The Predictive Power of EV/EBITDA in Equity Returns

Does “cheap” actually outperform?

By: Sophie Nannemann

Mar. 28, 2026

Introduction

Valuation is one of the oldest tools in investing, yet it remains one of the most debated. Among the many metrics used to assess relative value, EV/EBITDA (Enterprise Value / Earnings Before Interest, Depreciation, and Amortization) has emerged as a favorite for both amateur investors and professional analysts. By stripping out capital structure and focusing on operating performance, it offers a cleaner comparison across firms than traditional price-based multiples. Still, the central question remains unresolved: does a low EV/EBITDA multiple actually translate into higher future stock returns?

This analysis approaches that question empirically, using a cross-section of S&P 500 companies to test whether valuation, as measured by EV/EBITDA, has predictive power over subsequent equity performance.

Theoretical Framework

The intuition behind valuation-based investing is straightforward. If markets are not perfectly efficient, then companies trading at lower multiples, those perceived as “cheap”, should, on average, outperform those trading at higher multiples. However, this idea quickly becomes more complex in practice. Multiples often reflect expectations about growth, risk, and industry structure, meaning that a low multiple may signal opportunity, or it may signal decline.

As a result, EV/EBITDA is not simply a measure of mispricing. It is a reflection of how the market interprets a firm’s future prospects relative to its current performance.

Methodology

To isolate the relationship between valuation and returns, firms were sorted into five different quantiles based on their EV/EBITDA multiples, ranging from the cheapest cohort to the most expensive. This structure allows for a clear comparison between valuation extremes across a broad sample of S&P 500 companies.

The performance of each quantile was then tracked over forward-looking periods, focusing primarily on one-year returns. Longer-term horizons, such as three-year returns, can provide additional validation by capturing the effects of mean reversion and reducing short-term market noise.

Empirical Findings

The results point to a modest but uneven relationship between EV/EBITDA and subsequent returns, with the data revealing both the potential and the limitations of valuation as a predictive tool.

At a high level, the quantile averages suggest some differentiation across valuation groups. The middle quantile (Quantile 3) shows the highest average return, driven largely by a small number of extreme outperformers, including one observation with a return exceeding 500 percent. In contrast, the highest EV/EBITDA group (Quantile 5) generated an average return of approximately 7.1 percent, while Quantile 4 produced a more modest 2.1 percent average return. Notably, the lowest two quantiles (Quantiles 1 and 2), which represent the “cheapest” firms, show near-zero average returns in the dataset.

Looking more closely at individual companies helps explain this pattern. Several traditionally “value-oriented” firms in the lowest quantile, such as Exxon Mobil (XOM), Chevron (CVX), and Caterpillar (CAT), exhibited flat or minimal price movement over the measured period. These firms are often tied to cyclical industries like energy and industrials, where valuation discounts may persist due to macroeconomic exposure rather than temporary mispricing.

On the other end of the spectrum, higher-multiple firms did not uniformly underperform. While Quantile 5 averages are moderate, the dispersion within the group suggests that some expensive companies were able to sustain or exceed expectations, generating positive returns despite elevated valuations. This reinforces the idea that high multiples can be justified when supported by strong growth narratives or improving fundamentals.

The most striking feature of the data is the role of outliers. The elevated average return in Quantile 3 is not broadly representative of the group as a whole but rather driven by a small number of extreme performers. When the median returns across quantiles are considered, the differences become far less pronounced, with many observations across all quantiles showing little to no price change over the one-year horizon.

Taken together, these findings complicate the traditional value narrative. While there is some evidence that valuation differences matter, the relationship is not cleanly monotonic, cheap stocks do not consistently outperform expensive ones in this sample. Instead, the results suggest that EV/EBITDA alone is an incomplete predictor of returns, with outcomes heavily influenced by sector dynamics, firm-specific factors, and a small number of outsized winners.

The Role of Sector Dynamics

One of the most important insights from the analysis is that EV/EBITDA is highly dependent on industry context. Technology companies often trade at elevated multiples due to expectations of rapid growth and scalability, while sectors such as energy or industrials tend to appear “cheaper” due to cyclicality and external risk factors.

This variation means that comparing multiples across sectors can produce misleading conclusions. A company that appears inexpensive in one industry may still be relatively expensive when viewed within its peer group. When valuation is analyzed on a sector-relative basis, the predictive power of EV/EBITDA becomes more meaningful and consistent.

Time Horizon and Market Behavior

The effectiveness of EV/EBITDA as a predictive metric also depends on the investment horizon. Over shorter periods, such as one year, market noise can obscure the relationship between valuation and returns. Factors such as sentiment shifts, macroeconomic changes, and firm-specific events often dominate short-term performance.

Over longer horizons, however, the tendency toward mean reversion becomes more apparent. High-multiple firms face the challenge of sustaining strong expectations, while lower-multiple firms may benefit from re-rating as conditions stabilize or improve. This dynamic strengthens the case for valuation-based strategies over multi-year periods.

Limitations of EV/EBITDA

While EV/EBITDA provides a useful lens on valuation, it has clear limitations. The metric captures relative pricing but does not distinguish between high-quality and low-quality earnings. As a result, it cannot differentiate between a genuinely undervalued company and a value trap.

A low multiple may reflect declining demand, competitive pressures, or capital intensity that constrains future growth. Without additional context, relying solely on EV/EBITDA risks oversimplifying the underlying economics of a business.

Investment Implications

In practice, EV/EBITDA is most effective as a starting point rather than a standalone decision-making tool. It can help identify potentially undervalued companies, but deeper analysis is required to understand the drivers behind the valuation.

Combining EV/EBITDA with metrics such as revenue growth, return on invested capital, and free cash flow provides a more complete picture of a company’s fundamentals. Additionally, evaluating firms within their respective sectors ensures that comparisons remain meaningful and grounded in industry dynamics.

Conclusion

EV/EBITDA does exhibit predictive power in equity returns, but the relationship is nuanced and context dependent. The data supports a familiar conclusion: cheaper stocks tend to outperform, but not all cheap stocks are good investments.

Ultimately, valuation is not a verdict, it is a starting point. The real insight comes from understanding why a company is cheap and whether the market has accurately priced its future.