DNTH · Dianthus Therapeutics, Inc. · Stock research

Current Thesis

Two weeks of silence have followed the loudest fortnight of the year. Since the 2026-08-04 Q2 print and the DNTH312 announcement the same day, no company disclosure has reached the tape; the most recent DNTH-linked item located is third-party — Agomab's 2026-08-10 release proposing Dianthus president and CEO Marino Garcia to its own board. Into that vacuum the stock still marked a new 52-week high of $115.96 and closed 2026-08-21 at $111.10, 4.2% below it, with a three-month price change of +28.5%.

The narrative leg an investor is buying is unchanged from the July and early-August coverage: claseprubart (DNTH103), an antibody selective for activated C1s that shuts the classical complement pathway while leaving the lectin and alternative arms intact, carries CIDP, gMG and MMN on a single molecule, funded by roughly $1.2B of cash at 2026-06-30 with runway guided into 2030. What has changed is the calendar's shape. The first randomized-era readout, MoMeNtum in multifocal motor neuropathy, is guided to December 2026 on 46 enrolled patients against a 36-patient target. CAPTIVATE Part B still has no top-line date — the 2026-08-04 language commits only to guidance on timing by year-end 2026.

RSI(14) reads 56.6 at $111.10 against 59.5 at the 2026-08-07 close of $108.92. Measured: the higher price carries a lower oscillator. Inferred, and it is an inference: participation behind the August leg is moderating rather than broadening.

Bullish and bearish views on Dianthus Therapeutics, Inc.

The model's bull view on Dianthus Therapeutics, Inc. (DNTH), in brief: Structure held through a news vacuum. The 52-week high advanced from $112.64 (pre-2026-08-08) to $115.96 with no fresh company disclosure in between — the August re-rate has not given back the July breakout zone (2026-07-09 high $102.24, close $100.03). A readout now exists… The bear view: The mechanism question is unresolved. Sanofi's riliprubart, the same selective anti-activated-C1s approach, had MOBILIZE halted on 2026-06-10 for insufficient efficacy in refractory CIDP. Dianthus's GO decision rests on open-label responder data with no placebo arm; only Part B… Both cases follow in full.

Bull Case

• Structure held through a news vacuum. The 52-week high advanced from $112.64 (pre-2026-08-08) to $115.96 with no fresh company disclosure in between — the August re-rate has not given back the July breakout zone (2026-07-09 high $102.24, close $100.03).
• A readout now exists inside 2026. MoMeNtum finished enrolling at 46 patients versus a 36-patient target, top-line "on track for December '26" per the 2026-08-04 release. Before August, the nearest randomized datapoint was undated.
• Part A interim survived restatement. 75% response rate in the first 40 CAPTIVATE Part A completers against a ≥50% GO threshold, with no related serious infections, no clinical symptoms of drug-induced lupus and no safety discontinuations (2026-08-04). The early GO itself came 2026-03-09 on 20 confirmed responders with fewer than 40 planned participants.
• Balance sheet removes financing urgency. ~$1.2B cash at 2026-06-30 against Q2 R&D of $48.7M and G&A of $13.6M; runway guided into 2030.
• Pipeline widened on 2026-08-04. DNTH312, a bifunctional fusion of claseprubart and TACI targeting aC1s plus BAFF/APRIL, is targeted Phase 1 ready by year-end 2027; DNTH212 Phase 1 healthy-volunteer data is anticipated by year-end 2026; EMERGE, the gMG Phase 3, was announced initiated 2026-06-29.
• Sell-side marks sit above spot. Evercore ISI $134 (2026-08-03), Wedbush $122 (2026-08-05), Truist $128 from $110 (2026-08-06), HC Wainwright $158 (2026-08-06).

Bear Case

• The mechanism question is unresolved. Sanofi's riliprubart, the same selective anti-activated-C1s approach, had MOBILIZE halted on 2026-06-10 for insufficient efficacy in refractory CIDP. Dianthus's GO decision rests on open-label responder data with no placebo arm; only Part B settles the comparison, and Part B has no date.
• Guidance about guidance. A November or January update dating Part B top-line in 2H'27 or later would extend the thesis by a full year without any negative clinical datapoint.
• Nothing is dated for roughly three months. The Q1 print landed 2026-05-05 and Q2 on 2026-08-04, implying early November for Q3. Between 2026-08-22 and the December window there is no company-scheduled event on the record.
• No product revenue. The $0.761M booked in Q2 2026 is collaboration income. Q2 net loss was $50.2M, or $(0.90) per share, against $31.6M and $(0.88) in Q2 2025 on 56,029,124 weighted-average shares.
• The shelf is live. The March 2026 underwritten offering priced at $81.00 — Runway into 2030 does not preclude a take-down while the stock trades near a 52-week high.
• The broker cluster is behind the move. All four target raises fall inside 2026-08-03 to 2026-08-06, after the June–July advance off roughly $82.

