Wall Street analysts spend their careers researching companies and issuing ratings. But what do terms like “Strong Buy,” “Hold,” or “Underweight” actually mean? And should you follow their recommendations?

What Are Analyst Ratings?

Analyst ratings are recommendations issued by research analysts at investment banks, brokerage firms, and independent research companies. These analysts study companies in depth—reviewing financials, interviewing management, analyzing competitors—and publish their conclusions.

A typical analyst report includes:

A rating (Buy, Hold, Sell, or variations)
A price target (expected stock price in 12 months)
Investment thesis (reasoning behind the rating)
Financial projections (revenue, earnings estimates)

Understanding Rating Scales

Different firms use different terminology, but they generally map to three categories:

Bullish	Neutral	Bearish
Strong Buy	Hold	Sell
Buy	Neutral	Strong Sell
Outperform	Market Perform	Underperform
Overweight	Equal Weight	Underweight
Accumulate	—	Reduce

What Each Rating Means

Buy / Outperform / Overweight The analyst expects the stock to outperform the market or its sector. They recommend increasing your position.

Hold / Neutral / Market Perform The analyst expects the stock to perform in line with the market. Not recommending new purchases, but not selling either.

Sell / Underperform / Underweight The analyst expects the stock to underperform. They recommend reducing or eliminating your position.

What Are Price Targets?

A price target is the analyst’s prediction of where a stock’s price will be in 12 months. It’s calculated using valuation models like:

Discounted Cash Flow (DCF) – Future cash flows discounted to present value
Comparable Analysis – Multiples compared to similar companies
Sum of the Parts – Valuing divisions separately

How to Interpret Price Targets

Current Price vs Target	Implication
Target 20%+ above current	Strong upside expected
Target 5-20% above	Moderate upside
Target near current price	Fairly valued
Target below current	Downside risk

Example: If a stock trades at $100 and an analyst sets a $130 target, they expect 30% upside.

The Analyst Upgrade/Downgrade Cycle

Rating changes often move stocks more than initial ratings:

Action	Meaning	Typical Impact
Upgrade	Rating improved (e.g., Hold → Buy)	Positive price reaction
Downgrade	Rating lowered (e.g., Buy → Hold)	Negative price reaction
Initiation	Analyst starts coverage	Depends on rating
Price Target Raise	Increased expected value	Moderately positive
Price Target Cut	Decreased expected value	Moderately negative

Upgrades and downgrades are particularly impactful because they signal a change in the analyst’s view.

Tracking Analyst Ratings Data

You can access analyst ratings and price targets through the FinBrain API:

import pandas as pd
from finbrain import FinBrainClient

fb = FinBrainClient(api_key="YOUR_API_KEY")

# Get analyst ratings as DataFrame
df = fb.analyst_ratings.ticker("AAPL", as_dataframe=True)

print(df.head())
#                institution      action      rating targetPrice
# date
# 2024-01-15   Goldman Sachs     Upgrade         Buy        $210
# 2024-01-10  Morgan Stanley  Reiterated  Overweight        $225

# Filter for upgrades only
upgrades = df[df["action"] == "Upgrade"]
print(f"\nRecent upgrades: {len(upgrades)}")

# targetPrice comes back as a formatted string ("$275") or None,
# so strip it before doing arithmetic.
targets = pd.to_numeric(
    df["targetPrice"].str.replace(r"[$,]", "", regex=True),
    errors="coerce",
)
print(f"Average analyst target: ${targets.mean():.2f}")

# Count by rating
rating_counts = df["rating"].value_counts()
print(f"\nRating distribution:\n{rating_counts}")

Building Signals from Analyst Data

Consensus Tracking

Track how many analysts are bullish vs bearish:

def analyst_consensus(df):
    """Calculate analyst consensus from ratings"""
    bullish_ratings = ["Buy", "Strong Buy", "Outperform", "Overweight"]
    bearish_ratings = ["Sell", "Strong Sell", "Underperform", "Underweight"]

    bullish = df[df["rating"].isin(bullish_ratings)]
    bearish = df[df["rating"].isin(bearish_ratings)]

    total = len(df)
    if total == 0:
        return "no_coverage"

    bull_pct = len(bullish) / total
    bear_pct = len(bearish) / total

    if bull_pct > 0.7:
        return "strong_consensus_buy"
    elif bull_pct > 0.5:
        return "moderate_buy"
    elif bear_pct > 0.5:
        return "moderate_sell"
    elif bear_pct > 0.7:
        return "strong_consensus_sell"
    else:
        return "mixed"

consensus = analyst_consensus(df)
print(f"Analyst consensus: {consensus}")

Upgrade Momentum

Track recent upgrades vs downgrades:

from datetime import datetime, timedelta

def upgrade_momentum(df, days=30):
    """Calculate upgrade/downgrade momentum"""
    cutoff = datetime.now() - timedelta(days=days)
    recent = df[df.index >= cutoff]

    upgrades = len(recent[recent["action"] == "Upgrade"])
    downgrades = len(recent[recent["action"] == "Downgrade"])

    if upgrades > downgrades:
        return f"positive ({upgrades} upgrades, {downgrades} downgrades)"
    elif downgrades > upgrades:
        return f"negative ({upgrades} upgrades, {downgrades} downgrades)"
    else:
        return "neutral"

Limitations of Analyst Ratings

Keep these factors in mind:

Factor	Consideration
Conflicts of interest	Investment banks may rate clients favorably
Herding behavior	Analysts often move together
Lagging indicators	Ratings often follow price moves
Sell ratings are rare	Most ratings are Buy or Hold
Consensus can be wrong	Popular stocks get over-covered

The Sell Rating Problem

Studies show that fewer than 10% of analyst ratings are Sells. This “ratings inflation” means:

A Hold often means “we’d sell but can’t say it”
A downgrade to Hold from Buy is often bearish
True Sell ratings are significant events

Key Takeaways

Analyst ratings range from Strong Buy to Strong Sell (with various terminologies)
Price targets represent 12-month expected values
Rating changes (upgrades/downgrades) often move stocks more than initial ratings
Analyst consensus provides sentiment but isn’t always predictive
Consider conflicts of interest and herding behavior

Track analyst ratings and price targets with the Analyst Ratings Dataset and API Reference.

What Do Analyst Ratings Mean? Buy, Hold, Sell Explained