Analyst Research Reports: How to Read One in 15 Minutes

Page one of an analyst report holds the rating, the price target and a headline. Most readers stop there. That page is a summary of a summary, and the parts that tell you whether to believe it sit further back: the estimates, the valuation, the risks and the disclosures. Read the report in six passes, in a fixed order, and you can judge it in about 15 minutes. The order matters, because each pass tells you how hard to lean on the next.

Check the date and the trigger

Look at the publication date before anything else. Then find what prompted the report: an earnings release, a guidance change, a deal, a new analyst picking up coverage, or nothing in particular. A report written before the latest quarter describes a company that has since changed. Skip it, or read it only for background.

The trigger also tells you what the analyst had to work with. A note written the morning after results was often put together in hours, so its estimates may be a first pass, while an initiation report written over weeks usually carries a full model and the longest explanation of how the business makes money.

Write the thesis in one sentence

Every report argues something. Find the paragraph that says why the stock should go up or down, usually near the front, and compress it into a single sentence of your own. Margins expanding as a product line scales. A balance sheet repair nearing its end. Can’t write it? The report may not have one.

Keep that sentence. Each later section either supports it or doesn’t.

Compare the estimates

The estimates page shows the analyst’s forecasts for revenue, earnings per share and often cash flow over the next two or three years. Put them next to two things, the company’s own guidance from the latest earnings release and the same firm’s previous report if you can find it, and expect the text to explain any estimate that sits far above guidance.

Estimates that moved sharply since the last report need an explanation too. Check what changed and why. A forecast that rose while the thesis stayed word for word the same is a warning sign.

Rebuild the price target

Most price targets come from a simple calculation: a multiple applied to a forecast. The report should state both. If it says the target is based on a multiple of next year’s earnings, you can redo the sum and see how much rides on each input.

A 10% miss on earnings and a slightly lower multiple take nearly a fifth off the target. Neither input is extreme. That is why price targets are the weakest part of any report: they compound two estimates, each of which can be wrong. Ask where the multiple came from. A peer average? The stock’s own history? A number the analyst simply prefers? The P/E ratio page covers what a multiple does and does not capture.

Targets built on a discounted cash flow model hide more inputs, chiefly the growth rate and the discount rate. Look for a sensitivity table showing the target under different assumptions. No table, no stated rates? Weigh the target accordingly.

Read the risks

Reports list risks near the end, often as a block of generic items: competition, the economy, the loss of senior staff. Look past those for the risks specific to this company, the ones that would break the thesis sentence you wrote down.

Then ask whether the estimates already allow for any of them, since a report that names a customer concentration risk but models every customer renewing has listed the risk without pricing it.

Read the disclosures and the rating history

The last pages carry the disclosures. FINRA Rule 2241 governs research produced by member firms, and in practice it means the report has to tell you about relationships that could bias it, such as the firm having been paid by the company for banking work, along with how the firm’s ratings split across its whole coverage. Read them. A firm that rates most of what it covers a buy is using a different scale from one with an even split.

Many reports also include a chart of the stock’s price with the firm’s past ratings and targets marked on it. Look at it. Did earlier targets get hit, or quietly cut as the price fell? Reading analyst ratings covers rating scales and consensus figures in more depth.

Keep a log

When a report changes what you do, write down the date, the rating, the target, the thesis sentence and what you did. Review the log every few months against what happened, and measure the positions properly, because a vague sense that a research source has been helpful survives far longer than it should when nobody is checking the actual results. Calculating portfolio returns shows how to measure them with cash flows included.

The stock research subscription guide covers what to check before paying for reports, and free and paid ways to get analyst research reports covers where they come from.

When two reports disagree

Sometimes two firms cover the same stock with opposite ratings. Good. Put the two estimates pages side by side and find where they part: revenue growth, margins, the multiple, or the year the target is set for. Usually one input explains most of the gap.

Then decide which input you believe, using the filings rather than either analyst’s say-so, and you have done more useful work than averaging two price targets into a consensus figure that hides the very disagreement you were trying to understand.

Note the dates as well. A newer report may simply include a quarter the older one never saw.

What goes wrong

• Acting on the rating without reading the estimates.
• Reading a report written before the latest results.
• Accepting a target without knowing its multiple.
• Skipping the disclosures.
• Never checking how the firm’s past calls turned out.