is reasoned, not appraised; the Longview datapoint ($53/sq ft for a whole mall) says the B/C tail may be worth little beyond store FCF. If the market re-marks the RE optionality down, the SOTP high end (~$549/sh) collapses toward the low (~$322/sh). Mall-JV expansion is also a capital-allocation yellow flag.
6. Tariff/consumer exposure (F/I): DDS traded as a high-beta tariff proxy in April 2025 (−19% peak-to-trough around “Liberation Day”); management’s FY2025 commentary flagged “unpredictable costs.” Import-cost inflation on national brands and a discretionary-spending downturn both hit the GM floor the thesis rests on.
7. Key-person / succession (F/I): CEO William T. Dillard II is ~80. The WDC merger (estate planning) and the Voting & Exchange Agreement (next-generation voting lock: WTD II/Alex/Mike as Authorized Representatives) have pre-engineered control continuity — so succession is an estate event rather than a control vacuum — but any estate-driven distribution, tax-motivated share movement, or change in the return-of-capital posture post-succession is a genuine, unpriceable event risk (A).
8. Tax-rate step-up (F): ETR guides from 17.9% (FY25) to ~23% (FY26) — a ~500bps structural EPS headwind now that special-dividend ESOP deductions are lapped; worth ~$2–3 of EPS by itself.
9. Buyback-pause signal (F/I): zero buybacks since Aug-2025 means the most reliable marginal buyer of the float is gone at current prices; if the special dividend were also skipped (~Nov-2026 decision), both return engines would be off simultaneously.

Risk of a catastrophic loss? Chance of a total loss? (I) Total-loss risk is low — near the floor for equities. (F) Net cash ~$862M (Q1-FY26), an undrawn $800M revolver to 2030, ~93% owned real estate (no rent, no landlord, no lease expiry), no pension blowout ($314M unfunded is manageable), no covenant exposure while availability >$80M, and a management family whose entire wealth rides on survival and which conserved cash even at the 2020 bottom. There is no leverage-driven path to zero, and the owned boxes retain alternative-use value even in store-closure scenarios (Willow Bend). (I) The realistic bear is a large drawdown, not zero: the melt-path DCFs in Section 10 of the main report cluster at ~$328–436/sh and the bear scenario zone at ~$220–350/sh — i.e., a ~−40–60% multiple-and-margin reset if the GM step-up proves cyclical and the premium multiple collapses toward where these assets actually transact (4–5x EBITDA). Precedent for violence exists: lifetime max drawdown −92.9% (GFC era), −49.6% max DD over the trailing 5 years (F). Catastrophe here means owning a melting annuity bought at a stabilized-annuity price — severe but bounded by net cash, the real-estate floor, and the family’s demonstrated willingness to return rather than destroy capital.

8. Recent News & Events

Has the business environment changed recently? (F) Yes — the first stabilization evidence in three years: Q1-FY2026 (reported 2026-05-14): comps +3% (FY25 0%, FY24 −3%, FY23 −4%), retail gross margin 45.8% (+30bps YoY), net sales +2.6% YoY — the strongest print since the normalization began, though the headline EPS $16.04 includes the one-time $104.1M pretax payment-card interchange litigation settlement ($79.6M AT, $5.10/sh; 10-Q Note 9). Ex-gain, Q1 pretax was ~$223M (+4.3% YoY). (F) The tape gave the print no traction (stock −21% off the Dec-2025 high) — the market is not yet paying for stabilization. (OQ: whether further settlement installments remain — a second gain would distort TTM multiples again.)

Significant acquisitions? (F) None in the M&A sense. Two asset-level moves: the Longview Mall JV (Aug-2025, with Trademark Property Co., ~646K sq ft, ~$34M from Washington Prime — first mall-ownership JV; $34.3M contributed to mall JVs in FY2025, equity method) and the opportunistic Dayton, OH store opening replacing a Macy’s anchor. Willow Bend (Plano, TX) closed Jan-2026 as that mall redeveloped — pad value realized through the process.

