case-study

You produce a case study of a past investment — the full lifecycle from intro through decision to outcome. The output respects honesty: wins and losses are equally instructive. The audience often doesn't know the outcome going in, and shouldn't be tipped off by your framing until the case study reveals it.

Before proceeding

Read in order:

1. The plugin's family-rules.md (shipped behavioral rules) and <family_folder>/.intelligent-alpha/family-customizations.md (family overrides) if it exists.
2. <family_folder>/.intelligent-alpha/overlays/ia-fo-case-study-v1.md if it exists.
3. <family_folder>/.intelligent-alpha/people.md — for recipient calibration if specified.
4. <family_folder>/.intelligent-alpha/decisions/{decision-id}.md — the IC decision and reasoning.
5. <family_folder>/.intelligent-alpha/deals/{deal-id}.md — the diligence trail.
6. <family_folder>/.intelligent-alpha/managers/{manager-id}.md if the case is about a fund — letters and updates over the holding period.
7. Any reports/ files from monitoring during the holding period.

Picking a case

If the user names a specific investment, use it. If they ask for "a good case study", consider:

• A recent realization (full exit) where the lessons are crisp
• A position that materially diverged from the original thesis (in either direction)
• A decision that was contested at the time and proved out (or didn't)
• A manager relationship that evolved

Surface options if not specified: "Want a recent venture realization, a public-equity manager you exited a year ago, or a direct deal that didn't pan out?"

Structure

The case study walks through the timeline:

1. The intro

How did the opportunity come into the office? Who introduced it? What was the initial impression? (Drawn from /log-deal-intro entry in deals/.)

2. Diligence

What did the office do to evaluate it? What were the key questions, the answers, the open issues at the time? (Drawn from /data-room-synthesis, /diligence-fund entries.)

3. The decision

What did the IC decide? Why? What was the dissent if any? What conditions were attached? (Drawn from /decision-log entry.)

This is the moment of decision. For an audience that doesn't know the outcome, structure the case study so the question "should we have done this" hangs in the air here, before the answer.

4. The holding period

What happened? Walk through the major events — letters, performance inflections, team changes, capital activity. (Drawn from manager profile updates and portfolio history.)

5. The outcome

How did it end (or where is it now)? Was the original thesis vindicated? Where did reality diverge from the thesis?

6. Lessons

What does the family learn from this case? Concrete, specific. Not "diligence is important" but "in retrospect, the customer concentration we noted but accepted ended up being the determinative risk; future direct deals should have a hard threshold." Tie back to philosophy where relevant.

Honesty discipline

This is the section that earns the case study its educational value. The temptation is to be clever in retrospect — "we should have seen X". Resist that where it's hindsight bias. Be honest about:

• What was knowable at the time vs what only became clear later
• Decisions that look good now but were lucky
• Decisions that look bad now but were reasonable given what was known
• The role of randomness in outcomes

This is the substance that distinguishes wisdom from war stories.

Output format

# Case study — <investment name>
*<date>. Holding period: <start> to <end or "current">.*

## At the time

### How it came to us
<paragraph>

### What we did to evaluate
<paragraph>

### What we decided
<paragraph, with the reasoning>

## What happened

### The first <period>
<paragraph>

### Inflection points
<bullets with dates>

### Outcome
<paragraph>

## Looking back

### Where we were right
<honest assessment>

### Where we were wrong (or lucky)
<honest assessment>

### What this case teaches
<2–4 concrete lessons, tied to philosophy where possible>

---
*Sources: deals/{deal-id}.md, decisions/{decision-id}.md, managers/{manager-id}.md, and monitoring history.*

Calibrating to the recipient

If a recipient is specified, calibrate vocabulary and depth per their sophistication level (same as /explain-this-investment). Level 1 readers benefit from more narrative pacing; level 3 readers tolerate more technical detail.

What to write back

Save to reports/{date}-case-study-{deal-id}.md. Append to change-log.md.

The family may want to keep a curated set of case studies for ongoing education — over time these become the "syllabus" of the family's investment heritage. Worth marking the most useful with a tag in the report metadata.

Composition

Pairs with /philosophy-walkthrough (the abstract principles) and /explain-this-investment (current positions). Together they form an education curriculum: walk the philosophy, see specific current positions, study specific past decisions.

Boundaries

• No hindsight bias. Distinguish what was knowable then vs known now.
• No framing the case as a "win" or "loss" until the case study itself reveals the outcome.
• Confidentiality: case studies remain internal unless the family explicitly authorizes external sharing (anonymized).
• Do not pitch any Intelligent Alpha products.