Investment process
How to review a changing investment thesis
Update an investment thesis through evidence, preserving original assumptions, contrary observations and reasons for change.
01Preserve the original thesis
Specify the customer need, business mechanism, required resources and failure conditions. Separate facts, assumptions and conclusions, with dated information boundaries. Later announcements must not be retroactively treated as known. A vague expression of optimism cannot be meaningfully tested.
02Choose indicators with a mechanism
Use measures close to value creation: paid demand, delivery, retention, cost, capital use and collection. Test major assumptions from more than one angle. Preserve definitions, cohorts and periods. A change in reporting convention may break comparability even when the chart appears continuous.
03Separate noise from structural change
Seasonality and acceptance timing can move a period's results without invalidating the thesis. Look at persistence, breadth and mechanism. A single delayed receipt differs from deteriorating ageing across many customers. Growth investment needs milestones; indefinite delays cannot be justified by invoking patience alone.
04Seek disconfirming evidence
Make contrary evidence a required part of review, but verify it as carefully as favourable material. Rumours are not stronger merely because they disagree. Define in advance what observation would make the original explanation insufficient. This reduces the temptation to move the evidential threshold after the fact.
05Review operations and price separately
Business improvement and investment attractiveness are not identical. Update related assumptions together: capital intensity affects cash and competition affects margins. For IPOs, post-listing share structure and disclosure add new evidence. Neither the original offering narrative nor short-term price performance should settle every business question.
06Look for shared drivers
Different sectors may depend on the same funding, customer spending, energy or technology conditions. Geographic labels do not establish economic independence. Options also require separate review of maturity, volatility and obligations; an unchanged company thesis does not imply unchanged instrument risk.
07Document revisions proportionately
Record what changed, why and what remains uncertain. Keep prior versions. An exploratory partnership may add a watch item; persistent customer or funding deterioration may require rebuilding scenarios. Match the size of the revision to the evidence, and specify which additional disclosure would resolve an open question.
08Review dependent conclusions
If growth comes from one-off projects rather than expanding recurring use, revisit staffing, collection and valuation together. Preserve what remains valid and define evidence for repeatability and reusable products. A changed premise can affect an entire reasoning chain.
09Make review a repeatable practice
Combine scheduled reviews with material-event triggers. Classify whether evidence still supports the thesis, requires partial revision or undermines its mechanism. These are discussion outcomes, not automatic trading rules. Discipline means both persistence and change can be explained and tested against subsequent information.