Setup & Price Structure

Reference levels, all from the adjusted daily series through 2026-08-21: last close $111.10; 52-week high $115.96; distance from that high -4.2%; three-month price change +28.5%; RSI(14) 56.6. The July shelf that defines the breakout sits at the 2026-07-09 high of $102.24 and close of $100.03. The 2026-08-03 inducement grants — carry a $106.02 exercise price, a dated marker of where the stock traded at the start of the month.

The narrative is maturing. The attention arrived and has already been priced: four target raises inside four sessions (2026-08-03 → 2026-08-06), a Q2 print with two new disclosures on 2026-08-04, and then nothing company-sourced for the following seventeen calendar days while price ground to a marginal new high on a lower oscillator reading. That is not the expanding-headline pattern of accelerating, and no evidence of mainstream or retail-forum saturation was found in the sources searched, so saturated is not supported either.

Crowding and positioning, stated as observables rather than a verdict: the target-raise cluster is concentrated in a single week and no raise dated after 2026-08-06 was located; the S-3 shelf remains available with the last close well above the $81.00 March take-down price; the most recent insider transactions located are CFO Ryan Savitz selling 8,224 shares for approximately $739,000 on 2026-04-09 under a 10b5-1 plan and director Paula Soteropoulos exercising 599 options at $8.44 and selling at $80.11 on 2026-03-31 — nothing newer surfaced. No earnings date falls inside the next 30 days.

Catalyst Calendar (next 30 days)

The 2026-08-22 → 2026-09-21 window contains no company-dated event on the public record. Nearest dated items, all beyond the window:

• ~2026-11-05 (est.) — Q3 2026 results. Q1 landed 2026-05-05, Q2 on 2026-08-04, so early November is the pattern-implied date. The live question is whether CAPTIVATE Part B top-line finally receives a calendar entry.
• ~2026-12-01 (est.) — MoMeNtum Phase 2 MMN top-line, 46 patients, guided "December '26".
• ~2026-12-31 (est.) — CAPTIVATE Part B top-line timing guidance, promised by year-end 2026.
• ~2026-12-31 (est.) — DNTH212 Phase 1 healthy-volunteer data, anticipated by year-end 2026.
• 2028-07-01 (guided 2H'28) — EMERGE gMG Phase 3 top-line.

What Would Change Our Mind

The structure that has to hold is the July shelf. Losing it would mean the August advance — made on a print that added no efficacy data, plus four target raises — round-tripped into the pre-breakout base with the December readout still ahead: a weekly close below $100 marks that break. Above that level the drift is intact even with an empty calendar.

Three non-price conditions carry equal weight. First, a Q3 or January update that dates Part B top-line in 2H'27 or later, or repeats "guidance expected" without a date — the duration of the entire thesis is set by that sentence. Second, the December MoMeNtum top-line coming and going without a clean primary hit; MMN is commercially small, but it is the same molecule that carries CIDP and gMG. Third, a 424B5 prospectus supplement, an ATM sales agreement or an overnight block at a discount to the prior close, which would put supply into the high rather than leaving the shelf dormant.

What would strengthen the case rather than break it: a dated Part B top-line inside 2026 or 1H'27, or clean DNTH212 Phase 1 human data landing before year-end and broadening the story off one molecule.

Correlation Notes

• Same-mechanism read-across. Sanofi's riliprubart is the closest comparator; the 2026-06-10 MOBILIZE halt in refractory CIDP is the reason the randomized Part B question is priced as a genuine binary rather than a formality. Any further Sanofi disclosure on that program moves DNTH without DNTH filing anything.
• FcRn class in CIDP. Approved FcRn agents set the competitive share ceiling in the indication; sell-side revisions to DNTH peak-sales assumptions that cite FcRn uptake would compress the terminal value independent of trial outcome.
• Clinical-stage biotech beta. No product revenue means the multiple between events is set by sector risk appetite and rates. Between 2026-08-22 and December, price action here is more likely to track XBI-type flows than anything company-specific, since the company has scheduled nothing.
• Single-molecule concentration. CIDP, gMG and MMN share claseprubart, so the three programs are not independent draws; DNTH312 and DNTH212 are too early to offset a negative randomized result.