Value investing has produced some of the most successful investors in history. Benjamin Graham, Warren Buffett, and Joel Greenblatt all built their approach on this discipline. Every value investing approach centers on a systematic method. That method identifies stocks that trade below their intrinsic worth.
Over the decades, academics and practitioners have developed numerous scoring models that turn complex financial research into actionable numbers. But which scoring models actually work best, and how do they differ?
This comparison examines the most widely used value investing scoring models in 2026. It covers what each one measures, how to calculate it, its strengths, limitations, and when to use it.
Understanding these models sharpens your investment process. This applies whether you build a long-term portfolio or screen for short-term opportunities. Every model discussed here is available on ValueMarkers so you can screen and compare stocks using any combination.
Why Scoring Models Matter for Value Investors
The core challenge of value investing is separating undervalued stocks from those that are cheap for a reason. Without a systematic framework, investors rely on gut feeling or a single metric like price-to-earnings. Scoring models solve this problem by evaluating multiple dimensions at once. The result is a composite assessment more reliable than any single indicator.
Academic research has consistently shown that multi-factor models outperform single-factor screens. A stock might look cheap on one metric while hiding serious problems. Those problems only appear when you examine earnings quality, balance sheet health, or cash flow strength. Scoring models force this multi-dimensional evaluation, reducing the risk of falling into value traps.
The best scoring models share several characteristics. They use public financial data and produce objective results. Research and backtesting have validated all of them.
Any investor can implement them without specialized software. The models compared below all meet these criteria.
One practical benefit of combining models is protection from value traps. These are stocks that look cheap because the business is broken at its core. A low P/E ratio can mask shrinking margins, rising debt, or even false earnings.
Running multiple scoring models before committing to a position improves your odds. It helps find stocks that are cheap for the right reasons, not the wrong ones.
Piotroski F-Score
What It Measures
The Piotroski F-Score evaluates a company's financial health improvement across nine binary tests spanning earnings strength, leverage and liquidity, and operating efficiency. Each test awards a 0 or 1, producing a total score between 0 and 9. Joseph Piotroski developed it in 2000 to identify which cheap stocks have genuinely improving fundamentals.
How It Works
The nine tests cover earnings, liquidity, and efficiency. The tests check ROA, cash flows, leverage, current ratio, dilution, gross margin, and asset turnover. Each is compared to the prior year. A score of 8 or 9 signals strong fundamental momentum while 0 to 2 signals decline.
A company with positive cash flow, improving margins, and no dilution scores 3 or 4 on those tests alone. Other criteria can push the total score higher.
Strengths
The F-Score is remarkably simple to calculate and interpret. Its binary nature removes bias. Piotroski's original research showed it generated 7.5 percentage points of annual excess returns when applied to value stocks.
The F-Score works across industries and international stock markets. It is especially effective at avoiding value traps by confirming that cheap stocks are genuinely improving.
Limitations
All nine tests carry equal weight despite varying importance. The binary scoring means a company barely passing a test gets the same credit as one passing by a wide margin. The model is entirely backward-looking and does not incorporate any forward-looking estimates. It does not consider valuation directly, meaning a perfect 9 score on an overvalued stock is less useful.
Best Used For
Secondary screening of value stock universes. Apply it after filtering for cheap stocks to identify those with improving fundamentals. Ideal for buy-and-hold long term investing strategies focused on quality at reasonable prices.
The F-Score complements any valuation model. It confirms the business is actually improving, not just looking cheap.
Altman Z-Score
What It Measures
The Altman Z-Score estimates the probability that a company will go bankrupt within two years. Developed by Edward Altman in 1968, it combines five weighted financial ratios into a single discriminant score. The Z-Score is primarily a risk avoidance tool rather than a stock picking tool.
How It Works
The formula combines five components. These are working capital, retained earnings, EBIT, market equity, and asset sales. Each is divided by total assets or liabilities and then weighted.
Sum the five terms to get the Z-Score. Scores above 2.99 indicate safety, 1.81 to 2.99 is the grey zone, and below 1.81 signals distress.
A company in the grey zone with rising debt and shrinking EBIT needs close watching. Small drops can push the score below the distress threshold.
Strengths
The Z-Score has decades of empirical validation with high accuracy in predicting bankruptcy. It provides clear, interpretable thresholds. The model correctly classified 95 percent of bankrupt firms one year before failure in Altman's original study. It captures multiple dimensions of financial health including liquidity, earnings strength, leverage, and efficiency in a single number.
