DNUT · Krispy Kreme, Inc. · Stock research

once# DNUT — Krispy Kreme, Inc.

Current Thesis

The last read closed with the observation that the 2026-08-06 Q2 binary had passed and nothing company-scheduled sat inside 30 days. That is still true of the fundamentals — and the tape moved anyway. The stock closed $3.59 on 2026-08-21 against $3.33 on 2026-08-14, RSI(14) at 69.9 versus 53.4 a week earlier, and 23.6% below the $4.70 52-week high with a three-month price change of +9.8%. Two dated items sit under that move: Bank of America raised its target to $4.90 from $4.20 on 2026-08-10 while maintaining Buy, and the Pokémon 30th-anniversary doughnut collection went on sale in participating U.S. shops on 2026-08-18 (advance ordering from 2026-08-13, scheduled through 2026-09-13).

The narrative leg being bought has not changed: a company-operated doughnut manufacturer converting itself into a royalty collector, with franchise share of system-wide sales about 42% at Q2 against a 50% year-end 2026 target, and net leverage at 5.4x versus 6.7x at Q4 2025 and 7.5x a year earlier — already inside the company's own sub-5.50x year-end target. What changed in the past week is attention, not arithmetic. A limited-time offer that ends 2026-09-13 and a single target revision are the whole of the new information, and the next company-controlled datapoint is the Q3 report, estimated around 2026-11-05.

The narrative is accelerating, dated to the 2026-08-18 collection launch, the 2026-08-10 BofA revision and the 2026-08-21 close of $3.59 reclaiming the $3.50 strike that dominated August open interest. The acceleration is in the attention leg. The deleveraging leg — five quarters of the same refranchising mechanic — is well known and moderating, and the two can decouple quickly once the promotional window closes.

Bullish and bearish views on Krispy Kreme, Inc.

The model's bull view on Krispy Kreme, Inc. (DNUT), in brief: Net leverage 5.4x at Q2-end (2026-08-06 release) versus a stated year-end 2026 target below 5.50x and 7.5x in the prior-year quarter. The bear view: The promotion is a September expiry, not a business change. Both cases follow in full.

Bull Case

• Net leverage 5.4x at Q2-end (2026-08-06 release) versus a stated year-end 2026 target below 5.50x and 7.5x in the prior-year quarter. On a sub-$1B-cap equity carrying that debt load, each turn removed accrues disproportionately to the equity stub.
• Q2 net revenue $331M against a $303.35M consensus, with organic revenue roughly flat and system-wide sales +2.6% excluding the lapsed McDonald's business — the -13% reported decline is the arithmetic of selling Japan and the Western U.S. joint venture into franchise agreements.
• Fourth consecutive quarter of margin expansion: consolidated Adjusted EBITDA $28.8M, +43% YoY, an 8.7% margin against 5.3% in Q2 2025; U.S. segment margin about 8.0%, up roughly 370bps.
• Cash conversion inflected: H1 2026 free cash flow -$6.1M against -$107M in H1 2025, a $101M swing, with FY capex guided $50–60M and FY free cash flow above $15M.
• Bank of America raised its target to $4.90 from $4.20 on 2026-08-10 with a Buy rating — the first upward revision after the print, against Evercore ISI's 2026-08-07 cut to $4.00 with In-Line maintained. Coverage is thin enough that one revision visibly moves the quoted aggregate.
• The Pokémon collection (six doughnuts, collectible cups, launched 2026-08-18) is a marketing-driven traffic event that costs the balance sheet nothing at a moment when capex is being rationed to $50–60M for the year.
• Franchised unit growth continues without company capital: new agreements announced for the Netherlands, Estonia and Mauritius at the Q2 presentation, and 100+ shop openings planned for 2026, nearly all franchised.

Bear Case

• The promotion is a September expiry, not a business change. The collection is scheduled to run only through 2026-09-13; a limited-time offer that pulls traffic forward leaves an easier comparison this quarter and a harder one next.
• Reach shrank ahead of the promotion: Krispy Kreme exited roughly 2,400 McDonald's outlets, reported as about 13.3% of its former worldwide access points. A demand event lands on a smaller U.S. distribution footprint than the one that existed a year ago.
• The bottom line is still negative — Q2 Adjusted EPS -$0.03, wider than the -$0.01 modeled on the per-share line even as revenue beat.
• The 5.4x leverage ratio is denominator-sensitive. It is struck against trailing Adjusted EBITDA guided to $140–150M for the year; a shortfall lifts the ratio with no new borrowing, and the guide has not yet been tested by a quarter without refranchising activity to lean on.
• Balance-sheet slack is minimal: current ratio 0.42, debt/equity above 2.0, dividend suspended since 2025. There is no income-holder base to absorb a drawdown.
• Aggregator targets disagree by a wide margin — a $3.73 median across 14 analysts on stockanalysis.com against a $4.45 three-analyst average shown by Google Finance (both retrieved 2026-08-23). The dispersion is a function of how few real estimates exist, and a reader should treat any single "consensus" number here as provider-specific.

