Brock, Lakonishok & LeBaron (1992): Can Simple Trading Rules Really Predict the Stock Market?
For decades, investors have tried to answer one fascinating question: Can past movements in stock prices tell us anything about what prices will do next? If stock prices follow a completely unpredictable path, as suggested by the simplest version of the random-walk view of markets, then looking at historical price patterns should not provide investors with a reliable advantage. But technical analysts have long believed otherwise. They argue that patterns in prices can reveal trends and signals that may help investors decide when to buy or sell.
In 1992, William Brock, Josef Lakonishok, and Blake LeBaron published one of the most influential empirical studies examining this debate. Their paper, “Simple Technical Trading Rules and the Stochastic Properties of Stock Returns,” appeared in The Journal of Finance, Volume 47, Issue 5, pages 1731–1764. (Wiley Online Library)
The paper asked a remarkably simple but powerful question: Do basic technical trading rules actually contain useful information about future stock returns?
Their answer was striking. Using historical data from the Dow Jones Index covering 1897 to 1986, the authors found substantial evidence that two simple technical strategies—moving-average rules and trading-range breakout rules—generated patterns in returns that were difficult to reconcile with several standard models of stock-price behavior.
The Big Question Behind the Research
To understand why this paper mattered, we first need to understand the debate surrounding market efficiency.
The efficient-market view suggests that publicly available information should already be reflected in stock prices. If this is true, simply studying yesterday's or last month's price movements should not allow investors to consistently earn abnormal returns.
Imagine that a stock has been rising steadily for several months. A technical analyst might say, “The trend is strong, so the stock may continue rising.” But under a strict random-walk interpretation, the previous movement of the stock should not provide a dependable prediction of its next movement.
This creates a fundamental disagreement.
Technical analysis says:Past price patterns may contain information about future prices.
The random-walk/efficient-market perspective says:Past price movements should not provide a systematic trading advantage.
Brock, Lakonishok, and LeBaron designed their study to examine this disagreement using historical evidence rather than simply arguing from theory.
What Did the Researchers Actually Test?
The researchers deliberately chose simple and widely used technical trading rules rather than complicated mathematical strategies.
They focused primarily on two types of rules:
Moving-average rules and trading-range breakout rules. (Wiley Online Library)
This choice was important. If an extremely complicated strategy produced profitable results, someone could argue that the result was simply the product of excessive optimization. Instead, the authors examined rules that were relatively straightforward and familiar to technical traders.
The underlying logic was simple.
When prices move strongly above a relevant historical average, this can generate a buy signal.
When prices move below the relevant benchmark, this can generate a sell signal.
Similarly, a trading-range rule looks for situations where the market breaks above a previous range or below it, treating these movements as potential signals of a new trend.
The philosophy behind these strategies is that markets may not move randomly from one price to another. Instead, prices may sometimes develop trends or persistent movements.
Understanding the Moving-Average Rule
A moving average is essentially a way of smoothing price movements.
Suppose we observe the daily closing prices of an index. Instead of looking at every individual daily movement, we calculate an average over a particular period.
As new prices arrive, the oldest observation is removed and the newest observation is added. The average therefore “moves” through time.
Technical traders can then compare the current price with this moving average.
If the price moves sufficiently above the moving average, it can be interpreted as a buy signal.
If the price moves below it, it can be interpreted as a sell signal.
The important idea is not the mathematical complexity—it is actually very simple.
The strategy is essentially asking:
Is the current market price behaving strongly relative to its recent history?
If the answer is yes, the trader interprets that movement as evidence of a possible trend.
What Is a Trading-Range Break?
The second major strategy examined by the researchers involved trading-range break rules.
Imagine that a stock or market has been moving between a relatively high price and a relatively low price for some time.
For example, suppose an index repeatedly moves between 100 and 110.
A technical trader might interpret a movement above 110 as a breakout.
The reasoning is that the market has escaped its previous trading range and may be beginning a new upward trend.
Similarly, a movement below the previous lower boundary may be interpreted as a bearish signal.
The important question for the researchers was therefore:
After such a signal occurs, do future returns actually behave differently?
If they do, then the signal may contain information about future market behavior.
The Data: Almost a Century of Market History
One of the most impressive aspects of the research was the length of the historical sample.
The authors used the Dow Jones Index from 1897 through 1986, giving them almost nine decades of market observations. (Wiley Online Library)
This long period was valuable because it allowed the researchers to examine their trading rules across many different market environments.
There were periods of strong economic growth, financial crises, wars, recessions, booms, crashes, and changing investor expectations.
