RE: AI modelling25 Feb 2026 10:55
Its all about the prompt, AI has a history of confirmation Bias. Its a great tool but its results are only as good/bad as the prompt for example my prompt was for it to look bullish but realistic:
In “bullish but not totally fantasy” land, a 2026 takeout around £1–2bn enterprise value is about as far as I’d stretch it this year – which, on today’s share count, translates very roughly to something like:
£1bn → ~£2.30/share
£1.5bn → ~£3.40/share
£2bn → ~£4.60/share
vs today’s market cap of ~£270m at ~60p.
The £3–4bn+ (£6.8–£9+/share) stuff people talk about is more “later, with much more data / competition for the asset” than “this year”.
Let me explain how I get there in your bull case:
Good AVA6000 breast/TNBC readout + SGC survival still strong
Early AVA6103 human data look in line with / better than the mouse data
Big pharma with exatecan / ADC interests feels “we need this platform”
1. Where Avacta actually is on the food chain
Right now Avacta is:
AVA6000 – Phase 1b:
90% disease control in SGC, encouraging PFS, cardio safety strong, dose limit relaxed, path to dose selection for efficacy studies.
TNBC & other cohorts reading out in H1 2026.
AVA6103 – just cleared IND, Phase 1 starting Q1 2026:
Very nice preclinical pharmacology,
AI analysis vs Enhertu suggesting quicker, higher, more selective tumour delivery in mouse models.
No late-stage trials, no approvals, no human 6103 data yet.
Compare that to:
Seagen, which Pfizer bought for $43bn: fully commercial ADC company with multiple marketed drugs and late-stage pipeline.
Enhertu deal (AZ/Daiichi):
AstraZeneca paid $1.35bn upfront, plus up to $5.55bn in milestones (total up to $6.9bn) for co-dev & commercial rights to a late-stage exatecan ADC.
Other hot oncology buys: Gilead buying Arcellx for up to $7.8bn for a late-stage CAR-T platform.
Those deals are for much later-stage, de-risked assets.
Avacta today is more like:
“Platform with strong early human PoC in one asset (AVA6000) and very promising preclinical + just-starting Phase 1 in a second (AVA6103).”
So any 2026 takeout is early / opportunistic, not “we’re buying a mini-Seagen”.
2. What would have to happen to get a serious bid this year?
In your bullish scenario, this year we’d need something like:
SGC & TNBC look undeniably good
SGC PFS/OS continues to beat historical controls with more follow-up.
TNBC (and maybe breast) data show real, reproducible responses, not just noise.
AVA6103 human data don’t break the magic
Phase 1 runs clean:
Safety within expectations, no unexpected dose-limiting toxicity pattern.
PK in humans broadly matches the mouse/AI story (strong tumour exposure, limited plasma).
Maybe even early tumour shrinkage in at least one of the four indications (pancreatic/gastric/SCLC/cervical).
Big pharma really cares about owning this mechanism. Particularly those with exatecan/ADC exposure:
AstraZeneca / Daiichi Sankyo – co-owners of Enhert