Limitations
The original model was calibrated for publicly traded manufacturing companies and may be unreliable for financial firms, utilities, or service companies. The specific threshold values were derived from a historical sample and may not apply equally to all industries. It does not indicate whether a stock is a good investment, only whether the company is likely to survive.
Best Used For
Risk screening before investing. Check the Z-Score on any company with significant debt before investing. It can catch potential bankruptcies before they become obvious.
Essential for credit evaluation and bond investing. Useful as a first-pass filter before applying other scoring models.
This check matters most for turnaround situations and cyclical industries. Cash flows in these companies can drop sharply during downturns.
Greenblatt's Magic Formula
What It Measures
Joel Greenblatt's Magic Formula identifies companies that combine high business quality with low valuation. It ranks stocks on two factors. Return on capital measures how efficiently a business operates. Earnings yield measures how cheaply you can buy the business relative to its operating profits.
How It Works
Return on capital equals EBIT divided by the sum of net working capital and net fixed assets. Earnings yield equals EBIT divided by enterprise value. Stocks are ranked separately on each metric, then the two rankings are combined.
The top-ranked stocks represent the best combination of quality and cheapness. Greenblatt recommended buying the top 20 to 30 ranked stocks. He excluded financial companies and set a minimum market capitalization threshold.
For instance, a company ranked 15th on earnings yield and 22nd on return on capital has a combined rank of 37. This beats a company ranking 1st on one metric but 80th on the other.
Strengths
The Magic Formula integrates quality and value into a single framework. Screening on both dimensions is more powerful than either one alone. Greenblatt's backtesting showed approximately 30 percent annualized return over 17 years.
The ranking methodology is intuitive and easy to implement. It works well for investors who want a disciplined, rules-based approach to fundamental analysis.
Limitations
The strategy can underperform for two to three years at a time, testing investor patience. It does not account for balance sheet risk, meaning some top-ranked stocks may have excessive debt. The exclusion of financial companies limits its applicability. EBIT-based metrics can miss companies with unusual capital structures or significant non-operating income.
Best Used For
Building a systematically selected portfolio of quality value stocks. Best for long term investing with annual rebalancing. Suitable for investors who want a simple, proven strategy they can implement without deep financial expertise.
It works best when combined with the Altman Z-Score and Beneish M-Score as pre-filters. Those two eliminate the weakest names before the ranking begins.
Beneish M-Score
What It Measures
The Beneish M-Score detects potential earnings manipulation. Messod Beneish developed it in 1999. It uses eight financial ratios to calculate the probability that reported earnings are artificially inflated. A score above negative 1.78 suggests possible manipulation.
How It Works
The model evaluates eight variables. These cover receivables, gross margins, asset quality, sales growth, depreciation, SG&A expenses, leverage, and accruals. Each is measured year over year.
Each variable captures a different pattern associated with earnings manipulation. Companies that manipulate earnings tend to show unusual patterns across several of these dimensions simultaneously.
A company with growing receivables, shrinking margins, and rising accruals triggers multiple warning signs. This pushes the M-Score well above the negative 1.78 threshold.
Strengths
The M-Score correctly identified several high-profile accounting frauds before they were publicly revealed. It provides an early warning system for investors who rely on reported financial data. The model is based on rigorous academic research and has been validated across multiple time periods and geographies. It fills a gap that other scoring models ignore by questioning the quality of the underlying data.
Limitations
False positive rates are relatively high, meaning some legitimate companies may be flagged incorrectly. The model works best for detecting accrual-based manipulation and may miss other types of fraud. It requires year-over-year financial data, which limits its applicability for newly listed companies. A clean M-Score does not guarantee that financial statements are accurate.
Best Used For
Due diligence screening before making significant investments. Run the M-Score on any company you are considering adding to your portfolio. It is most valuable when combined with other scoring models.
It helps ensure the financial data those models rely on is trustworthy. Ideally, apply it before running any other model. There is little point scoring earnings or growth if the underlying numbers cannot be trusted.
Graham Number
What It Measures
The Graham Number calculates a theoretical maximum fair value for a stock. It uses earnings per share and book value per share as inputs. Benjamin Graham created it to represent the highest price a conservative investor should pay. The formula builds in a margin of safety.