A strategy that appears successful during only one particular market environment may simply be benefiting from unusual circumstances. A very long historical sample provides a stronger test of whether the pattern appears repeatedly.
The Most Important Finding: Buy Signals Performed Better
The results were striking.
The researchers found that buy signals consistently generated higher subsequent returns than sell signals. They also found that returns following buy signals were less volatile than returns following sell signals. (Wiley Online Library)
This is important because the result was not simply:
“Prices go up after buy signals.”
The more meaningful comparison was between the behavior of the market following different types of signals.
The market appeared to behave differently depending on whether the technical rule generated a buy or sell signal.
Even more interestingly, the researchers found that returns following sell signals were negative. According to the authors, this result was difficult to explain using the equilibrium models they considered. (Wiley Online Library)
In simple terms, the trading rules appeared to contain information that standard models were struggling to explain.
Why Was This Finding So Important?
The importance of the paper goes beyond the question of whether moving averages “work.”
The deeper issue was predictability.
If past price information can help identify periods when future returns are systematically different, then stock returns may not behave like a completely unpredictable random process.
That challenges a very strong interpretation of market efficiency.
The paper therefore became part of a larger literature questioning whether financial markets always behave exactly as traditional models predict.
It provided empirical evidence suggesting that patterns in historical prices may contain information about future market behavior.
But the Researchers Did Not Stop at Simple Statistics
One of the strongest features of the study was its attempt to determine whether the results could simply be explained by statistical characteristics of stock returns.
The authors compared their findings with several alternative models, including:
the random walk model
an AR(1) model
a GARCH-M model
an Exponential GARCH model
The authors reported that the returns generated following the technical trading signals were not consistent with these four null models. (Wiley Online Library)
This strengthened the importance of the findings.
If the results had been completely consistent with a standard model of return behavior, one could argue that technical rules were merely repackaging an already understood statistical property.
Instead, the authors found evidence that was difficult to reconcile with these alternatives.
The Role of Bootstrap Testing
Another important methodological contribution was the use of bootstrap techniques alongside conventional statistical analysis. (Wiley Online Library)
Bootstrap methods are useful because researchers can repeatedly resample data to understand how likely a particular result might be under alternative assumptions.
In financial research, this is particularly valuable because stock returns often have complicated statistical properties.
Markets can experience periods of high and low volatility, extreme observations, and other characteristics that make simple statistical assumptions unreliable.
By using bootstrap techniques, Brock, Lakonishok, and LeBaron attempted to make their statistical conclusions more robust.
What Could Explain the Results?
The findings naturally lead to a deeper question:
Why would a simple technical rule work at all?
There are several possible explanations.
One possibility is that markets contain trends.
Suppose new information arrives about a company or the economy. Investors may not immediately incorporate the full implications of that information into prices. Instead, prices may adjust gradually.
This gradual adjustment could create persistence in price movements.
A price begins moving upward.
Other investors notice the movement.
More investors respond.
The trend becomes stronger.
Eventually, the market may incorporate the information more fully.
Under this explanation, technical rules may not be “predicting the future” in a magical sense. They may simply be detecting a persistent process already developing in the market.
Investor Psychology May Also Matter
Another explanation involves human behavior.
Financial markets are ultimately made up of people and institutions making decisions.
Investors do not always process information instantaneously or identically.
Some investors may be slow to react.
Others may follow trends.
Some may become overly optimistic after prices rise.
Others may become excessively pessimistic after prices fall.
These behaviors can create persistence in market prices.
This connects the paper to an important idea in behavioral finance:
Prices may sometimes reflect the gradual evolution of investor beliefs rather than an instantaneous adjustment to perfect information.
The paper itself primarily establishes the empirical behavior rather than proving one psychological explanation. But its findings became important for later discussions about market efficiency and behavioral explanations.
A Simple Example
Imagine that a market index has been moving sideways for several months.
Investors have seen prices fluctuate between two boundaries.
Then suddenly the index rises strongly above the previous trading range.
A technical trading rule generates a buy signal.
If markets were completely unpredictable, there would be no reason for the future returns following this signal to differ systematically from returns following a sell signal.
But Brock, Lakonishok, and LeBaron's evidence suggests that historically, such signals were associated with different subsequent return behavior.
The market was therefore showing signs of structure rather than complete randomness.
Why “Simple” Is the Most Interesting Part
The title of the paper emphasizes “Simple Technical Trading Rules.”
This is actually very important.
The authors were not testing an incredibly sophisticated algorithm containing hundreds of variables.
They were asking whether very basic information—such as the relationship between current prices and historical averages—could have predictive value.
