Editorials

Joel Greenblatt's Magic Formula is often misquoted in financial education. Most explainers compress it to a simple rule: buy stocks with high earnings yield (E/P) and strong return on capital. But Greenblatt's original formula pairs earnings yield with return on capital employed (ROCE) using a more sophisticated earnings metric: EBIT divided by enterprise value, not EPS divided by stock price. This distinction matters enormously once a company carries meaningful net cash or debt. A simple P/E ratio ignores the balance sheet entirely; Greenblatt's approach captures it. Understanding both, and why they diverge, is essential to reading how cheap a company's earnings power actually is.

The intuitive version is:

Earnings Yield = EPS / Stock Price

Or equivalently, 1 / P/E Ratio.

If a stock trades at a P/E of 14×, its earnings yield is 7.14%. This is easy to compute and widely understood: it tells you how much of each dollar of stock price you are earning per year, before the company reinvests. For a value investor hunting cheapness, this is the first screen. A higher E/P suggests a bargain.

But this formula has a blind spot. It measures only earnings attributable to shareholders, ignores the company's cash position, and treats net debt as irrelevant. Consider two identical businesses side by side: one with $0 in the bank, the other with $5 billion in cash. By simple P/E, they might look equally cheap, yet the cash-rich version has given you a claim on liquid value you haven't paid for.

Greenblatt's original metric is:

Earnings Yield (Magic Formula) = EBIT / Enterprise Value

Where Enterprise Value = Market Capitalization − Net Cash (or + Net Debt).

This is earnings before interest and taxes, divided by what the business as a whole is worth to its stakeholders—equity holders and debt holders. EBIT is the earnings power generated by the assets themselves, uncolored by capital structure. Enterprise value is the price paid for the entire claim on those earnings, net of any cash the buyer receives.

The virtue of this formula is that it adjusts for balance sheet reality. A company swimming in cash is cheaper on this basis than simple E/P suggests, because the buyer inherits that cash. Conversely, a highly leveraged business is more expensive, because the buyer must assume the debt.

Infosys (INFY), the Indian IT services and AI consulting firm, is a textbook case where the two approaches diverge in instructive ways.

For fiscal year 2025 (ended March 31, 2025), Infosys reported:

• EPS: $0.76
• P/E Ratio: 14.04× (per Moatkeep computed metrics, fundamentals as of July 1, 2025)
• Simple Earnings Yield = 1 / 14.04 = 7.12%

By this metric alone, INFY looks reasonably cheap for a global enterprise software and services player. The earnings yield of 7.12% exceeds the risk-free rate and suggests modest margin of safety at that valuation.

Now fold in the balance sheet. For the same fiscal year:

• Operating Income (EBIT): $4,071 million
• Market Capitalization: $43,273 million
• Cash and Cash Equivalents: $2,861 million
• Total Debt: Negligible (finance costs only $49 million)
• Net Cash Position: $2,861 million
• Enterprise Value = $43,273M − $2,861M = $40,412 million
• EBIT / Enterprise Value = $4,071M / $40,412M = 10.08%

This is the capital: the enterprise value earnings yield for Infosys is 10.08%, nearly 3 percentage points higher than the simple E/P reading of 7.12%.

Why the gap? Infosys carries $2.86 billion in net cash—roughly 6.6% of its market capitalization. When you strip that out of the price, you are valuing the operating business itself at a lower multiple. The buyer gets the earnings power at 10.08% yield plus the cash as an interest-free kicker. The gap between the two yields is the implicit return embedded in the cash position.

The two metrics answer slightly different questions:

1. Simple E/P (7.12%): "How much earnings am I buying per dollar of stock price?" This is equity-centric. It is useful for comparing stocks within the same capital-structure profile (say, among utilities with consistent leverage). It misses balance-sheet opportunities.

2. EBIT/EV (10.08%): "What earnings power am I acquiring for the total price, after accounting for the cash I inherit?" This is asset-centric. It reflects the true earning power available to all stakeholders and reveals whether cash is a buried asset or a drag.

For Infosys, the math suggests the business is cheaper than a simple P/E scan would indicate. An investor shopping for a 10% earnings yield would skip INFY at 14× P/E; yet the enterprise value lens reveals nearly that return. This is no accident: Infosys has been a persistent cash generator and modest capital allocator, returning capital to shareholders via dividends and buybacks (totaling $2.4 billion in FY2025) rather than deploying it into acquisitions or excessive reinvestment. The result is a fortress balance sheet with a secular cash accumulation.

The Magic Formula pairs earnings yield with quality. For Infosys, the quality picture is strong:

• Equity of $11.2 billion; Net Income $3.16 billion → ROE ≈ 28.1% (FY2025)
• Operating leverage is durable: gross margins of 41.7%, operating margins exceeding 29% despite a competitive IT services market
• Capital intensity is low: the business requires modest PPE ($1.5B) relative to $19.3B in assets, generating strong cash conversion

A high earnings yield paired with 28%+ ROE and sub-zero net debt is the profile Greenblatt's formula targets: cheapness without value traps.

When screening for value, neither metric alone is sufficient. The simple P/E is a quick filter, useful for its transparency. But Greenblatt's EBIT/EV method catches what simple P/E misses: the balance sheet as a source of value or drag.

For Infosys, the divergence between a 7.12% simple earnings yield and a 10.08% enterprise value yield is material enough to reshape the investment case. A screen looking for 10% earnings yields would miss the stock using P/E alone. A screen using Greenblatt's method would catch it. Over a large universe, these gaps compound. The most systematic value investors use both approaches in tandem—the simple P/E as a high-level filter, then Greenblatt's metric to separate genuine bargains from balance-sheet illusions.

Table notes: Key valuation and yield metrics for Infosys Ltd (INFY), as of July 1, 2025. Note that the earnings yield of 7.12% (inverse of the 14.04× P/E) is the simple E/P formula; the enterprise value earnings yield, adjusting for Infosys's $2.86B net cash position, reaches approximately 10.08%.