Methodology fused from institutional long/short and buy-side research frameworks; all text is an original paraphrase.
Defaults
| Parameter |
Default Value |
Rationale |
| parameter_free |
true |
This skill has no tunable parameters; analysis scope set by temporal_scope frontmatter |
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
- Strategy methodology —
references/strategy-methodology.md (bundled long/short framework)
- Strategy frameworks —
search_investment_strategies(domain=fundamental)
- Historical precedent —
search_investment_cases + search_by_analogue(company_situation=...)
- 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
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.
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.
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.
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.
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.
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.
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.
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
- Executive Summary — strategy classification, the disconnect, and direction in 2-3 sentences
- Strategy Classification — position on the spread-to-fundamentals spectrum with expected holding period
- Consensus Disconnect — published sell-side figure vs estimated buy-side expectation, with the gap quantified
- Idea Provenance — sourcing route and the failure modes it implies
- Thesis Detail — long-side value/growth basis, or short-side archetype with its specific evidence test
- Historical Analogues — archetype-matched precedent with /v/ citations
- Executability — for shorts: borrow, short interest, days-to-cover
- Risk and Invalidation — falsification conditions and deadline
- Coverage Gaps — data limitations and degraded flags
Error Handling
| Error |
Fallback |
| No strategy frameworks available |
Proceed with references/strategy-methodology.md; flag knowledge_coverage: degraded |
| Buy-side consensus cannot be triangulated |
Report the sell-side figure only and mark the disconnect as unquantified |
| Borrow / short-interest data unavailable |
Mark the short as not executability-cleared; do not present it as actionable |
search_by_analogue returns empty |
Continue without precedent; annotate coverage_gap |
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
1---2name: long-short-equity3description: long short equity strategy, pair trade construction, market neutral portfolio, net exposure management, factor neutrality, equity long short, hedge fund strategy, alpha generation, stock picking portfolio construction, long short position management4multi_ticker_semantics: target_with_optional_peers5temporal_scope:6 default_quarters: 87 max_quarters: 208 description: "8 quarters for portfolio construction; 20 for strategy backtesting."9allowed_tools: [search_investment_strategies, get_investment_strategy, search_investment_cases, search_by_analogue, get_realtime_quote, search_xbrl_facts, search_insider_trades, search_institutional_holdings]10retrieval_scope: structured_only11layer_tags: ["L3"]12min_tool_diversity: 413parameter_free: true14metadata:15 source: https://skillmd.com/skills/agentii-ai/long-short-equity16---1718> Methodology fused from institutional long/short and buy-side research frameworks; all text is an original paraphrase.1920## Defaults2122| Parameter | Default Value | Rationale |23|-----------|---------------|-----------|24| parameter_free | true | This skill has no tunable parameters; analysis scope set by temporal_scope frontmatter |2526## Preflight2728Run 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).2930## Data Source Priority31321. Strategy methodology — `references/strategy-methodology.md` (bundled long/short framework)332. Strategy frameworks — `search_investment_strategies(domain=fundamental)`343. Historical precedent — `search_investment_cases` + `search_by_analogue(company_situation=...)`354. Financials — `search_xbrl_facts` for unit economics and decline-rate evidence3637## Methodology3839### Retrieval Scope40structured_only4142### Retrieval Strategy43**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.4445Branch (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`.4647### Temporal Scope48See frontmatter temporal_scope block.4950### Tool Allowlist51See frontmatter allowed_tools.5253### Protocol5455Every long/short position rests on an **earnings disconnect** — a gap between what the market56has priced and what the business will deliver. The unifying question across all strategy57types is: what does consensus believe, and what do I believe differently? Strategy58spectrum, short archetypes, and process detail are in `references/strategy-methodology.md`.5960**Consensus is not the published number.