What Works on Wall Street: The Classic Guide to the Best-Performing Investment Strategies of All Time

Overview

What Works on Wall Street (1997, updated 2005, 4th edition 2011) is the most ambitious empirical study of investing ever published for a general audience. James P. O'Shaughnessy, founder and chairman of O'Shaughnessy Asset Management, built what he called "The Great Database" — a standardized historical record of all US stocks going back to 1926 — and tested 12+ investment strategies across hundreds of thousands of individual stock observations.

The goal was not theoretical. O'Shaughnessy wanted to know, concretely, which investment approaches actually made money over very long time horizons — and why. The result is a book that systematically dismantles popular myths, exposes the limitations of the efficient market hypothesis, and identifies a handful of strategies that have consistently outperformed.

The Great Database

Before O'Shaughnessy, no one had built a clean, standardized, backward-extensible database of all US equity returns. Main Street databases covered only current large-cap stocks. Academic datasets were short. Survivorship bias was baked into every available series.

O'Shaughnessy spent years constructing a database that included:

• All stocks listed on NYSE, AMEX, and NASDAQ from 1926 onward
• Consistent breakpoints and definitions across the entire period
• Fully delisted (bankrupt) stocks included — not just survivors
• Market value, valuation, and price-momentum calculations standardized

The result: statistical control across 85+ years of US equity history, spanning the Great Depression, multiple world wars, the inflation of the 1970s, the tech bubble, and the 2008 crisis. No other investing book of its era had this level of empirical grounding.

---

The Central Question

O'Shaughnessy's book asks three nested questions:

1. Which characteristics — market cap, valuation ratios, price momentum — are associated with outperformance?
2. For how long does a given strategy actually beat the market? Years? Decades? Or only in specific environments?
3. Why do these characteristics work? Are they compensation for risk, or are they structural inefficiencies that intelligent investors can exploit?

---

Author Context

James P. O'Shaughnessy graduated from the University of Pennsylvania's Wharton School and spent the early part of his career on Wall Street before founding O'Shaughnessy Asset Management (OSAM) in 1988. He was one of the first investment managers to market factor-based institutional mandates — what we now call "smart beta" — and he has run quantitative strategies using a disciplined, systematic approach for over three decades.

His firm has published model portfolios based on the book's findings that have consistently outperformed the S&P 500 over full market cycles, including through the dot-com crash and 2008. O'Shaughnessy has also developed software tools (including the O'Shaughnessy Stock Screening System) that let individual investors apply the book's strategies themselves.

---

Place in the Genre

What Works on Wall Street occupies a unique position: it is both an investing book and a serious piece of financial scholarship. It is more data-intensive than The Intelligent Investor, more accessible than most academic finance papers, and more empirically rigorous than most popular market books.

The book sits at the center of a long-running debate about market efficiency. Eugene Fama's Efficient Market Hypothesis (EMH) claims that, after costs, it is impossible to consistently beat the market using publicly available information. O'Shaughnessy accepts EMH in limited form — markets are "mostly" efficient — but demonstrates that certain persistent, measurable anomalies produce excess returns over decades. This directly contradicts the strong form of EMH.

---

What This Book Is Not

This is not a successor to The Intelligent Investor in spirit or method. Graham's approach is qualitative — buy below net asset value, focus on margin of safety, be conservative. O'Shaughnessy's approach is quantitative: sort on specific ratios, measure performance across 85 years, isolate what works statistically.

The book also does not offer get-rich-quick guidance. Every strategy is tested across full market cycles, including severe bear markets. The book is intended for investors who want to understand markets at a deep level, not speculators looking for a short-term edge.

---

The 4th Edition Polemic

The 4th edition (2011) includes O'Shaughnessy's sharpest broadside yet against the efficient market hypothesis. He lays out a "proof by counterexample" — if markets were truly efficient, value and small-cap strategies would not produce such consistent, statistically significant outperformance over 85+ years. The addition of this material made the book required reading in the "behavioral finance" and "factors" debate of the 2010s.

Core Concepts

The Great Database

The empirical starting point for everything in the book: O'Shaughnessy's "Great Database" — the first clean, standardized, complete record of all publicly traded US stocks from 1926 onward. Standard, commercially available data suffers from three problems the Great Database was built to solve:

• Survivorship bias: Standard databases include only currently surviving companies, inflating historical returns because bankruptcies and delistings are excluded.
• Short time horizon: Academic data and provider data typically start around 1962 (CRSP's first year of full coverage), cutting out the Depression and pre-WWII era.
• Inconsistent methodology: Breakpoints, float definitions, and valuation calculations changed across decades; no single provider used uniform methodology for the full period.

