How we build a number you can defend
LOE Edge produces generic and biosimilar opportunity estimates from primary regulatory, patent, epidemiological and pricing records. This page sets out the method. The full source register and calculation specification are shared on request.
Five stages, and each one shows its working
An opportunity estimate is not a single model. It is a chain in which each stage is owned by one module, carries its own sources and uncertainty, and is read, never recalculated, by the stages downstream.
True entry date
Patent protection decomposed into layers, because compound expiry is rarely the date competitive entry actually becomes possible.
Patient population
Built upward from national registry incidence and prevalence, filtered by subtype, stage and diagnostic confirmation.
Treatment reality
Guideline eligibility narrowed to what is actually prescribed in that market. The gap between the two is usually where the opportunity sits.
Volume and price
Dosing, duration and unit price, with list and net prices distinguished rather than conflated.
Capture
Post-exclusivity erosion and share capture, split by the routes through which a generic entrant actually wins patients.
One value, one owner
Each figure in the chain is calculated in exactly one place and referenced everywhere else. A treatment rate on the market sizing screen can never quietly disagree with the same rate on the epidemiology screen, a failure most internal models carry without knowing it.
Uncertainty compounds, it does not average
Every step carries an error band that is carried forward by section rather than smoothed into one headline range. A chain resting mostly on modelling declares itself instead of borrowing confidence from the registry data around it.
We size every market twice, and show where the two disagree
One estimate reads the commercial data you already license. The other is built bottom-up from public records and needs no subscription. Where both exist, the gap between them is itself a finding, and we analyse it rather than picking the friendlier number.
Licensed basis
Reads the commercial audit data you hold. We never buy, store or resell licensed data. If your licence lapses, this basis goes dark and the other continues.
Public-source basis
Built from registry, literature and government pricing records. Available in every market we cover, including where no commercial audit exists.
Divergence analysis
The gap is quantified and explained. Where it exceeds tolerance, scoring is held at reduced confidence until an analyst reconciles it.
Where the data comes from
Primary records from the issuing authority, the registry and the patent office, rather than secondary aggregators. The register itself, with each source's coverage, refresh cadence, retrieval method and confidence tier, is shared on request.
For every source the register records what it gives, what it cannot tell you, geographic coverage, refresh cadence, retrieval method and confidence tier.
Regulatory
Approval status, exclusivity, therapeutic equivalence and price control, taken from national authorities rather than resellers.
Patent and IP
Full-text patent records and legal status across the major jurisdictions, searched by classification.
Scientific and epidemiological
National cancer and disease registries, peer-reviewed literature and international guideline bodies.
Commercial
Licensed audit and pipeline data, read under the client's own subscription.
Primary research
Commissioned fieldwork where no published source exists: hospital audits, expert interviews, retail pricing surveys, supplier quotations.
Three problems most opportunity models get wrong
These are the places where a plausible-looking estimate quietly breaks. Each is handled explicitly in the platform and documented in full in the dossier.
Compound expiry is the wrong date
A molecule can be free of its compound patent for years while a formulation, device or packaging patent still blocks the commercially relevant presentation. Planning entry on compound expiry alone has cost entrants whole launch windows.
The filer is not published
Regulators disclose that a patent challenge has been filed, and when, but not by whom. Pending applications are confidential. Competitor views built on assumption at this point are the most common source of a wrong entrant list.
Guideline-eligible is not treated
A guideline describes what should happen. In most markets the share of eligible patients actually receiving therapy is far lower, and that gap, not the treated pool, is usually the largest driver of a post-exclusivity opportunity.
Every number arrives with a confidence rating attached
Confidence is set by rule, not by analyst judgement. Where modelled inputs dominate a chain, the rating is capped automatically. An analyst cannot talk a weak estimate up.
Evidence tiers
- Tier 1Primary record, published by the issuing authority
- Tier 2Secondary or partial coverage that needs parsing or adjustment
- Tier 3Proxy or modelled, never used as a sole basis
Cross-method validation
Where a figure can be reached two independent ways, we run both and report the gap rather than choosing.
- Close agreementReported as convergent
- A wider gapReported with the difference stated
- A material gapHeld at reduced confidence pending interpretation
- Beyond thatFlagged for review rather than averaged away
Request the full methodology dossier
Ask our team about the complete source register, calculation specification, evidence-layer framework and worked molecule examples. This form records your request for review; it does not send a dossier automatically.