Change in accounting policies? (F) None found in the filing corpus. KPMG remains auditor (ratified 2026-05-28). The quality-of-earnings flags are item distortions, not policy changes: the $104.1M settlement gain, recurring ESOP special-dividend tax benefits (+$16–35M/yr), recurring property-sale gains, and the guided FY2026 ETR step-up to ~23% (see Sections 4 and 7 of this questionnaire). CDI Contractors uses cost-to-cost accounting (estimate-revision risk noted in the 10-K) — immaterial at ~4% of sales and $3.5M pretax.

Recent changes — new markets, facilities, management? (F) A dense corporate-event cluster in 2025–2026, in date order:

• Apr-2025: Director Warren Stephens resigned from the board (confirmed US Ambassador to the UK) — removes the board’s most prominent outside financial mind.
• Jul-2025: Poison-pill remnant eliminated (Certificate of Elimination for the Series A Junior Participating Preferred; the rights plan had expired 2012, never issued) — housekeeping; the dual-class makes a pill redundant anyway.
• Aug-2025: Reincorporation Delaware → Texas (special meeting 2025-08-19; effective 2025-08-31) — part of the post-Tesla move to a management-friendlier jurisdiction; weakens minority recourse (I).
• Nov-2025: Record $30.00 special dividend declared (paid 2026-01-05, ~$473M) + quarterly raised to $0.30; fifth consecutive annual special.
• 2024–ongoing: Credit-card program transition Wells Fargo → Citibank (WF terminated Sep-2024; new long-term Citi alliance) — card income down 41% over two years; Q1-FY2026 service-charge income +12% YoY is the first possible stabilization signal.
• Since ~Aug-2025: Buyback pause — $9.8M repurchased TTM; $165.2M authorization unused; return-of-capital fully pivoted to specials.
• Mar–Jun-2026: WDC merger — family holdco W.D. Company (holder of 99.99% of Class B) merged into DDS (signed 2026-03-20; approved 2026-05-28 by 14,199,181 for / 28,127 against; closed 2026-06-04), share-for-share and economically/voting neutral for public holders; a tax-free reorganization under IRC Section 368 with an IRS private letter ruling; stated purpose = administrative simplification + estate planning. Simultaneous Voting & Exchange Agreement locks recipient family Class B into bloc voting (majority of WTD II/Alex/Mike), irrevocable proxies, and lineal-descendant-only transfers — expressly to preserve controlled-company status across generations (F). Post-merger 13D: WTD II 30.0% of Class A (F). (I) Read together with the CEO’s age (~80), Texas reincorporation, and the pill cleanup, the 2025–26 cluster is a coherent estate-and-succession program: entrenchment extended to the next generation, not a prelude to sale.
• (F) No management turnover at the operating level; no new markets (footprint stable at 271–272 stores / 30 states); the June-2026 Form 4 wave was merger bookkeeping + routine grants, with zero open-market insider transactions.

Open Items Carried Forward (OQ register)

1. Durability of the +650bps retained gross margin — one quarter of favorable evidence (Q1-FY2026); Q2 print ~Aug-2026 is the next test.
2. Real-estate market value — the $1–3B band is reasoned; no appraisal exists; Longview JV ownership %/economics and whether the $34.3M covers assets beyond Longview are undisclosed.
3. Sixth consecutive special dividend (~Nov-2026) — board discretion; no filing commits.
4. Interchange settlement — which case/network and whether further installments remain (10-Q legal note not fully detailed).
5. Citibank card-alliance economics and trajectory post-transition.
6. Class B paper trail — recipient Form 3s post-WDC merger not yet filed/observed.

APPENDIX B — Source Appendix

Dillard’s, Inc. (NYSE: DDS) — Source Appendix

Report date: 2026-07-18 · Coverage type: INITIATION (first coverage of DDS by the author) · CIK: 0000028917 · CUSIP: 254067101 Access date: all sources accessed 2026-07-18 unless otherwise noted. Prices as of 2026-07-17 close ($550.93) unless otherwise noted.

This appendix contains no price target and no BUY/SELL recommendation.

1. Primary — SEC filings (EDGAR, CIK 0000028917)

1.1 Current primaries (the two documents this report is built on)

Key content relied on: 10-K MD&A KPIs, segment note, service-charges/card-income table, debt & liquidity notes, Item 2 Properties (owns 43.0M of 46.0M sq ft, ~93%); 10-Q income statement, Note 9 ($104.1M pretax payment-card interchange litigation settlement gain; $79.6M after tax, $5.10/sh), Note 8 (repurchases).