That makes the finding more intuitive.
If simple rules can identify periods with different return characteristics, then perhaps financial markets contain relatively persistent patterns that investors can observe without requiring extremely complicated models.
This idea later became highly influential in quantitative finance.
Technical Analysis vs. Efficient Markets
The study sits directly at the intersection of two competing views.
The technical-analysis perspective
Technical analysts believe that historical price and volume information can contain patterns that may help identify future market movements.
Trends, breakouts, moving averages, and other signals are therefore treated as potentially useful information.
The efficient-market perspective
The efficient-market view argues that publicly available information should already be reflected in prices, making it difficult to systematically earn abnormal returns using information that everyone can observe.
The Brock, Lakonishok, and LeBaron results created an important challenge to the strongest version of the second view.
Their evidence suggested that simple historical-price rules were associated with economically meaningful patterns in subsequent returns. (Wiley Online Library)
What the Paper Does NOT Prove
It is important not to overstate the findings.
The paper does not prove that every moving average strategy will always make money.
It does not mean that investors can simply put a moving average on a chart and become consistently profitable.
And it does not establish that technical analysis is universally superior to fundamental analysis.
The study examined specific rules, a specific historical market index, and a particular historical period.
Real-world implementation also involves transaction costs, taxes, bid-ask spreads, execution delays, and other practical considerations.
Therefore, the correct interpretation is not:
“Technical analysis always works.”
A more academically accurate interpretation is:
“The historical return patterns following these simple technical signals were sufficiently strong and systematic that they were difficult to explain using several standard models of stock returns.”
That is a much more powerful and defensible conclusion.
Why the Paper Remains Important
The real contribution of Brock, Lakonishok, and LeBaron was not simply the discovery of a profitable chart pattern.
The deeper contribution was methodological and conceptual.
The study demonstrated how an apparently simple trading idea could be subjected to rigorous statistical testing.
Instead of asking whether technical analysis “looks convincing” on a chart, the researchers asked:
Do the returns following trading signals behave differently from what standard statistical models would predict?
That transformed the discussion from opinion into empirical finance.
The paper was published in The Journal of Finance in December 1992 and occupies pages 1731–1764 of Volume 47, Issue 5. (Wiley Online Library)
The Bigger Lesson for Investors
The most important lesson from this research is that market prices may contain patterns that are more complicated than a simple random walk suggests.
A price chart is not merely a collection of random numbers.
Sometimes prices can develop trends.
Sometimes those trends can persist.
And sometimes very simple rules can identify periods in which the distribution of future returns appears different.
But markets are also adaptive.
Once investors discover a pattern, they may trade on it. Their trading can change prices and potentially reduce the opportunity.
This means that the existence of a historical pattern does not guarantee that the same pattern will remain equally profitable forever.
Connection With Momentum Research
The Brock, Lakonishok, and LeBaron paper is especially interesting when viewed alongside later research such as Jegadeesh and Titman (1993).
Both studies examine, from different perspectives, the possibility that past price behavior contains information about future returns.
Brock, Lakonishok, and LeBaron examine technical trading signals, particularly moving averages and trading-range breakouts.
Jegadeesh and Titman examine cross-sectional momentum, where recent winning stocks tend to outperform recent losing stocks over subsequent months.
Together, these studies contributed to a broader challenge to the idea that stock returns are completely unpredictable based on publicly observable historical information.
Conclusion
Brock, Lakonishok, and LeBaron's 1992 research became an important landmark in the study of technical analysis and market efficiency. By examining nearly ninety years of Dow Jones Index data, the authors tested simple moving-average and trading-range breakout strategies and found strong evidence that the signals were associated with systematically different subsequent returns. Their results showed that buy signals produced higher returns than sell signals, while returns following buy signals were also less volatile. (Wiley Online Library)
The significance of the paper goes far beyond moving averages. It asks a much bigger question about financial markets:
Are stock prices truly as unpredictable as traditional theories suggest?
The evidence presented by Brock, Lakonishok, and LeBaron suggests that the answer may be more complicated.
Their research reminds us that markets can display trends, persistence, and patterns, even when investors are operating in highly competitive environments. It also demonstrates the importance of testing financial theories against real historical data rather than accepting either technical analysis or market efficiency purely on faith.
Ultimately, the paper's most powerful message is simple:
Sometimes, the history of prices contains more information than a purely random market model would lead us to expect.
And that simple observation helped open the door to a much deeper investigation into technical analysis, return predictability, market efficiency, behavioural finance, and quantitative trading.
Comments