ODD · ODDITY Tech Ltd. · Stock research

Current Thesis

The compounder narrative that carried ODDITY to a 52-week high of $63.68 is finished, and the replacement narrative — normalization of the IL MAKIAGE paid-acquisition engine — has no dated evidence behind it until 2026-09-09. Since the last note the two things that changed both cut against the recovery camp: price gave back the bounce (a 2026-07-03 close near $17.88 to $14.12 on 2026-08-07, roughly -21%), and the Q2 print was pushed to September 9 (announced 2026-07-22), five weeks later than the 2025-08-04 date used for Q2 2025. That leaves a month of no company-dated event with the stock still trading above every published post-Q1 target except Evercore's $15. Over three months the name is nearly flat (-4.5%) having travelled from roughly $14.8 down to $9.25 and back up to $17.88 — churn, not a base. RSI(14) at 38.3 is neither washed out nor trending.

Bullish and bearish views on ODDITY Tech Ltd.

The model's bull view on ODDITY Tech Ltd. (ODD), in brief: Q1 2026 (2026-06-02): revenue $197.9M, -26% YoY — inside the pre-announced ~-30% cut, the single datapoint the normalization camp anchors on. The bear view: Q2 2026 was guided down 25–30% YoY at the Q1 report. Both cases follow in full.

Bull Case

• Q1 2026 (2026-06-02): revenue $197.9M, -26% YoY — inside the pre-announced ~-30% cut, the single datapoint the normalization camp anchors on.
• Repeat orders -15% against first orders -50% (Q1 disclosure). The installed base held; management frames the shortfall as an advertising-partner auction dislocation rather than demand erosion.
• Balance sheet is not the problem: $667.4M in cash, equivalents and investments plus $350M undrawn credit at Q1 2026. A market cap in the mid-hundreds of millions against that stack limits solvency risk.
• The issuer is a buyer. Q1 2026: ~6.1M Class A shares repurchased for $82.3M, with $167.3M of authorization left. Separately, on 2026-06-12 the company retired $50M face of 0% exchangeable notes due 2030 for $35M — roughly 70 cents on the dollar, ~$15M accretive, and it shrinks the 2030 dilution overhang.
• FY2026 guide still calls for positive adjusted EBITDA despite a Q1 adjusted EBITDA of -$7.0M, which implies management expects an H2 swing.
• Prior-cycle economics were real: FY2025 adjusted EBITDA of $163M with Q4 2025 revenue +24% YoY (reported February 2026). The model produced cash before the acquisition channel jammed.
• Product pipeline is dated: Methodiq launched 2025-11-18; a fourth brand has been slated for 2026.

Bear Case

• Q2 2026 was guided down 25–30% YoY at the Q1 report. The September 9 print is not scheduled to show growth; the debate is only over the shape of the decline.
• Gross margin fell to 69.7% in Q1 2026 from 74.9% a year earlier, with a $21.4M net loss. Margin, the core of the long case, went with the cohorts.
• First orders -50% is a cohort hole that feeds forward. DTC repeat revenue is a function of prior new-customer cohorts, so a lost H1 2026 acquisition year pressures 2027 repeat revenue even if CPA normalizes tomorrow.
• Price sits above the Street. Post-Q1 targets: Barclays $8, Morgan Stanley $10, Jefferies $10.25, Truist $12, Evercore ISI $15; Goldman moved to Sell and KeyBanc to Sector Weight. Consensus target was cut from $17.33 to $10.09. At $14.12 the tape is still discounting a recovery the analyst community has not yet underwritten.
• The reporting calendar has slipped twice this year — Q1 landed 2026-06-02 against a historical early-May cadence, and Q2 moved to 2026-09-09 against 2025-08-04 for the prior-year quarter. The company has not tied the change to any disclosed problem; the observable is the slip itself.
• Single-partner dependency. The company has not named the advertising partner, so the CPA claim cannot be cross-checked against a third party's disclosures.

Setup & Price Structure

Life-cycle: DEAD. The dating is explicit: guidance suspension and the CPA disclosure in February 2026; the 2026-06-02 Q1 print with revenue -26% and first orders -50%; a drawdown to -77.8% from the $63.68 52-week high; and the failure of the summer bounce, which topped near $17.88 on 2026-07-03 and closed at $14.12 on 2026-08-07. Structure is broken in the sense the label describes — there is no base, no rising trend to hold, and the recovery attempt has already surrendered most of its gain. A turnaround narrative may eventually be written here, but nothing dated supports relabelling it before September 9.

Crowding and positioning observables, stated without a verdict:

• Distance from the 52-week high is -77.8% (basis 2026-08-07), and the 2026 low near $9.25 is about 35% below the last close — the downside reference is closer than the upside one.
• RSI(14) 38.3: no oversold capitulation reading, no momentum.
• The last close is above four of five published post-Q1 price targets ($8/$10/$10.25/$12) and below only Evercore's $15.
• Supply is being absorbed by the issuer, not sold by it: $82.3M of Q1 repurchases, $167.3M authorization remaining, and a discounted convert retirement on 2026-06-12. No insider selling or equity issuance into the bounce has been disclosed in the period covered here.
• The next binary is 31 days out, so the near-term tape has no scheduled information event to price against.

Catalyst Calendar (next 30 days)

• No company-dated event between 2026-08-09 and 2026-09-08. The window is empty; the bounce and fade have both happened on flow rather than disclosure.
• 2026-09-09 (confirmed, announced 2026-07-22): Q2 2026 results, before US market open, call at 8:30 a.m. ET. One day outside the 30-day window and the only thing that resolves anything: first-order trend, CPA direction, whether Q2 landed inside the -25% to -30% guide, and whether the FY2026 positive-adjusted-EBITDA guide survives.
• ~2026-11 (est.): Q3 2026 results. The second read, and the one that tests whether H1 cohort damage shows up in repeat revenue.

What Would Change Our Mind

The structure that has to hold is the post-crash shelf built between the $9.25 low and the July bounce; losing it re-opens the low. Concretely, a weekly close below $13 abandons that shelf and marks the recovery attempt as a failed bounce rather than a base. The fundamental condition that would break the recovery case independently: a Q2 print on 2026-09-09 showing first orders still down more than 40% YoY with no CPA improvement, or withdrawal of the FY2026 positive-adjusted-EBITDA guide.

What would force an upgrade instead: first orders inflecting to better than -20% YoY with gross margin recovering back toward the 74.9% of Q1 2025, Q2 revenue landing above the -25% end of the guide, and price reclaiming the $17.88 area on volume. Absent one of those, the September 9 print is a gap-risk event with a two-sided distribution and no informational edge available before it.

Correlation Notes

• Performance-marketing-dependent DTC cohort. The read-across runs to names whose unit economics live and die on paid-social CPA. Any disclosure from a large advertising platform about auction or algorithm changes is a direct input to ODDITY's cost line; the company has not named its partner, so this is an inference, not a confirmed linkage.
• Beauty comparables (e.l.f. Beauty, Coty and the broader mass-prestige complex) drive sentiment on category demand, but ODDITY's problem is channel-specific — its 2026 drawdown has been idiosyncratic rather than a sector move.
• Small-cap consumer discretionary beta. With a broken chart and a thin institutional target band, the name will amplify risk-off moves in the small-cap consumer complex more than it participates in rallies.
• Israel-domiciled foreign private issuer. Reporting cadence, filing format and disclosure granularity follow 20-F/6-K conventions rather than US quarterly practice, which is part of why the calendar has moved twice this year.