How It Works
The formula is the square root of 22.5 times earnings per share times book value per share. The 22.5 coefficient reflects Graham's own criteria. A stock should not trade above 15 times earnings and 1.5 times book value at the same time. When the stock price is below the Graham Number, the stock may trade below fair value by Graham's criteria.
For example, a stock with EPS of $3.00 and book value per share of $20.00 has a Graham Number of about $36.74. If the stock trades at $28, it sits comfortably below that ceiling.
Strengths
The Graham Number is extremely simple to calculate and understand. It provides a single, concrete stock price target rather than a relative score. It enforces discipline by requiring both earnings power and asset backing. The approach has decades of validation behind it and remains a useful sanity check on any valuation analysis.
Limitations
The model was designed for an era when book value was a more meaningful metric. For asset-light technology and service companies, book value significantly understates true business value. The formula assumes all companies should be valued the same way regardless of growth prospects, competitive position, or industry. It tends to screen out fast-growing companies that may still be undervalued relative to their future cash flows.
Best Used For
Quick valuation screening for classic value stocks, particularly in capital-intensive industries where book value is meaningful. Useful as one of several valuation checks rather than a standalone investment criterion. Pairs well with the F-Score for a complete Graham-style value investing approach.
Banks, insurers, and industrial companies suit the Graham Number better. Software and consumer-brand businesses are often poor fits.
Peter Lynch Fair Value (PEG Ratio)
What It Measures
Peter Lynch popularized the PEG ratio. It evaluates whether a stock's P/E ratio is justified by its expected earnings growth rate. It bridges value investing and growth investing. The key question is whether you are paying a fair price for the growth you expect to receive.
How It Works
The PEG ratio equals the price-to-earnings ratio divided by the expected annual earnings growth rate. A PEG of 1.0 means the P/E ratio equals the growth rate, which Lynch considered fair value. A PEG below 1.0 suggests the stock is undervalued relative to its growth.
A PEG above 1.0 suggests it may be overvalued. Lynch generally sought stocks with PEG ratios below 1.0 paired with strong fundamentals.
As a practical illustration, take a company growing earnings at 20 percent per year with a P/E of 15. Its PEG of 0.75 suggests it may be attractively priced.
Strengths
The PEG ratio elegantly combines valuation and growth into a single metric. It prevents investors from automatically dismissing high-P/E stocks that are growing rapidly. Nor should investors automatically embrace low-P/E stocks that are shrinking.
It is intuitive and easy to calculate. It works across different market environments because it adjusts for growth expectations.
Limitations
The PEG ratio depends heavily on earnings growth estimates, which are inherently uncertain and subject to analyst bias. It does not account for balance sheet strength, cash flow quality, or competitive position. Very low growth rates produce misleadingly high PEG ratios, and negative growth rates make the ratio meaningless. It is less useful for cyclical companies or companies in turnaround situations.
Best Used For
Evaluating growth-at-a-reasonable-price (GARP) investments. It helps compare stocks within the same sector to find those with the best growth relative to their valuation. Best combined with fundamental analysis of earnings quality and sustainability.
It becomes most powerful when paired with the Beneish M-Score. Growth stories are the most common targets for earnings manipulation.
Comparing the Models Side by Side
Each scoring model captures a different dimension of value investing. The Piotroski F-Score measures financial health improvement. The Altman Z-Score measures bankruptcy risk. The Magic Formula measures the combination of business quality and valuation.
The Beneish M-Score measures earnings reliability. The Graham Number measures absolute valuation against conservative standards. The PEG ratio measures valuation relative to growth expectations.
No single model captures everything an investor needs to know. The F-Score tells you fundamentals are improving but not whether the stock price is attractive. The Z-Score tells you the company is safe but not whether it is a good investment.
The Magic Formula identifies quality-at-a-bargain but does not check for earnings manipulation. The M-Score validates the data but does not indicate value.
This is precisely why the most sophisticated value investors use multiple models together. Start with the Z-Score to cut bankruptcy risk and the M-Score to remove potential manipulators. From there, use the Magic Formula or Graham Number to find attractively priced stocks. Confirm the selection with the F-Score to verify that fundamentals are improving.
Common Mistakes When Using Scoring Models
Scoring models are powerful tools, but they are easy to misuse. Many investors make the same avoidable errors when they first adopt a model-based approach. Understanding these pitfalls will help you get more out of every model you use.
The most common mistake is relying on a single model. No scoring model was designed to do everything on its own.