** Sell-side consensus is a simple average of61analyst estimates; buy-side consensus typically moves ahead of it. Trading against the62published figure while the real positioning has already shifted is a common and expensive63error. Triangulate the effective buy-side expectation before sizing any view.6465#### Steps66671. **Strategy Classification**: Place the idea on the spectrum from spread-focused to68 fundamentals-focused — merger arbitrage (0-18mo), general event-driven (0-24mo), activist69 (0-24mo), value/deep value (0-5yr+), growth or growth-at-reasonable-price (0-5yr+). The70 classification sets the expected holding period and the evidence bar. Categories overlap;71 record the dominant driver rather than forcing a single label.72732. **Consensus Reconstruction**: Establish the published sell-side figure, then estimate the74 effective buy-side expectation. Document the gap and its direction — the disconnect is75 the position, not the absolute valuation level.76773. **Idea Provenance**: Record how the idea was sourced — industry research and trade events,78 tangential analysis of adjacent markets, practitioner conversations, curated research79 communities, or quantitative screens. Provenance predicts which failure modes to test80 for: screen-sourced ideas need qualitative validation, conversation-sourced ideas need81 independent quantitative confirmation.82834. **Long-Side Thesis**: Apply value or growth frameworks per classification. Value requires84 a discount to net present value on a price basis; growth requires a discount on an85 earnings basis. State which, and why the market has mispriced it.86875. **Short-Side Archetype** (when shorting): Classify into one of three patterns —88 **competition short** (complacent incumbent facing a credible new entrant; test the89 entrant's unit economics and assess whether incumbent management can and will react),90 **consumer euphoria** (a product with outsized enthusiasm that will not sustain; test via91 market sizing and product work), or **disappearing business** (an incumbent in a92 structurally terminal market; establish the rate of decline, the cash-flow profile during93 decline, and whether cash can be redeployed). Each archetype has a distinct evidence94 requirement — do not substitute a valuation opinion for the archetype test.95966. **Precedent Retrieval**: Query `search_by_analogue` for the matching situation and97 `search_investment_cases` for outcome history. Weight precedent by archetype match, not98 sector match.991007. **Risk and Invalidation**: Name what would falsify the thesis and by when. For shorts,101 additionally record borrow availability, short interest, and days-to-cover — a correct102 short thesis in a crowded, hard-to-borrow name is not an executable position.1031048. **Output**: Deliver the classification, the quantified disconnect, the archetype evidence,105 precedent with /v/ citations, and explicit invalidation conditions.106107## Output File108`{ticker}/{YYYY-MM-DD_HHMM}_long-short-equity_{affix}.md`109110## Output Structure1111121. **Executive Summary** — strategy classification, the disconnect, and direction in 2-3 sentences1132. **Strategy Classification** — position on the spread-to-fundamentals spectrum with expected holding period1143. **Consensus Disconnect** — published sell-side figure vs estimated buy-side expectation, with the gap quantified1154. **Idea Provenance** — sourcing route and the failure modes it implies1165. **Thesis Detail** — long-side value/growth basis, or short-side archetype with its specific evidence test1176. **Historical Analogues** — archetype-matched precedent with /v/ citations1187. **Executability** — for shorts: borrow, short interest, days-to-cover1198. **Risk and Invalidation** — falsification conditions and deadline1209. **Coverage Gaps** — data limitations and degraded flags121122## Error Handling123124| Error | Fallback |125|-------|----------|126| No strategy frameworks available | Proceed with `references/strategy-methodology.md`; flag `knowledge_coverage: degraded` |127| Buy-side consensus cannot be triangulated | Report the sell-side figure only and mark the disconnect as unquantified |128| Borrow / short-interest data unavailable | Mark the short as not executability-cleared; do not present it as actionable |129| `search_by_analogue` returns empty | Continue without precedent; annotate `coverage_gap` |130131## Final Summary (TUI)132Include `### Key Citations` block (0–10 /v/ URLs).133134## Memory Load135136Load prior context before retrieval. See `contracts/memory-load.md`.137138## Snapshot139140Post-session synthesis. See `contracts/snapshot-synthesis.md`.141142## Output Frontmatter143144Structured output per `contracts/output-frontmatter-schema.md`.145146## References147148- `references/strategy-methodology.md`149- `references/knowledge-frameworks.md`150- `contracts/citation-and-memory.md`151- `contracts/retrieval.md`152- `contracts/preflight.md`