O'Shaughnessy standardized all value and price metrics, applied consistent universe definitions across 85+ years, and — critically — included every delisted stock. This is why his results look different (and more conservative) than many published academic papers.

---

The 11 Factor Categories

O'Shaughnessy tested 11 broad categories of investment strategy. Each was evaluated on historical return, risk profile, and robustness across time periods:

| Factor | Key Metric(s) | Description | |--------|---------------|-------------| | Market Cap | Size | Can small stocks reliably beat large stocks? | | Price-to-Earnings | P/E | Cheap on earnings — the classic value screen | | Price-to-Book | P/B | Graham's ratio, updated | | Price-to-Sales | P/S | The book's most important finding | | Price-to-Cash Flow | P/CF | Value measured on operating cash | | Dividend Yield | DY | Income-focused strategy | | Price-Momentum | 6mo–1yr | Past price trends predicting future returns | | Earnings Growth | EPS momentum | Earnings acceleration as a signal | | Earnings Stability | Low volatility of EPS | Consistent earners outperform erratic ones | | Financial Strength | Interest coverage, debt | Solid balance sheets matter | | PEG Ratio | P/E ÷ growth rate | "Growth at reasonable price" screening |

---

The Market Value Effect

One of the most robust findings in finance is the market value effect: small stocks have returned significantly more than large stocks over long time horizons.

O'Shaughnessy quantified this across every 10-year rolling period in the 1926–1997 database:

• Large-cap stocks (top 10% by market value)
• Mid-cap stocks (middle 40%)
• Small-cap stocks (bottom 50%)
• Micro-cap stocks (bottom 20%)

On average, small-cap value — small stocks with low P/E or P/S ratios — outperformed all other combinations. But the key nuance is that simple "small-cap" (unfiltered for value) produced significant outperformance only when combined with value characteristics.

Small size alone is not enough. Small-cap growth stocks severely lagged. It is the combination of small with cheap that is the engine.

---

Small-Cap Value: The Best-Performing Strategy

After testing every combination, the book identifies one strategy as consistently outperforming across all periods tested:

flowchart LR
    CS["Common Stocks<br/>(All)"] --> LG["Large Growth"]
    CS --> LV["Large Value"]
    CS --> SG["Small Growth"]
    CS --> SV["Small Value"]

    LV -->|"Better than LG"| LV_OUT["Good: Beats large growth"]
    SG -->|"Worse than large"| SG_OUT["Poor: Trails index"]
    SV -->|"Best performer"| SV_OUT["Best: Beats all others<br/>~12-14% annualized<br/>over 85-year sample"]

The small-cap value portfolio in O'Shaughnessy's backtests:

• Outperformed the S&P 500 by 3–5% annualized across the full 1926–1997 period
• Outperformed through the Depression, the 1970s, and the tech bubble
• Had a higher worst-drawdown but recovered faster — value and size are different risks

---

The Superiority of Price-to-Sales

Perhaps the book's most actionable and surprising finding is that P/S is the single best standalone valuation ratio tested. Not P/E. Not P/B. Not dividend yield.

P/S outperforms P/E because:

• Sales are harder to manipulate than earnings
• P/E breaks down when earnings are negative or near-zero (common for distressed and growth stocks)
• P/S captures both profitability and scale without elimination bias

The O'Shaughnessy "Best Value" screen (small-cap, low P/S, positive earnings, high relative strength) produced an annualized return roughly double the market's across the full period.

---

Value vs. Growth

The book delivers a clean verdict: value systematically beats growth over long periods, with significant statistical and practical margins.

Value strategies (low P/E, low P/B, low P/S) outperformed growth strategies (high P/E, high P/B) in every long-term subperiod examined. Growth's best years tend to cluster in the late stages of bull markets — precisely when valuation risk is highest.

Growth stocks do not create the most wealth; they destroy the most on a relative basis. The companies that genuinely grow earnings into excessive valuations are rare. The rest are priced for perfection and disappointed.

---

The Limitations of the Efficient Market Hypothesis

O'Shaughnessy accepts that markets are directionally efficient — you cannot consistently beat them by reading newspapers or trading tips. But his empirical results directly contradict the strong form of EMH.

His argument against strong EMH is structural:

1. Systematic, repeatable anomalies exist. If markets were perfectly efficient, well-known factors like value and size would not produce persistent excess returns after decades of publication.
2. The anomalies are risk-adjusted. Even after adjusting for volatility (Sharpe ratios, Sortino ratios), value and small-cap strategies produce higher risk-adjusted returns than simple indexing.
3. Behavioral biases explain the gaps. Anchoring, extrapolation bias, and overconfidence cause investors to systematically overpay for glamour and underpay for value.

