Your research should survive the next quarter
The investment memo captures a view at one moment. The assumptions, sources, and unresolved questions need a life beyond that moment.
Not investment advice.
You return to a company you researched six months ago. The memo is still there. The spreadsheet opens. But you cannot remember why a particular customer mattered, which assumption you were least comfortable with, or what evidence would have made you walk away.
The conclusion survived. Much of the reasoning did not. The next review becomes an exercise in rebuilding context rather than judging what changed. In a team, the problem becomes more visible when an analyst changes coverage or leaves. For an individual investor, it happens whenever attention moves to another part of the portfolio.
The answer is not simply to write longer memos. It is to preserve the parts of the research that future decisions will need, and keep those parts connected to new evidence.
A price target is not institutional memory
A target price can tell you where an analyst landed. It rarely tells you which assumption did the most work. The conclusion may depend on stronger demand, lower funding costs, a product launch, or a customer's willingness to keep spending. If those dependencies are hidden inside a model or a long narrative, they are difficult to monitor.
Imagine a hypothetical investment in a software company whose growth depends on expanding within existing customers. The final memo says the company has attractive recurring revenue. The more useful memory is specific: expansion must offset slower new-customer growth, the largest customers must keep renewing, and sales costs must not rise faster than revenue.
When the next results arrive, those conditions let you ask a focused question. Is the original growth mechanism intact? Without them, the discussion can drift toward whichever metric management emphasized most recently.
Keep a thesis record you can actually revisit
A thesis record should be short enough to use and explicit enough to challenge. Start with the reason you expect the investment to work. Separate that reason from the valuation you are willing to pay and from the conditions that would make the position unsuitable for the portfolio.
For each important assumption, record the supporting evidence and its date, what remains uncertain, and the next check. The purpose is not to complete a form. It is to leave a clear account of the decision for the person who will review it later, including your future self.
A practical record contains:
- The expected business change and why it matters to the investment.
- The assumptions that must hold, with sources and dates.
- The evidence that would weaken or invalidate the view.
- The unresolved questions and the next planned review.
- The role of the position in the portfolio and the risks already accepted.
Record why the view changed
Research continuity requires more than keeping the latest version. If the original case expected customer expansion to accelerate and the current case depends on cost reductions instead, the investment has changed. Replacing the old paragraph with a new one can hide that change.
Keep a dated record of material revisions. Explain the evidence that prompted each revision, which assumption changed, and whether the change affected conviction, valuation, or position suitability. A view can evolve for good reasons. The record should make those reasons visible.
This is especially useful when a position performs well for reasons different from the original thesis. A favorable outcome does not establish that the reasoning was correct. Preserving the original case creates room to learn from both the decision and the result.
Review the change, not the entire archive
A quarterly review should begin with the assumptions from the last review and the evidence that has arrived since. Some assumptions will be supported, some will be challenged, and others will remain unresolved. That comparison is the core of the work.
For the hypothetical software company, a decline in reported expansion could challenge the growth case. Before changing the view, establish whether the metric definition changed, whether currency affected the comparison, and whether the weakness is concentrated in a particular customer group. The old research gives the new investigation a starting point, not a predetermined answer.
Organize the review around three questions: what changed in the business, what changed in our understanding, and what changes in our decision? Keeping those questions separate makes it harder for a strong narrative to substitute for a material update.
Make handoffs an investment exercise
A handoff should transfer the open questions as well as the completed work. The incoming researcher needs to know what is still unverified, which relationships may be stale, and where the outgoing researcher made a judgment call. An archive of polished conclusions will not answer those questions by itself.
A useful test is to ask someone unfamiliar with the name to explain the thesis, identify its most fragile assumption, and locate the source behind that assumption. If they cannot do that without a lengthy conversation, the research still depends heavily on the original author.
The same test applies to a personal portfolio. Put a name aside and return later. If your notes explain only what you liked, rather than what had to remain true, they are an incomplete memory of the decision.
Compounding starts with continuity
Clargus is built to connect new evidence to research that already exists and keep investment theses under watch. That continuity matters because research has value beyond the day it is written. An assumption established today can focus a review next quarter. A relationship discovered on one name can inform work on another.
The goal is not an archive that remembers everything indiscriminately. It is a research process that preserves what matters to the decision, makes changes visible, and lets investors challenge their earlier reasoning. The next quarter should add understanding, not require a fresh start.
If this is how you see the desk, we should talk.