Long Short Equity

Methodology fused from institutional long/short and buy-side research frameworks; all text is an original paraphrase.

Defaults

Preflight

Run canonical pre-flight per contracts/preflight.md. Include the X-Agentii-Trace header on every tool call per contracts/x-agentii-trace-header.md — carry the _run_id from your first tool result and name yourself (and your parent, if you were spawned).

Data Source Priority

1. Strategy methodology — references/strategy-methodology.md (bundled long/short framework)
2. Strategy frameworks — search_investment_strategies(domain=fundamental)
3. Historical precedent — search_investment_cases + search_by_analogue(company_situation=...)
4. Financials — search_xbrl_facts for unit economics and decline-rate evidence

Methodology

Retrieval Scope

structured_only

Retrieval Strategy

Ownership & insider signals: search_institutional_holdings (top-10 holders + whale portfolios, direction=accumulating|reducing|new|exited) and search_insider_trades (Form-4 transactions with SEC URLs) are available as signal inputs.

Branch (a) Structured Data Query from contracts/retrieval.md. Retrieve frameworks via search_investment_strategies, precedent via search_investment_cases and search_by_analogue. Detailed methodology in references/strategy-methodology.md.

Temporal Scope

See frontmatter temporal_scope block.

Tool Allowlist

See frontmatter allowed_tools.

Protocol

Every long/short position rests on an earnings disconnect — a gap between what the market has priced and what the business will deliver. The unifying question across all strategy types is: what does consensus believe, and what do I believe differently? Strategy spectrum, short archetypes, and process detail are in references/strategy-methodology.md.

Consensus is not the published number. Sell-side consensus is a simple average of analyst estimates; buy-side consensus typically moves ahead of it. Trading against the published figure while the real positioning has already shifted is a common and expensive error. Triangulate the effective buy-side expectation before sizing any view.

Steps

1. Strategy Classification: Place the idea on the spectrum from spread-focused to fundamentals-focused — merger arbitrage (0-18mo), general event-driven (0-24mo), activist (0-24mo), value/deep value (0-5yr+), growth or growth-at-reasonable-price (0-5yr+). The classification sets the expected holding period and the evidence bar. Categories overlap; record the dominant driver rather than forcing a single label.

2. Consensus Reconstruction: Establish the published sell-side figure, then estimate the effective buy-side expectation. Document the gap and its direction — the disconnect is the position, not the absolute valuation level.

3. Idea Provenance: Record how the idea was sourced — industry research and trade events, tangential analysis of adjacent markets, practitioner conversations, curated research communities, or quantitative screens. Provenance predicts which failure modes to test for: screen-sourced ideas need qualitative validation, conversation-sourced ideas need independent quantitative confirmation.

4. Long-Side Thesis: Apply value or growth frameworks per classification. Value requires a discount to net present value on a price basis; growth requires a discount on an earnings basis. State which, and why the market has mispriced it.

5. Short-Side Archetype (when shorting): Classify into one of three patterns — competition short (complacent incumbent facing a credible new entrant; test the entrant's unit economics and assess whether incumbent management can and will react), consumer euphoria (a product with outsized enthusiasm that will not sustain; test via market sizing and product work), or disappearing business (an incumbent in a structurally terminal market; establish the rate of decline, the cash-flow profile during decline, and whether cash can be redeployed). Each archetype has a distinct evidence requirement — do not substitute a valuation opinion for the archetype test.

6. Precedent Retrieval: Query search_by_analogue for the matching situation and search_investment_cases for outcome history. Weight precedent by archetype match, not sector match.

7. Risk and Invalidation: Name what would falsify the thesis and by when. For shorts, additionally record borrow availability, short interest, and days-to-cover — a correct short thesis in a crowded, hard-to-borrow name is not an executable position.

8. Output: Deliver the classification, the quantified disconnect, the archetype evidence, precedent with /v/ citations, and explicit invalidation conditions.

Output File

{ticker}/{YYYY-MM-DD_HHMM}_long-short-equity_{affix}.md

Output Structure

1. Executive Summary — strategy classification, the disconnect, and direction in 2-3 sentences
2. Strategy Classification — position on the spread-to-fundamentals spectrum with expected holding period
3. Consensus Disconnect — published sell-side figure vs estimated buy-side expectation, with the gap quantified
4. Idea Provenance — sourcing route and the failure modes it implies
5. Thesis Detail — long-side value/growth basis, or short-side archetype with its specific evidence test
6. Historical Analogues — archetype-matched precedent with /v/ citations
7. Executability — for shorts: borrow, short interest, days-to-cover
8. Risk and Invalidation — falsification conditions and deadline
9. Coverage Gaps — data limitations and degraded flags

Error Handling

Final Summary (TUI)

Include ### Key Citations block (0–10 /v/ URLs).

Memory Load

Load prior context before retrieval. See contracts/memory-load.md.

Snapshot

Post-session synthesis. See contracts/snapshot-synthesis.md.

Output Frontmatter

Structured output per contracts/output-frontmatter-schema.md.

References

• references/strategy-methodology.md
• references/knowledge-frameworks.md
• contracts/citation-and-memory.md
• contracts/retrieval.md
• contracts/preflight.md