Event-driven & merger arbitrage
From deal announcement to a cited investment memo in minutes.
SAMWISE runs the full event-driven research pipeline: screen, thesis, probability tree, regulatory path, memo. Every figure traces to a primary filing; anything unsourced is marked UNVERIFIED, not guessed.
Built for UK and European takeover mechanics. US deals covered where relevant.
| Situation | Type | Spread | Ann. | Status |
|---|---|---|---|---|
| THG LN | Scheme | 3.4% | 11.2% | LIVE |
| ALD FP | Cash offer | 2.1% | 8.9% | LIVE |
| SOBI SS | Rule 2.4 | — | — | PUSU 21d |
| DWNI GY | Tender | 1.6% | 6.1% | WATCH |
fail 8% · early 22% · base 58% · late 12%
p-weighted IRR 10.4% vs hurdle 7.0% · sources: 2.7 announcement, scheme doc
Illustrative output. Names and figures shown are examples, not live research.
Event-driven research is a race you run by hand.
An RNS drops at 07:01. By the time you have read the announcement, rebuilt the timetable, checked irrevocables and sized the spread, the easy basis points are gone.
Sources everywhere
Terms in the 2.7, register in the TR-1s, flow in the Form 8.3s, timetable in the scheme doc. Four sources before a single number is safe to use.
Aggregators guess
News wires round, paraphrase and lag. A wrong deal term that looks confident is worse than an honest gap, and generic AI tools make it worse, not better.
Memos take days
The story, the probability tree, the regulatory path and the workbook inputs are all separate documents. Assembling them by hand is the bottleneck.
One pipeline: catalyst to conviction.
SAMWISE runs the same five stages a disciplined analyst runs. It just does the fetching, cross-checking and drafting in minutes instead of days.
Triage the batch, keep only the tradeable.
Every situation tiered against your hurdle: premium, gross spread, implied annualised return, liquidity, recommendation status, irrevocables.
- Tier 1 / Tier 2 / Watch / Skip in one compact table
- One-line rationale per name
- Default 7% annualised hurdle, configurable
| Name | Tier | Ann. | Why |
|---|---|---|---|
| Target A | 1 | 11.2% | Clean cash, rec'd, irrevocables 28% |
| Target B | 2 | 9.8% | Contested, interloper possible |
| Target C | W | — | Rule 2.4, PUSU 21 days |
| Target D | S | 2.3% | 96% committed, thin spread |
Understand why this deal exists.
Board intentions, share register, management incentives, shareholder archetypes, deal history. Has this target been approached before, at what price, and what happened?
- Five-section thesis spine, in your house style
- Prior-approach history surfaced automatically
- Exported as a formatted .docx
2. Board intentions & share register
3. Management alignment & incentives
4. Shareholder archetypes
5. Rationale, why now, counterfactual
+ deal history: approached 2023 at 285p, rejected; lessons applied
Price the path, not the headline.
Explicit fail / early / base / late scenario trees with probability-weighted IRR against your hurdle. Regulatory path mapped by jurisdiction: antitrust, FDI, panel mechanics, long-stop.
- Bump-or-walk base rates built from named precedents
- UK Code mechanics handled precisely: 2.4, 2.7, PUSU, schemes, squeeze-out
- Probabilities always sum to 100%: enforced, not hoped
| Scenario | P | Price | Ann. |
|---|---|---|---|
| Fail | 8% | 412p | -18.4% |
| Early | 22% | 505p | 14.6% |
| Base | 58% | 505p | 10.9% |
| Late | 12% | 505p | 7.2% |
From analysis to a two-page memo and workbook inputs.
A PM-ready investment memorandum (terms, thesis, tree, regulatory path, risks, recommendation, kill-line) plus deal-monitor workbook inputs, cell by cell.
- Two-page A4 memo in house style
- Writes only to mapped input cells, never formulas
- Memo IRR reconciled to workbook within 50bps
Regulatory path · Key risks · Recommendation
Size · Entry · Kill-line
workbook: 14 input cells recommended, 0 formula cells touched
Built for the people who trade the event.
Not a general market data feed. A working analyst's pipeline, productised.
Event-driven & special-situations analysts
Cover more names without lowering the evidence bar. SAMWISE drafts the screen, story and tree; you keep judgement and the final word.
Request early access →Portfolio managers
Every memo arrives with sources attached and a kill-line stated. Act on research you can audit line by line, at the speed the spread demands.
Request early access →Family offices & prop desks
Institutional-grade merger-arb process without an institutional headcount. One seat does the work of a research associate.
Request early access →AI you can audit, because every number has a source.
Most AI research tools optimise for an answer. SAMWISE optimises for a defensible one.
Deal terms, prices, dates, conditions and timetables come from regulatory filings and offer documents. Never from an aggregator, never from model memory. If a primary source cannot be fetched, the figure is marked UNVERIFIED and the pipeline stops on it rather than guessing.
Form 8.3 disclosures, TR-1s and scheme appendices are parsed into the register view. The full source log ships with every deliverable, so any number can be traced back to the document it came from.
irrevocables 27.9% [2.7 appendix III]
long-stop 2026-11-30 [scheme doc p.44]
break fee UNVERIFIED (no primary source fetched; not stated)
Read a real deal report before you trust us with yours.
We publish a complete, unedited SAMWISE research pack for a completed situation: the screen entry, the story, the full probability tree, the regulatory read and the final memo, sources and all. Judge the output, not the marketing.
Get the sample packDelivered with early-access confirmation. Completed situations only; nothing live.
INVESTMENT MEMORANDUM: [COMPLETED SITUATION]
Situation. Recommended cash offer at 505p under a scheme of arrangement; premium 38.1% to undisturbed. [2.7, 2026-05-14]
Tree. Fail 8 / Early 22 / Base 58 / Late 12 → p-weighted IRR 10.4% vs 7.0% hurdle.
Path. CMA Phase 1 expected clear; no FDI trigger; court sanction est. 2026-10.
Kill-line. Exit if spread > 6.5% post-vote or interloper 2.4 lapses.
8 sources logged · 0 UNVERIFIED · generated in 6m 42s
Where SAMWISE fits next to what you already pay for.
It is a workflow for one strategy, not another data terminal. Most users keep their market data and put SAMWISE in front of it.
Search finds the document. It does not build the memo.
AlphaSense is horizontal search across a very wide corpus, and it is good at it. You still read the 2.7, rebuild the timetable, check the irrevocables and write the memo yourself. SAMWISE starts where the document ends: it extracts the terms, prices the spread, builds the probability tree and produces the memo, with each figure carrying the filing it came from.
Market data, not research production.
A terminal gives you the price, the register and the news feed, and nothing here replaces that. What it does not give you is a source-anchored regulatory path, a scenario tree with the derivation shown, or a memo a PM can act on. SAMWISE writes those, and the workbook it populates is Bloomberg-wired, so the two run together rather than against each other.
The same pipeline, run the same way every time.
Doing it by hand is the honest benchmark, and it is what most desks do. The cost is not only hours: it is that the fifteenth deal of the month gets less rigour than the first. SAMWISE runs the same five stages in the same order on every situation, and marks what it could not verify instead of quietly rounding.
Questions a careful buyer asks.
Where does the data come from?
How is this different from AlphaSense or a Bloomberg terminal?
What happens when a number cannot be verified?
Is this investment advice?
What does it cost?
What does early access involve?
The next catalyst won't wait for your memo.
Request early access and get the sample deal report by reply.