documenting-a-thesis-and-monitoring-triggers — evaluating investments

The practical point: Write a 1-page, date-stamped thesis at least 24 hours before you buy, attach ≥5 triggers (3 fundamental + 2 price-based), and commit to an execution window of ≤10 trading days when an exit trigger trips—because the behavioral penalty shows up in numbers like 4.4% per year, 7.1 percentage points per year, and 1.8× holding-time distortion when you don't pre-commit. (Odean, 1998; Barber & Odean, 2000; Kahneman & Tversky, 1979)

Why Thesis Documentation Matters

The point is 1 thing: your thesis document is a decision contract that converts a continuous market stream into discrete actions at pre-specified thresholds. Without that contract, investors become 1.5× more likely to sell winners than losers, and that pattern is associated with about 4.4% annual return erosion in the disposition-effect channel. (Odean, 1998)

You also pay for "monitoring without rules" as excess activity: the most frequent traders earned 11.4% annually versus 18.5% for buy-and-hold investors, a 7.1 percentage point gap attributed to undisciplined adjustments and timing mistakes. (Barber & Odean, 2000)

And you're fighting an asymmetry: losses are weighted about 2.25× more than gains, which pushes "wait one more week" into "hold 1.8× longer than optimal" unless you have written, binary thresholds. (Kahneman & Tversky, 1979)

Thesis Documentation: Build a 1-Page Spec (7 fields)

1) Thesis statement (1 paragraph, 3 assumptions)

Write 1 paragraph that includes 3 quantified assumptions (not vibes), each with a measurable variable and a numeric floor/ceiling.

Example pattern (3 numbers, 3 variables):

Revenue growth: 15%–25% annually through a dated horizon like 2026
Market share: ≥18%
Gross margin: ≥42%

2) Time horizon (1 range)

State a single range like 18–36 months, because "long term" without a number becomes "indefinitely" in ≥3 drawdowns.

3) Position size (2 caps)

Document (a) initial allocation (e.g., 4%) and (b) hard concentration ceiling (e.g., ≤10%), because concentration above 15% correlates with 3.1× higher portfolio volatility without commensurate return improvement for individuals.

4) What would prove you wrong (3 fundamental disconfirmers)

Write ≥3 fundamental triggers with numeric thresholds; fewer than 5 total triggers per position is associated with 2.3× higher incidence of disposition-effect behavior.

5) What would force risk reduction (2 price-based triggers)

Price triggers are risk management, not "truth," so you define 2 actions tied to drawdown/volatility, not round numbers.

6) Review cadence (3 clocks)

Lock 3 clocks into the thesis: within 48 hours of earnings, monthly quick checks, and quarterly deep reviews; delays beyond 72 hours correlate with about 18% lower trigger adherence.

7) Rewrite policy (1 expiry)

Set a hard rewrite interval of 18 months (even if performance is +80%), because theses older than 24 months show about 45% assumption staleness.

Trigger Identification: Turn Assumptions into Thresholds

Minimum viable trigger set: 3 fundamental + 2 price

Your baseline is 5 triggers per position: 3 that test the business thesis and 2 that control path risk.

A clean mapping rule (1/3 haircut): if your thesis assumes 15% revenue growth, set the "thesis at risk" trigger at 10% (a 33% shortfall), not an arbitrary "−20% stop-loss."

Fundamental triggers (examples with exact thresholds)

Use thresholds that match the thesis variables, with explicit measurement windows:

Growth trigger: revenue growth <10% for 2 consecutive quarters → mandatory thesis review
Profitability trigger: gross margin <40% → automatic 50% position reduction
Competitive trigger: market share loss >2 percentage points → full exit review

Tie each trigger to a source and deadline, e.g., "10-Q metrics within 5 business days of release."

Price-based triggers (examples with 48-hour and 35% rules)

Price is noisy, so you pre-commit to fast review and slow action:

+20% gain: rebalance back to 4% target weight (harvest concentration drift)
−25% decline: thesis review within 48 hours (not 48 days)
−35% decline: mandatory 50% reduction regardless of narrative

The point is 1 separation: fundamentals validate the thesis; price constrains tail risk in ≤10 days. (Barber & Odean, 2000)

Monitoring System: A Calendar That Enforces Action

The 3-layer calendar (48 hours / 15 minutes / 60 minutes)

You schedule monitoring as time budgets with lead-time:

Earnings: review within 48 hours, with alerts set 24 hours in advance
Monthly: 15-minute check against price triggers
Quarterly: 60-minute written reassessment against fundamental triggers

Data sources (3 sources, 1 deadline)

You define 3 primary sources (10-Q/10-K, earnings call transcript, competitor/sector metric) and a capture deadline like ≤5 business days, because unbounded "later" becomes ≥15 days, and delays beyond 15 days are associated with 67% lower adherence to documented exits.

Exit Criteria: Convert Triggers into a Decision Tree

Rule 1: "2-of-3 assumption breach" = exit within 10 trading days

If any 2 of 3 core assumptions break (e.g., growth, share, margin), you exit 100% within 10 trading days—because execution drift beyond 15 days collapses adherence by 67%.

Rule 2: Management change triggers a 90-day clock

If CEO/CFO changes, you impose a 90-day review window and freeze adds (0 new buys) until the written reassessment is complete.

Rule 3: Regime/sector signal triggers hedging (15% over 6 months)

If a sector index underperforms the benchmark by >15% over 6 months, you implement a hedge (size it explicitly, e.g., 25%–50% notional) or reduce gross exposure.