Setup & Price Structure

Price at $3.59 on 2026-08-21 is 23.6% under the $4.70 52-week high. The structure that matters is the shelf built after the 2026-08-06 print: the stock spent the following week in the low-$3.30s, then broke above the $3.50 line into the 2026-08-21 monthly expiry. The prior note flagged 2,106 puts of open interest at the August-21 $3.50 strike against a 43-contract call sweep near the ask on 2026-08-14 with the shares referenced at $3.505; the $3.59 close resolved that strike above the line, and that particular positioning is now off the board.

Crowding and positioning observables, stated as observables:

• RSI(14) at 69.9 as of 2026-08-21, against 53.4 on 2026-08-14 — the momentum reading has travelled 16 points in five sessions and sits at the edge of the conventional overbought band.
• Short-interest estimates diverge violently by provider and by denominator: MarketBeat's mid-August compilation showed roughly 24.95M shares, about 28.93% of float, up ~17% month-over-month; stockanalysis.com shows 18.26M shares short, 10.59% of shares outstanding (retrieved 2026-08-23). These are not the same measurement, and neither settlement date is interchangeable with the other. Any squeeze framing built on the higher number should carry that caveat.
• The retail-attention channel is live and dated: Google searches for "Pokémon 30th anniversary" were reported up over 200% around the collection reveal, and the launch is the first Pokémon tie-in for U.S. Krispy Kreme customers in the franchise's 30-year history.
• No Form 4 or registration filings appear in the recent filing record reviewed for this update. JAB Holding's majority stake remains standing potential supply against a thin public float, but there is no dated transaction to point at.
• No company-scheduled event sits inside 30 days. The nearest hard dates are the semi-monthly short-interest publications and the 2026-09-13 end of the promotional window.

Fresh engagement at $3.59 is engagement into a five-session extension, with the identified demand catalyst already on shelves and an end date attached to it. That is the reason this remains a low-conviction setup on the evidence available, notwithstanding the deleveraging progress: the fundamentals will not speak again until roughly 2026-11-05, and what is moving the price between now and then is flow.

Catalyst Calendar (next 30 days)

• ~2026-09-11 (est.) — Next short-interest publication, covering the 2026-08-31 settlement date. This is the one that captures the Pokémon launch week.
• 2026-09-13 — Scheduled last day of the Pokémon Doughnut Collection at participating U.S. shops. Ends the current traffic event; nothing announced replaces it.
• 2026-09-18 — September monthly options expiry. Relevant only as flow around whichever strike accumulates open interest, in a name where a $0.50 strike interval spans 14% of the share price.
• ~2026-11-05 (est.) — Q3 FY2026 results. Outside the 30-day window and named here because it is the only scheduled company disclosure between now and November: the first quarter that must show net leverage falling from 5.4x, and the $1.25–1.35B revenue and $140–150M Adjusted EBITDA ranges surviving, without a refranchising gain in the frame.

Elapsed catalysts

• ~2026-08-25 (est.) — NASDAQ semi-monthly short-interest publication for the 2026-08-14 settlement date. Shows whether the short base kept building through the print-and-promo window or began to retire. (passed 1d ago)

What Would Change Our Mind

The structure that would have to break first is the post-print shelf: the low-$3.30s held through the week after 2026-08-06 and became the launch point for the move to $3.59. Losing it says the August advance was promotional flow with no holder underneath. Concretely, a weekly close below $3.25 returns price inside the pre-print range and marks the Pokémon-window bid as spent.

Three non-price conditions would do the same work independently of the chart. First, 2026-09-13 arriving and passing with no follow-on promotion, no new franchise agreement and no September traffic commentary — the calendar then empties completely until November, and an accelerating label would no longer be defensible. Second, any revision to the $140–150M FY Adjusted EBITDA guide or the $1.25–1.35B revenue range: the entire deleveraging arithmetic is struck against that denominator, and a cut raises leverage without a dollar of new debt. Third, the ~2026-08-25 and ~2026-09-11 short-interest publications showing the share count short rising after a revenue beat and a demand event — bears underwriting the FY guide rather than the last quarter is a different, and slower, situation than the one described here.

On the other side: net leverage printing below 5.4x at the Q3 report with the FY guide intact, or a refranchising transaction that visibly funds paydown, would move the read from flow-driven to fundamentals-driven and would justify revisiting the conviction rating on its own merits.

Correlation Notes

• The name trades as a low-priced, high-leverage consumer discretionary equity, so it takes directional beta from small-cap risk appetite and from credit conditions more than from doughnut demand. With 5.4x net leverage, moves in high-yield spreads matter to the equity stub more than a comparable move in QSR same-store sales.
• The closest structural comparison set is franchisors that converted from company-operated to asset-light — the read-across is to the mechanic (royalty stream, capex reduction, leverage paydown), not to the multiple, since peers made the conversion from positions of far lower leverage.
• Promotional-LTO co-movement: the Pokémon collaboration puts the name temporarily in the same attention pool as other IP-collaboration consumer promotions. That correlation is to social-search volume and to retail flow, and it ends with the offer on 2026-09-13.
• Because the free float is thin against JAB's majority holding, and short-interest estimates run anywhere from 10.6% of shares outstanding to 28.9% of float depending on provider, realized volatility here decouples from consumer-sector indices during positioning events. Index-relative charts will understate what actually happens on squeeze days and on expiry weeks.
• Reported revenue falls partly by design as markets are refranchised, so any screen or factor model ranking on revenue growth will keep marking the name down mechanically. That is a persistent, structural source of negative flow independent of the operating result.