O'Shaughnessy's conclusion is now mainstream: markets are "efficient-ish" — enough to make beating them hard, but not so efficient that systematic factors cannot generate excess returns with discipline and patience.

---

The Core-Satellite Approach

O'Shaughnessy invented (or at least popularized) the core-satellite portfolio approach as a practical implementation of the book's findings:

flowchart TD
    PORTFOLIO["Full Portfolio"]
    CORE["Core: 60-70%<br/>Broad market index<br/>(S&P 500 or total market)"]
    SAT1["Satellite 1: Factor Tilt<br/>Large-cap value + momentum"]
    SAT2["Satellite 2: Small-cap tilt<br/>Small-cap value screen"]
    SAT3["Satellite 3: International value<br/>Developed-market factor index"]

    PORTFOLIO --> CORE
    PORTFOLIO --> SAT1
    PORTFOLIO --> SAT2
    PORTFOLIO --> SAT3

    CORE -.->|"Provides market return<br/>low cost, low tracking error"| BENEFIT1["Diversified base"]
    SAT1 -.->|"Adds factor alpha<br/>small tracking error"| BENEFIT2["Tilted return"]
    SAT2 -.->|"Highest expected alpha<br/>highest tracking error"| BENEFIT3["Growth kicker"]
    SAT3 -.->|"Geographic diversification<br/>+ value alpha abroad"| BENEFIT4["Global diversification"]

The core provides market-rate returns at minimal cost. The satellites apply proven factor tilts. Investors get the best of both worlds: indexed safety plus active upside.

---

The O'Shaughnessy Shark

O'Shaughnessy developed a proprietary screening methodology — sometimes called the "OSAM Shark" — that ranks all stocks on a composite of factors:

• Value: P/E, P/B, P/S, P/CF, dividend yield
• Quality: earnings stability, balance sheet strength
• Momentum: 6-month and 12-month price trends

The result is a ranked list where the "best" stocks (cheap, stable earners with recent momentum) cluster at the top. Portfolios constructed from the top decile have outperformed across all testing periods.

The methodology evolved from rules-based screens to a composite scoring system. The key insight: no single factor wins every year, but combining factors creates more robust returns with smaller drawdowns.

---

Strategy Decay and the Transience of Alpha

Perhaps the most important caution in the book: every successful strategy eventually decays. O'Shaughnessy documents why:

1. Publication effect: Once a strategy is widely known, capital flows into it quickly, compressing the anomaly until the excess return disappears.
2. Discovery and crowding: Academics cover new anomalies; the hedge fund industry arbitrages them away.
3. Survivorship bias in discovery: We study the anomalies that worked. We rarely hear about the hundreds of factor combinations tested that did not work.

O'Shaughnessy's recommendation: accept that alpha is temporary. Create robust processes. Diversify across multiple validated factors. Manage expectations. Do not extrapolate a great 10-year track record into infinity.

---

Process Over Strategy

A theme running through the entire 4th edition: the quality of your decision-making process matters more than any individual stock pick or strategy.

O'Shaughnessy argues that most investment failures are not caused by bad data — they are caused by:

• Abandoning a strategy at its worst point (after underperforming for 2–3 years)
• Strategy hopping when a new idea outperforms
• Failing to understand why a strategy is working (or stopped working)
• Letting emotions override disciplined execution

"The strategy that will make you the most money over the next 20 years is the one you will stick with through 15% market drawdowns and 3-year underperformance periods."

Analysis

Strengths

• Unprecedented empirical scope. No investing book for a general audience has matched the database scale — 85+ years, all US stocks, including delisted companies. This scope alone makes the book unique in the genre.
• Honesty about limitations. O'Shaughnessy openly acknowledges survivorship bias, look-ahead bias, and the decay problem. He does not oversell his findings.
• Demolishes the strong efficient market hypothesis. The empirical evidence presented is difficult to reconcile with perfectly efficient markets. The book provided intellectual ammunition for the "smart beta" revolution that followed.
• Actionable factor research. The P/S finding and the small-cap value combination can be applied today by any investor with access to basic screening tools. This is not theoretical — it is practical.
• Core-satellite as portfolio architecture. The book gave institutional and individual investors a framework that reconciles passive and active management. Core-satellite is now standard in wealth management.
• Accessible to non-academics. Despite the heavy quantitative foundation, the book is written in plain English. The charts and tables are explanatory, not intimidating. You do not need a PhD to follow.

