Questions: Technical Analysis and Market Efficiency Evidence

Question 1

A researcher tests a moving average crossover strategy on 20 years of historical stock data and finds statistically significant excess returns. The most important methodological threat to interpreting this as evidence against weak-form efficiency is:

• A. The strategy requires too many trades and transaction costs would eliminate any gains
• B. Data mining — with enough tested strategies, some will appear profitable by chance, especially with survivorship bias
• C. The researcher should have tested it on bonds instead of stocks
• D. Moving averages are too simple to exploit real market inefficiencies

Answer: (B)

Data mining (also called data snooping) is the dominant threat to technical analysis studies. If hundreds of rules are tested on the same historical data, some will appear profitable purely by chance. Survivorship bias compounds this: we only hear about strategies that worked, not the many that were tried and failed. A strategy's historical significance must be evaluated against the universe of all strategies tested, not just in isolation. This is why out-of-sample testing on fresh data is essential — documented anomalies often disappear after publication.

Question 2

The academic literature documents both short-term momentum (3–12 month continuation) and long-term mean reversion (3–5 year reversal) in asset returns. Together, these findings are best interpreted as:

• A. Definitive proof that financial markets are not weak-form efficient
• B. Evidence that price history contains some predictive information, but whether this reflects risk premia, behavioral biases, or data artifacts remains genuinely contested
• C. Evidence that technical analysis works in the short run but not the long run
• D. Proof that investors systematically underreact to information in all circumstances

Answer: (B)

The correct interpretation is genuinely contested. Momentum could reflect a risk factor (earning returns as compensation for bearing some risk), behavioral underreaction, or data mining. Mean reversion could reflect behavioral overreaction or rational long-horizon risk. The empirical patterns are real and replicated, but they do not straightforwardly 'disprove' weak-form efficiency, because the efficient market hypothesis predicts only that excess returns should be zero *after* controlling for risk — and identifying the right risk model is the central difficulty.

Question 3

A momentum strategy that earns excess returns above the market could still be consistent with market efficiency if it is bearing systematic risk not captured by the market return alone.

• T. True
• F. False

Answer: True

Market efficiency says prices reflect all information such that no excess *risk-adjusted* returns are available. If momentum strategies load on a priced risk factor — exposure to economic downturns, liquidity crises, or some other systematic risk — then their higher returns are fair compensation, not evidence of a free lunch. The challenge is identifying the correct risk model; apparent anomalies often shrink or disappear when more sophisticated factor models are applied. This is why 'markets are inefficient' requires both showing anomalous returns AND ruling out risk-based explanations.

Question 4

The existence of documented momentum anomalies in academic journals means individual investors can now earn consistent risk-adjusted excess returns by trading on them.

• T. True
• F. False

Answer: False

Once anomalies are published, arbitrage activity tends to erode them — this is sometimes called the 'anomaly decay' pattern. Post-publication returns for documented anomalies are typically much smaller than pre-publication estimates. Additionally, the historical estimates often suffer from data mining and survivorship bias that inflate their apparent profitability. Transaction costs, market impact for large trades, and the limits of arbitrage (short-selling constraints, capital requirements) further reduce the practical exploitability of any documented pattern.

Question 5

Why is weak-form market efficiency better understood as a claim about the *magnitude and exploitability* of price predictability rather than a simple yes-or-no question about whether prices follow a random walk?

Model answer: Some price predictability may exist but still be consistent with efficiency if the excess returns are too small to trade profitably after transaction costs, if they merely compensate for risk, or if they are artifacts of data mining rather than genuine. Efficiency does not require a perfect random walk — it requires that no reliable risk-adjusted profit opportunity remains after costs. The debate over momentum, mean reversion, and technical analysis is really a debate about whether the apparent predictability clears this bar, which depends on risk models, costs, and research methodology.

This framing — efficiency as a question of exploitability, not absolute randomness — is how modern empirical finance approaches the question. The honest answer is that markets are 'mostly efficient' at the weak-form level: patterns exist but are smaller, less consistent, and harder to exploit than naive backtesting suggests. The deeper lesson is methodological: rigorous testing of market efficiency requires careful attention to risk adjustment, data mining, and out-of-sample validation.