The Equity Research Process, Step by Step

The equity research process is the repeatable workflow an analyst follows to turn public filings and market data into a reasoned view of a company. It moves from a wide funnel of ideas down to a small number of businesses studied in depth, then forecast, valued, written up, and monitored.

This piece describes that process as a general framework. It is not specific to any one firm, and it is not advice. A separate piece covers how PMS firms specifically research stocks; this one stays at the level of the craft that most analysts share.

From a wide funnel to a short list

Research starts with idea generation. There are more listed companies than any analyst can cover, so the first job is to narrow the field to names worth the time.

Screening is the usual tool. An analyst sets filters on things like size, growth, returns on capital, leverage, or valuation, and lets the list fall out. Ideas also come from reading, from a sector shift, from a competitor’s results, or from a company that keeps appearing at the edge of other work.

The output of this stage is a short list, not a decision. It is a set of candidates that clear a first sanity check and deserve a deeper look.

Understanding the business and the industry

Before a single forecast, the analyst has to understand what the company actually does and how the industry around it is structured.

That means answering plain questions. How does the business make money. Who are the customers and how sticky are they. What does the competitive landscape look like, and who holds pricing power. Where does the company sit in its value chain. What regulation shapes it.

Industry structure matters as much as the single company. A good business in a brutal industry and a mediocre business in a benign one can look very different once the surrounding forces are clear. This stage produces the qualitative map that everything numerical later hangs on.

Gathering and cleaning the data

This is where the hours go, and it is the least glamorous part of the job.

Financial statements arrive across many years and many filings, in formats that shift over time. Companies restate, reclassify line items, change segment definitions, and report on different calendars. Before any of it can be trusted, it has to be pulled together, aligned, and reconciled so that a number in one year means the same thing as the number above it.

A large share of the total effort in equity research is spent gathering, standardising, and reconciling data. The judgement, the forecasting and the valuation, is a smaller share of the clock even though it carries most of the weight.

An analyst who skips this step builds on sand. Clean, consistent history is the foundation, and getting there is tedious, manual, and easy to get wrong.

Reading filings and listening to the calls

With a clean base in hand, the analyst reads the primary documents closely.

Annual reports, quarterly results, notes to the accounts, related-party disclosures, and the management discussion all carry signal. So do the earnings calls, where management frames the quarter and answers questions that are not in the printed numbers. The point is to read what the company says about itself, in its own words, and to notice what it does not say.

This is where a variant view often forms. If the numbers, the filings, and the call transcript point somewhere different from the consensus story, that gap is worth writing down and testing.

Building the model and forecasting the drivers

Now the work turns forward. A financial model is a structured way to translate a view of the business into projected statements.

The discipline is to forecast drivers, not just totals. Revenue is broken into the things that actually move it, volume and price, or subscribers and average revenue, or units and realisation, depending on the business. Margins are tied to cost structure and operating leverage. The balance sheet and cash flow are built to be internally consistent with the profit and loss.

A model is only as good as its assumptions, and its real value is in making those assumptions explicit and testable. Sensitivity to the key drivers, and a look at plausible upside and downside cases, matters more than a single precise-looking number.

Valuation

Valuation asks what the projected business is worth, and whether the current price already reflects it.

Analysts use a mix of approaches: discounted cash flow, multiples applied to earnings or cash flow, and comparison against peers and the company’s own history. No single method is definitive, so a range is more honest than a point. The output is a view on value, framed as a range and tied back to the forecast, not a target dressed up as certainty.

Writing the thesis, the risks, and the variant view

The write-up is where scattered work becomes an argument.

A useful research note states the thesis plainly, lays out the evidence, and is honest about what could break it. Three parts do most of the work:

• The thesis: the core reason the view holds, in a few sentences.
• The risks: what would make the thesis wrong, and what to watch for.
• The variant perception: where and why this view differs from the market, since a view identical to consensus rarely earns its keep.

Writing forces clarity. Arguments that felt solid in the head often fall apart on the page, and that failure is a feature, catching a weak thesis before it costs anything.

Monitoring over time

Research does not end when the note is filed. A view is a living thing, and it has to be maintained.

Each new quarter, each guidance change, each competitive move is a test of the thesis. The analyst checks whether the drivers are tracking the forecast, updates the model, and asks the uncomfortable question: given what I now know, would I still hold the same view. When the facts change, the view changes with them.

The stages at a glance

Stage	What it produces
Idea generation and screening	A short list of candidates worth studying
Business and industry study	A qualitative map of the company and its structure
Data gathering and cleaning	A clean, consistent financial history
Filings and earnings calls	Primary-source read and an early variant view
Model and driver forecasts	Projected statements tied to explicit assumptions
Valuation	A value range set against the current price
Thesis write-up	A stated view with risks and a variant perception
Monitoring	An updated view as new facts arrive

Sell-side and buy-side, briefly

The same craft splits into two homes. Sell-side research is produced by brokerages and published to a broad set of clients, often carrying a rating and reaching many readers. Buy-side research is internal to a fund or manager and exists only to inform that firm’s own decisions, so it is rarely published and answers to one audience. The stages above apply to both; what differs is who the work is for and whether it ever sees daylight.

A note on infrastructure: much of the friction in this process, especially the data gathering and reconciliation, is repetitive work that tooling can carry. Altys Labs builds infrastructure that supports this kind of workflow for Indian equities, so that more of an analyst’s time can go to judgement rather than plumbing.

Practical takeaway

If you are learning this process, resist the urge to jump straight to the model. The order matters. Understand the business before you value it, clean the data before you trust it, and write the thesis before you believe it. And once the note is done, keep watching, because the hardest part is not forming a view but being willing to change it when the facts do. This is a process for thinking clearly about a business, not a shortcut to a recommendation.