Worked Example: You Document and Monitor a Cyclical Semiconductor Thesis

You manage a $500,000 portfolio and take a 4% initial allocation, so you size the position at $20,000 with a hard ceiling of ≤10% (≤$50,000) even if the stock doubles.

Step 1 (24 hours before entry): you write 1 paragraph with 3 thresholds

You write (with numbers, not adjectives): revenue growth 15%–25% annually through 2026, industry spending >12% annually, market share ≥18%, and gross margin ≥42%.

You date-stamp the thesis ≥24 hours before you trade, because same-day documentation correlates with a 34% higher reversal rate within 90 days.

Step 2: you define 3 fundamental triggers (10%, 40%, 2pp)

You pre-commit:

If revenue growth prints <10% for 2 quarters, you initiate a written review in ≤48 hours.
If gross margin drops <40%, you cut the position by 50% (from $20,000 to $10,000 at that time's price).
If market share falls by >2 percentage points, you run a full exit review with a 10-trading-day execution plan.

Step 3: you define 3 price actions (+20%, −25%, −35%)

You set mechanical actions:

At +20%, the position becomes $24,000; you rebalance to 4% weight (back toward $20,000, depending on portfolio drift).
At −25%, the position becomes $15,000; you run a thesis check inside 48 hours.
At −35%, the position becomes $13,000; you sell 50%, reducing exposure to about $6,500.

Step 4: you quantify outcomes (baseline / good / poor)

You then write the 3-case table so "surprise" is bounded by 3 pre-modeled states:

Baseline (24 months): revenue grows 18% annually, margins stay 44%, stock appreciates 45%; your $20,000 becomes $29,000, and rebalancing at +20% captures $4,000 while leaving upside exposure.
Good case (18 months): revenue grows 30% annually, stock appreciates 80%; your $20,000 becomes $36,000, and staged rebalancing at +20% / +40% / +60% realizes $9,600 while maintaining an $18,000 residual position.
Poor case: revenue growth slows to 5%, gross margin compresses to 38%, stock declines 40%; your −35% rule cuts exposure by 50% at about $13,000, limiting maximum loss to $8,500 versus $12,000 if you rode to −60%, for an estimated $3,500 behavioral benefit.

Historical Stress Tests: What Happens When You Ignore Triggers

Valeant (March 2015–March 2017): a trigger breach followed by an 89.8% slide

The documented violation window was October 2015 (revenue recognition/regulatory standing), with breach occurring around $108/share, but exit occurred at $11/share after an additional 89.8% decline post-breach. The total loss was about $4 billion, roughly 90% of peak value, and a "30-day after breach" exit would have limited loss to about $1.2 billion (a difference of $2.8 billion).

LTCM (January 1998–September 1998): leverage drift from 25:1 to 55:1

The documented leverage ratio was 25:1, but leverage expanded to 55:1 (a 120% overshoot) before a regime break in August 1998; the fund lost $4.6 billion (about 92% of capital) in under 4 months. A hard leverage trigger at 35:1 with mandatory reduction was estimated to preserve about $2.8 billion of capital. (Lowenstein, 2000)

Common Implementation Mistakes (and the quantified damage)

You write the thesis after you buy (0-hour lead time). You then rationalize instead of reassess, and evidence links post-purchase rationalization to 23% longer holding periods for losers and 2.1% additional annual return erosion; you fix it by enforcing a 24-hour waiting period between thesis completion and execution.
You use round-number triggers (like 20%) instead of assumption-derived thresholds. You then sell more often at local bottoms, with evidence associating arbitrary price triggers with a 31% higher rate of bottom-selling versus fundamental thresholds; you fix it by mapping "15% expected growth" to a "10% trigger" (a 33% miss), and by using volatility logic like 1.5× ATR instead of "nice" integers.
You omit sizing and rebalance rules (0 caps, 0 drift controls). You then allow concentration drift of 200%–400% through appreciation and suffer an estimated 1.7% annual shortfall from over-concentration; you fix it by setting "rebalance at 150% of target weight" and "never exceed 8%–10% of portfolio."

Implementation Checklist (Tiered by ROI)

Tier 1 (highest ROI, ~60 minutes setup, ~15 minutes/month)

Write a 1-page thesis with 3 quantified assumptions and a 24-hour pre-trade timestamp.
Define ≥5 triggers (3 fundamental + 2 price) and an exit execution window of ≤10 trading days.
Add the 48-hour post-earnings review rule and a quarterly 60-minute rewrite checkpoint.

Tier 2 (medium ROI, ~90 minutes setup, ~30 minutes/month)

Add a position cap of ≤10% and a rebalance rule at +20% and/or 150% of target weight.
Build a 3-case outcome table (baseline/good/poor) with at least 1 dollar-loss limit (e.g., −$8,500) and 1 profit-harvest rule (e.g., +$4,000 at +20%).

Tier 3 (situational ROI, ~120 minutes setup, quarterly)

Add 1 regime trigger (e.g., sector underperformance >15% over 6 months) and a predefined hedge size (25%–50% notional).
Add an 18-month thesis rewrite timer and a "stale thesis" audit that flags assumptions older than 24 months.

The Durable Lesson

You don't rise to your conviction level 1 time—you fall to your documentation level 100 times, and the numbers that follow from weak pre-commitment show up as 4.4% per year behavior drag, 7.1 percentage points per year activity penalties, and post-breach drawdowns like 89.8% when you let action windows slip past 10 trading days. (Odean, 1998; Barber & Odean, 2000)

Documenting a Thesis and Monitoring Triggers