1.2 Prior annual reports (FY2021–FY2024)

1.3 Proxy statements

Comp formula sourced from the DEFM14A (Senior Management Cash Bonus Plan: pool = 1.5% of pre-tax income + 3.5% of YoY pre-tax income increase; FY2025 pool $10.42M on $694.5M pre-tax income). Family ownership (directors+officers 34.7% of Class A; WDC 99.99% of Class B; Newport Trust/ESOP 38.8% of Class A; 11,630,838 A + 3,986,233 B outstanding at 2026-03-30 record date) and the FY2025 related-party schedule (6 additional family members on payroll ~$5–6M/yr aggregate; Stephens Insurance ~$1.07M; Connor Group $3.42M; no family store leases or family real-estate JVs) also per the DEFM14A.

1.4 Key 8-Ks (event-driven; the full 34-filing 8-K corpus for 2021–2026 was reviewed)

Special-dividend scorecard, from earnings-release exhibits + XBRL PaymentsOfDividends (correcting an early secondary-source discrepancy): $15 (Dec-2021), $15 (Dec-2022), $20 (Jan-2024), $25 (Jan-2025), $30 (declared 2025-11-20, paid 2026-01-05) ≈ $1.73B total. Quarterly dividend $0.20 (2022) → $0.25 (early 2024) → $0.30 (declared 2025-11-20).

1.5 Insider filings — Form 3/4/5 corpus (60 months, parsed in full)

989 Form 3/4/5 filings were fetched from EDGAR and parsed in full: 1,050 non-derivative transactions: 920 grants (A), 106 gifts (G), 15 open-market sales (S), 5 open-market buys (P), 4 WDC-dissolution entries (D). Zero 10b5-1 flags on any transaction.

• Zero open-market purchases and zero open-market sales by any Dillard family member or family entity in 60 months (FACT). Family share movements are grants, gifts (estate/charitable), and WDC-merger bookkeeping.
• Only conviction-size trade in the corpus: director Warren A. Stephens bought 20,000 sh @ ~$228–231 (~$4.6M) on 2022-02-23 (Form 4, 2022-02-23; in corpus) — after the stock had already ~3x’d off the squeeze base.
• All 15 sales are non-family and tiny (10,236 sh / ~$4.10M total; largest 2,300 sh — Tony Bolte, Nov-2022; J.C. Watts Jr. 7 lots 2021–2026 at $320→$609). Nobody sold into the 2021–22 squeeze except token non-family lots.
• June-2026 Form 4 wave (~35 filings, 2026-06-05/30): routine June annual stock-bonus grants + WDC-dissolution entries — no market transactions.
• Individual Form 4s are accession-indexed on SEC EDGAR (986 Form 4/4A filings in the 60-month window).

1.6 Form 144 (13 notices, 60 months)

All small and non-family or non-common: J.C. Watts Jr. ×7 (~4,450 sh total, 2024–2026); Tom Bolin ×2; Mike Litchford (2,660 sh); James I. Freeman + Freeman Charitable Trust (2 + 3,500 sh); Chris B. Johnson (400 sh, Sep-2025); Annemarie Jazic (Feb-2025 — Dillard’s Capital Trust I units, not common stock). Trivially small sale-notice volume. Accession-indexed on SEC EDGAR.

1.7 Beneficial ownership and other filings

1.8 XBRL company facts

SEC XBRL company-facts API, CIK 0000028917: https://data.sec.gov/api/xbrl/companyfacts/CIK0000028917.json (pulled 2026-07-18). Tags relied on: Revenues, NetIncomeLoss, NetCashProvidedByUsedInOperatingActivities, PaymentsToAcquirePropertyPlantAndEquipment, PaymentsForRepurchaseOfCommonStock, PaymentsOfDividends. Reconciliation: exact tie to ROIC.ai aggregated data for FY2018–FY2025 — see Section 5 of this appendix.