---

Weaknesses

• US-only universe. The database covers only US stocks. International, emerging market, and frontier market data are not tested. The strategies may work differently in less developed markets, but the book cannot say.
• Sample selection bias in factor discovery. O'Shaughnessy tested 11+ factors. The fact that value and small-cap worked does not prove that other untested (or unpublished) factors did not also work. We simply do not know if we are seeing the results of testing thousands of failed combos and cherry-picking the winners.
• Strategy decay is under-documented. The book shows that strategies decay, but it does not provide a systematic framework for detecting decay in real time — only after the fact. A reader might follow the book's guidance in 2002 and underperform for 15 years through 2017 as value fell out of favor.
• Data ends in 1997 (original edition). The 4th edition adds recent data but the core database conclusion still relies heavily on pre-2000 data. The post-2000 era — dominated by passive, ETF, and algorithmic flows — is a different market structure.
• Backtested returns do not survive transaction costs. The book presents gross-of-fee historical returns. Small-cap value strategies have high turnover and bid-ask spreads that eat into real-world returns for all but the largest investors.
• Small-cap liquidity constraints ignored. For institutional investors, the small-cap universe is too small to absorb meaningful capital. O'Shaughnessy acknowledges this but does not deeply explore the liquidity-capacity problem.

---

Comparison to Similar Books

| Book | Author | Key Difference | |------|--------|----------------| | A Random Walk Down Wall Street | Burton Malkiel | Malkiel defends EMH and indexing. O'Shaughnessy systematically refutes strong EMH with the same type of empirical data. | | A Commentary on the Common Sense | O'Shaughnessy (companion) | Practical application of WWOWS findings; more actionable, less academic than the original. | | Expected Returns | Antti Ilmanen | Similar empirical spirit but goes deeper into macro, macro-factor premia, and multi-asset class returns. Ilmanen is more skeptical of persistent anomalies. | | Stocks for the Long Run | Jeremy Siegel | Also empirical and long-term focused. Siegel is more bullish on pure indexing; O'Shaughnessy is more activist. | | The Intelligent Investor | Benjamin Graham | Graham's value method is qualitative and introspective. O'Shaughnessy's is quantitative systematic. Both reach similar conclusions about value, but through very different paths. | | AQR's Alternatives / Das Kapital | Cliff Asness | AQR builds the institutional quant firm on the same factors O'Shaughnessy discovered. Asness's work is more sophisticated but less accessible to individual investors. |

---

Practical Applicability

• For individual investors: Directly applicable. Open an online brokerage account. Use free screeners (Finviz, Yahoo Finance, StockRover) to replicate the small-cap value screen. Core-satellite is straightforward to implement.
• For financial advisors: The core-satellite framework is now industry-standard. O'Shaughnessy's research gave advisors the academic and empirical justification for factor-based active sleeves in client portfolios.
• For quantitative researchers: The database methodology is largely obsolete — modern data providers (CRSP, Compustat, Bloomberg) now have cleaner, more complete histories. But the conceptual structure of how to test strategies across 85-year periods is still relevant.
• For passive investors: The book is a helpful corrective. Indexing is excellent, but it is important to understand what you are indexing and why value and size factors earn their premia within a cap-weighted index.

---

Omissions

• International markets. Strategies tested only in the US. Small-cap value may perform differently in Japan, Europe, or emerging markets, but O'Shaughnessy does not address this.
• Multi-asset class testing. The Great Database covers only US equities. Bonds, real estate, commodities, and alternatives are not included, leaving a significant gap for portfolio construction discussion.
• Current period data (2020s). The 4th edition (2011) ends before the decade-long dominance of large-cap tech/growth (2010–2021) and its subsequent drawdown (2022). These periods provide useful real-world tests for the book's value-growth conclusions.
• Liquidity and capacity. O'Shaughnessy mentions the Small Firm Effect becomes harder to exploit with larger pools of capital. He does not fully model at what portfolio size the small-cap value premium disappears due to crowding.
• Cryptocurrency, hedge funds, and private markets. Not covered. The book is focused on publicly traded US equities, which was appropriate in 1997 but limits its full-spectrum relevance today.

---

Verdict

What Works on Wall Street is a landmark in the popularization of quantitative investing. It is not the most rigorous finance book ever written — better databases and more sophisticated factor models exist today — but it was, and remains, the most accessible rigorous investing book ever published.

The central insights (small-cap value works; P/S is the best standalone value metric; strong EMH is false; strategy decay is real; core-satellite is effective; process beats strategy) have stood up remarkably well across 25+ years of additional data. A few strategies underperformed in the 2010s, but the meta-lesson — that factors are cyclical, that discipline matters more than any single insight, and that alpha is transitory — is more true today than when the book was published.

For any investor seeking to understand why markets behave the way they do and how to exploit persistent inefficiencies systematically, this book belongs at the top of the list. It is dense, but it rewards close reading.