The analyst’s operating loop
A public-markets analyst isn't finished when a model is built or a pitch is delivered. The job is a recurring loop:
- define the market question;
- collect and test evidence;
- translate evidence into operating assumptions;
- compare those assumptions with the price and expectations;
- recommend a position or no position;
- monitor new information;
- update the thesis and the portfolio implication;
- review the outcome and process.
This loop runs across earnings seasons, industry events, management changes, regulatory decisions, competitor results, and market moves.
Building the company model around the debate
The model should expose the variables that matter to the security. A software company may require customer growth, retention, pricing, sales efficiency, stock compensation, and cash conversion. A bank requires balances, yields, funding cost, credit losses, capital, and share count. A retailer requires units, traffic, ticket, gross margin, inventory, occupancy, and working capital.
The analyst should be able to state which estimates differ from consensus, what evidence supports the difference, and what the current valuation implies. A large model that doesn't isolate the debate is less useful than a smaller model that does.
The research mosaic
Research can include filings, transcripts, competitors, suppliers, customers, former employees, industry data, regulators, technical experts, channel checks, and public web data. Every source has incentives and limitations. The analyst must distinguish public lawful research from material nonpublic information and follow the firm’s compliance procedures.
The strongest evidence often comes from triangulation. Management may say demand is stable, while customer data, competitor commentary, inventory, and pricing suggest otherwise. The analyst should document both the evidence and the confidence level.
Pre-earnings work
A pre-earnings note commonly contains:
- current thesis and position context;
- market or consensus expectations;
- the analyst’s estimates and major differences;
- key operating indicators;
- scenario outcomes for the quarter and guidance;
- the questions that matter on the call;
- valuation and price sensitivity;
- recommended action before the event.
The goal isn't to predict every line. It is to understand the distribution of outcomes and what the price appears to require.
Post-earnings work
After results, the analyst separates the reported fact from the market reaction. The update should explain:
- what exceeded or missed and why;
- whether the quality of the result was better or worse than the headline;
- how guidance and underlying drivers changed;
- changes to estimates and valuation;
- whether the thesis strengthened, weakened, or broke;
- the recommended change to position or monitoring.
A stock can fall after an earnings beat because the beat was already priced, the mix was weak, cash flow disappointed, or forward guidance declined. A useful update explains the expectation gap rather than repeating the press release.
Thesis log and decision history
Maintain the original thesis, date, price, assumptions, valuation, catalyst, risk, and falsifier. Record each material update and the reason for changing the view. This prevents hindsight from rewriting the original decision and reveals whether the process responds appropriately to evidence.
The log should also record non-investments. An idea rejected because the downside was unbounded can be a good decision even if the price later rises. An idea purchased for the wrong reason can be a poor process even if it makes money.
Communication cadence
The analyst communicates at different levels of detail:
- a one-line message for urgent P&L-relevant news;
- a short event note for a result or filing;
- a one-page thesis or earnings preview;
- a full initiation memo and model;
- a live discussion with the PM and trader;
- a postmortem after the position is closed.
Each format should lead with the decision implication. The reader shouldn't have to search through company history to discover whether the analyst recommends buying, reducing, waiting, or exiting.
Interaction with the portfolio
Research quality is necessary but not sufficient. The analyst should understand position size, liquidity, factor exposure, catalyst risk, borrow, and correlation with the rest of the book. A highly attractive idea may deserve a small position if the downside is discontinuous or the evidence is weak. A modest expected return can still be useful if it diversifies the portfolio and is easy to exit.
What changes by fund model
At a concentrated fund, the analyst may emphasize durable business value and management quality. At a pod, the workflow may be more estimate-intensive and tied to near-term catalysts. At a family office, the analyst may compare public securities with private investments, credit, or real assets. At a central-research team, the output may support multiple PMs rather than one book.
The job is therefore best understood as a decision system connected to a specific portfolio, not as a generic sequence of reading and modeling tasks.