Wall Street analysts publish research on public companies constantly — ratings, price targets, model updates. Financial media treats upgrades and downgrades as major news events. Stocks move on analyst calls. Understanding what these ratings actually mean — and what they don't — is worth a few minutes of your time before you let them influence a decision.
The ratings scale — and why it's not what it looks like
Different banks use slightly different terminology but the scale is similar everywhere:
- Buy / Overweight / Outperform — analyst recommends buying; expects stock to outperform the market
- Hold / Neutral / Equal Weight / Market Perform — analyst is neutral; neither recommending buying nor selling
- Sell / Underweight / Underperform — analyst recommends selling; expects stock to underperform
Here's the thing: across Wall Street research as a whole, buy ratings typically outnumber sell ratings by roughly 5-to-1 or more. If the rating scale were applied honestly, you'd expect a roughly even distribution. The reason for the imbalance is structural — investment banks that employ analysts also have investment banking relationships with the companies those analysts cover. Issuing a sell rating on a company's stock is a guaranteed way to damage that banking relationship. Analysts who cover a company rarely bite the hand that might feed the bank future IPO fees or M&A advisory work.
What "hold" actually means
When an analyst downgrades a stock from buy to hold, that is often the closest thing to a sell recommendation they'll publish while still maintaining a banking relationship. "Hold" in analyst language frequently translates to "we'd actually sell this but can't say that publicly." Watch for buy-to-hold downgrades on names with significant investment banking relationships — that's the signal. A hold on a stock that was a buy is meaningful. A hold from initiation is more neutral.
Price targets — useful context, not prophecy
Price targets are 12-month fair value estimates derived from the analyst's model — usually a DCF (discounted cash flow), sum-of-the-parts, or comparable company analysis. They're calibrated to the analyst's earnings estimates and the multiple they think the stock deserves.
They are not predictions. They're model outputs, and models are wrong constantly. The average analyst price target has historically been optimistic — set 10–15% above current prices even in flat markets. When the consensus price target is 5–10% above current price, that's not the same as genuine analyst conviction that the stock will hit that level.
What's more useful than the absolute target: the direction of change. A cluster of price target increases across multiple firms signals improving sentiment. A wave of target cuts after an earnings miss often continues — analysts update their models conservatively and the stock often doesn't fully bottom until the cut cycle ends.
What analyst research is actually useful for
Analyst reports are best used as a starting point for research, not a conclusion. The best parts: the industry analysis, competitive landscape, and model assumptions — not the rating or target. A good sell-side analyst has spent months building deep knowledge of a specific industry. Their understanding of how a company's revenue and margin structure works, who their competitors are, and what the key catalysts are — that's valuable context even if you ultimately disagree with their rating.
The initiations of coverage are worth reading carefully. When an analyst initiates on a company for the first time, the write-up is usually their most comprehensive and least incentive-distorted piece. Compare that against earnings-day "we're raising our target by $5" notes, which are often just acknowledging what happened rather than adding insight.
Frequently asked questions
Should I buy a stock because an analyst upgraded it?
Use upgrades as a research trigger, not a buy signal. When a stock gets upgraded, ask why — what changed in the analyst's view? Is it a valuation call, a fundamental improvement, or a near-term catalyst? Read the report if available. Then make your own judgment about whether the thesis is sound. Stocks often move significantly on upgrades before you can act — so acting on the news alone means buying after the immediate price impact has already occurred.
Why do analysts almost never say sell?
Structural conflict of interest: investment banks that employ analysts also want to maintain M&A and underwriting relationships with the companies those analysts cover. Issuing sell ratings damages those relationships. Regulatory reforms after the dot-com era required more separation between research and banking, but the incentives still push toward optimism. An analyst at a bank with no investment banking relationship — or at an independent research firm — tends to have a more honest distribution of ratings.
What is a consensus estimate?
The average of all analyst estimates for a company's earnings or revenue, compiled by data providers like FactSet and Bloomberg. When financial media says a company "beat estimates," the estimates are this consensus number. Companies are judged against the consensus each quarter — the beat/miss dynamic that drives much of the post-earnings stock reaction. See the full earnings report explainer for how this works in practice.
What is a "street high" or "street low" estimate?
The most optimistic (street high) or most pessimistic (street low) estimate among all covering analysts. When a company beats the street high estimate — the most optimistic analyst's model — that's a particularly strong result. When it misses the street low — failing to meet even the most bearish expectation — that tends to cause significant selling.