The Magic Formula does not check for fraud. The Piotroski F-Score does not measure valuation. The Graham Number ignores growth.
Investors who rely on one model end up with blind spots that can produce painful losses. Use at least two or three models in combination to cover each other's weaknesses.
A second mistake is skipping the Beneish M-Score. Many investors apply valuation and quality models without ever questioning whether the underlying numbers are accurate. If earnings data is false, the P/E ratio, return on capital, and F-Score all rest on false data. Always run the M-Score first - it takes only seconds and can save you from buying into a fraud before any other analysis matters.
A third mistake is ignoring sector differences. Scoring thresholds vary by sector. A rule that works for industrial companies may not apply to software firms, banks, or real estate trusts.
The Graham Number is nearly useless for asset-light technology companies. The Altman Z-Score needs a modified version for financial institutions. Always consider whether the model you are using was designed for the type of company you are analyzing.
A fourth mistake is treating backtested results as guarantees. Every model in this article was validated on historical data. But markets change, valuation norms shift, and accounting standards evolve.
A model that returned 30 percent per year in the 1990s may return much less today. Use backtested evidence as a reason for confidence in a model's logic, not as a promise of future performance.
Finally, many investors abandon a model during periods of underperformance. The Magic Formula, for example, can lag the market for two to three years at a time. Investors who sell out during these stretches miss the eventual recovery and lock in the worst of the drawdown.
Scoring models require patience and consistency. The discipline to stay committed during underperformance is often the hardest part of a rules-based strategy.
Building a Multi-Model Screening Strategy
Here is a practical framework for combining scoring models into a single screening process for long-term investing.
Step one: eliminate risk. Remove any company with an Altman Z-Score below 1.81. Also remove any company with a Beneish M-Score above negative 1.78. This filters out companies at risk of bankruptcy or those that may be manipulating their earnings.
In practice, this step alone can eliminate 20 to 30 percent of the market. When interest rates rise, heavily indebted companies often fail the Z-Score test. Their EBIT coverage shrinks relative to liabilities. Remove them early rather than discovering the problem later.
Step two: identify value. Screen for stocks trading below the Graham Number, or those with earnings yield in the top quartile. This ensures you focus only on stocks that appear genuinely undervalued.
In cyclical sectors like energy or materials, earnings yield spikes during commodity downturns. These moments can create attractive entry points that P/E alone would miss.
Step three: confirm quality. From the remaining stocks, select those with a Piotroski F-Score of 7 or above. This confirms that the cheap, safe stocks in your universe also have improving fundamentals. A score of 7 or higher means at least seven of nine financial health criteria are trending in the right direction.
Real-world examples often include small and mid-cap companies. They quietly improve margins and cash generation while their stock price lags the market.
Step four: validate growth. For any remaining candidates, check the PEG ratio to ensure the valuation makes sense relative to expected growth. Stocks with PEG ratios below 1.0 get additional consideration.
This final step distinguishes between two similar candidates. A stock with a PEG of 0.7 and stable estimates ranks ahead of one with a PEG of 1.4 and uncertain growth. This holds even if both clear the first three steps.
This four-step process dramatically narrows the investment universe. The result is a compact shortlist of safe, well-reported, attractively valued stocks with improving fundamentals. It combines the best insights from each scoring model into a comprehensive framework. Running this process quarterly keeps the portfolio current without generating excessive trading activity.
How ValueMarkers Brings It All Together
ValueMarkers calculates all of the scoring models discussed in this comparison for every stock in its database. The platform's stock screener lets you combine any of these models into custom screens. Set F-Score minimums, Z-Score thresholds, Magic Formula rankings, M-Score filters, Graham Numbers, and PEG limits all in one screen.
Each stock's individual analysis page displays all relevant scores alongside traditional valuation metrics, cash flow data, and fundamental analysis indicators. This gives you a full picture of every company across value, quality, risk, and growth in one place.
Conclusion
The best value investing scoring model is not a single model at all. It is a combination of complementary models that together address valuation, quality, risk, earnings integrity, and growth. Each of the six fills a gap the others leave open. Together they cover valuation, quality, risk, fraud detection, and growth in ways no single metric can.
Layering these models produces a shortlist of safe, honestly reported, attractively valued stocks with improving fundamentals. This multi-model approach takes more work than relying on one metric. The result is a significantly more robust investment process. Start exploring all of these scoring models at ValueMarkers and build a screening strategy tailored to your investment goals.