2. Quantitative cross-check feeds (third-party, reconciled to filings)

3. News / press (validated at source)

4. Frameworks

• Greenwald, Competition Demystified — moat taxonomy (demand-side captivity, supply-side cost advantage, economies of scale plus captivity), the share-stability and ROIC empirical tests, and the operational-effectiveness resolution for advantage-free industries. Basis for the moat verdict in Section 4 of the main report (no durable advantage; one narrow supply-side cost advantage — owned real estate plus lean operations).
• Marathon/Chancellor, Capital Returns — capital-cycle positioning of a declining industry (deep capital-exit phase; technology disruption breaking the normal cycle’s completion). Basis for the industry read in Section 3 of the main report.

5. Reconciliation & caveats

1. ROIC.ai vs EDGAR XBRL: revenue, net income, operating cash flow, and capex tie EXACTLY for FY2018–FY2025. ROIC.ai EPS uses rounded share counts — filing EPS quoted instead ($36.42 FY2025; $16.04 Q1 FY2026). ROIC.ai’s displayed ROE (~9%) is inconsistent with the filings-implied figure (true ROE ~32%) and is quoted nowhere in the report. ROIC.ai’s net-debt line ignores $211.5M of short-term investments — net cash figures were rebuilt by hand ($515M FYE25; $862M at Q1-FY2026).
2. $104.1M one-time gain stripped: Q1-FY2026 includes a $104.1M pretax / $79.6M after-tax ($5.10/sh) payment-card interchange litigation settlement (~14% of FY2025 pre-tax income). All TTM multiples in the report use ex-gain TTM EPS ≈ $37.0 (reported $42.11); FCF likewise quoted both ways ($755M reported / ~$650M ex-settlement). ROIC.ai nets the gain into “other non-op” — labeling obscures; strip manually.
3. ESOP-dividend tax benefits flagged non-operational: +$16.3M (FY2022), +$21.1M + $7.3M (FY2023), +$30.8M (FY2024), +$35.0M (FY2025) — recurring but scaling with the special dividend, not operations; stripped in core-EPS work (core FY2025 EPS ≈ $33.2 after also stripping $1.00/sh property-sale gains). FY2026 ETR guides to ~23% vs 17.9% FY2025.
4. Fiscal-year convention: FYE = Saturday nearest Jan 31. FY2025 = 52 weeks ended 2026-01-31; FY2023 was a 53-week year. ROIC.ai labels FY2025 as “FY2026” — company labels used throughout the report.
5. FactorsToday annualization: leaderboard returns are annualized at all horizons; 3m/6m figures were de-annualized and cross-checked against the price series before quoting.
6. Short interest: current figures (35.25% of float; 9.26 DTC; settlement 2026-06-30) are from yfinance — a third-party aggregator, flagged as such. The 2021–22 squeeze-era series (~30%+ of float) is approximate and press-consistent, not from a pulled FINRA/exchange series — treated as INTERPRETATION-grade context, not FACT.
7. Special-dividend amounts were checked against primary earnings releases: $15 (Dec-2021), $15 (Dec-2022), $20 (Jan-2024), $25 (Jan-2025), $30 (Jan-2026) — an early secondary-source figure ($25/$25 for 2022/Jan-24) was superseded by the primary releases.
8. FY-labeling note on buybacks: buyback cash figures use ROIC.ai FY labels with company-FY mapping noted; XBRL PaymentsForRepurchaseOfCommonStock cross-check confirms the series.
9. No earnings calls: Dillard’s holds no regular earnings conference calls (press-release-only disclosure). No transcript sources exist to cite — itself noted as a transparency input to the analysis.

Unverifiable / open items

• Longview Mall JV: DDS ownership %, capital account, and whether the $34.3M “mall joint ventures” line covers assets beyond Longview — not itemized in the 10-K.
• Interchange settlement: which network/case and whether further installments remain — the 10-Q legal note is not fully detailed.
• 2026-06-15 −7.3% price drop: no discrete catalyst found (no 8-K, no company release); GuruFocus frames as valuation-driven unwind — unattributed.
• Real-estate market value: no appraisal exists; the $1–3B SOTP band is reasoned (book $948M, 75% depreciated; Longview $53/sq-ft distressed datapoint), not measured.
• Class B paper trail: no Class B non-derivative transactions appear in the Form 4 corpus (all Class A); the 3,985,758 Class B issuance sits in the 8-K/Form D/13D only — recipient Form 3s may be pending.