How to find undervalued stocks: a 5-step research method

The 5 steps at a glance

How to tell if a stock is undervalued: three tests

A price says nothing on its own. Three comparisons tell you whether a stock is undervalued or overvalued:

1. Against its own history. Is the P/E or EV/EBIT below its 10-year median? If so, what changed?
2. Against its peers. Is it cheaper than companies with similar growth, margins and debt?
3. Against intrinsic value. Is the price clearly below your value range? This is the only test that looks at the business rather than at other prices.

A stock that passes all three is a research candidate. Passing only the first two means the stock is cheap relative to other prices, not to the business: the whole sector may be expensive, or the company may be in decline. A stock well above its value range is overvalued on your assumptions, however good the business.

Step 1: Screen on valuation (P/E, EV/EBIT, FCF yield)

A screen does not find bargains; it shrinks thousands of companies to a few dozen worth reading about. Use several ratios, since each can be distorted.

These thresholds are starting points, not rules: “normal” multiples differ by sector, and higher interest rates tend to lower the multiples investors will pay. For growth companies, our GARP (growth at a reasonable price) guide adapts the screen with the PEG ratio.

Step 2: Estimate a range of intrinsic value

Intrinsic value is what a business is worth based on the cash it can return to shareholders, discounted back to today. Three methods are common:

• Discounted cash flow (DCF): project free cash flow for 5 to 10 years, add a terminal value and discount at your required return. Complete, but sensitive to assumptions.
• Graham formula: a quick earnings-based estimate for stable companies (see below).
• Normal multiple: a reasonable P/E or EV/EBIT on average earnings, as a cross-check.

Always produce a range. In the worked example below, moving the discount rate from 9% to 10% cuts the DCF value from about $57 to $49, while 8% lifts it to $68. Test your own assumptions in the intrinsic value calculator, or read how to calculate the intrinsic value of a stock.

Step 3: Demand a margin of safety

Benjamin Graham called the margin of safety the central concept of investment (The Intelligent Investor, chapter 20). It is the gap between your estimated value and the price, there to absorb your own errors.

Margin of safety = (intrinsic value − price) / intrinsic value. With a base-case value of $57 and a price of $40: (57 − 40) / 57 = 30%.

Many value investors look for roughly 15–25% on stable, high-quality companies and 30–50% on cyclical, indebted or young ones. If the margin disappears in your bear case, the price offers little protection.

Step 4: Filter for quality (ROIC, margins, balance sheet)

Cheap shares of weak businesses tend to stay cheap. Quality checks tell you whether your estimated value will grow, hold or erode:

• ROIC above the cost of capital: earning 12% on invested capital with a 9% WACC creates value as the company grows; earning 6% destroys it.
• Stable margins over a full cycle, not just the last year.
• Manageable debt: net debt below about 2x EBITDA and comfortable interest cover, for non-financial companies.
• Real cash: free cash flow close to net income over 3 to 5 years.
• Limited dilution: a rising share count shrinks your share of the business.

Check these figures in the 10-K and 10-Q filings, free on SEC EDGAR.

Step 5: Run the value-trap checklist

A value trap looks cheap on its ratios but keeps getting cheaper, because the business is worth less every year. Before adding a stock to your research list, ask:

• Is revenue falling for structural reasons (technology, regulation, lost customers) rather than a passing cycle?
• Are margins shrinking year after year?
• Is debt rising, or is a large refinancing due at higher rates?
• Are dividends or buybacks funded with borrowed money?
• Does the low P/E rest on peak-cycle or “adjusted” earnings, with free cash flow well below profit?
• Is the share count rising, or is management buying growth through expensive acquisitions?

Then answer the question that matters most: why is this stock cheap, and what would change the market’s mind? Write down what would prove your thesis wrong, such as two years of falling margins, and check it at each quarterly report.

Is there a formula to find undervalued stocks?

No single formula finds undervalued stocks, but three classics make useful screens:

• Graham formula: V = EPS × (8.5 + 2g) × 4.4 / Y, with g the expected annual growth in % and Y the current AAA corporate bond yield. Try it in the Benjamin Graham formula calculator.
• Graham number: √(22.5 × EPS × book value per share), the most Graham would pay for a defensive stock.
• Margin of safety: (value − price) / value, which turns any estimate into a research threshold.

Each is only as good as its inputs: optimistic growth inflates the Graham formula, and the Graham number penalises asset-light companies.

How Warren Buffett and Benjamin Graham look for undervalued stocks

Graham, Buffett’s teacher, looked for statistical bargains: a low P/E and P/B, a strong balance sheet and a wide margin of safety.

Buffett kept the price discipline but moved toward quality. In his 1989 shareholder letter he explained why a wonderful company at a fair price beats a fair company at a wonderful price. He also added three filters: a circle of competence, a durable competitive advantage and management that allocates capital well. Graham gives you steps 1 to 3, Buffett steps 4 and 5; more in 15 investing lessons from Warren Buffett.

Worked example: a hypothetical company through the five steps

Company A is a hypothetical US-listed maker of industrial components: revenue $3.0 billion, 100 million shares, price $40, earnings per share $3.20, free cash flow $3.00 per share.

Outcome: Company A becomes a research candidate, not a purchase. The next steps are to read the contract terms in the 10-K and to write down invalidation points, such as “contract lost” or “operating margin below 12%”. If either happens, the bear case ($37, below the price) becomes the base case. Our stock research reports follow this structure: thesis, intrinsic value range, margin of safety, scenarios, catalysts and invalidation points.

Educational research, not advice. Company A is hypothetical; thresholds are common starting points, not rules; scanner scores are research priorities, not recommendations. See the full disclaimer.

Using the stock valuation scanner as a first filter

Software does step 1 well. The Bullish Stock Alerts undervalued stock screener covers US-listed stocks and gives each eligible one a research-priority score out of 100: 40% valuation, 25% quality, 15% growth, 15% risk and 5% technical context. A score of 75–100 means high priority, 60–74 research candidate, 45–59 watchlist and below that limited data. None of these categories is a recommendation.

The free public preview shows up to 8 rows (ticker, company, sector, priority category, update date), delayed by at least 24 hours. Members get the full table, detailed scores, advanced filters, sorting by score or margin of safety and company research views. Missing values show as N/A, never as zero. Steps 2 to 5 remain your job; the scoring methodology explains where the score stops. Full access: €228 a year, or €683 for lifetime access (see pricing).

Value investing for beginners: habits that matter more than ratios

Ratios are the easy part. Most beginner mistakes come from process:

• Keep a watchlist, not a buy list. Candidates can wait.
• Write the thesis first, including what would prove you wrong.
• Diversify and avoid borrowing to invest. Careful analysis is still sometimes wrong.
• Review on results days, not price moves. Quarterly reports test the thesis.

Checklist: before a stock goes on your research list

• I can explain how the company makes money in two sentences.
• P/E, EV/EBIT and FCF yield are low against history and peers.
• I have bear, base and bull values with written assumptions.
• The margin of safety matches how predictable the business is.
• ROIC has beaten the cost of capital for several years.
• Debt, free cash flow and share count pass the quality checks.
• I know why the stock is cheap.
• My invalidation points are written down.
• Key figures match the latest 10-K and 10-Q.

Frequently asked questions

Version française : comment analyser une action en bourse