State the claims of technical analysis and sort the ones backed by evidence from the ones that are not.
Only one part of technical analysis, momentum, has strong research support. Moving average rules and chart patterns have little, and no part of technical analysis guarantees a return. This module sets out what each part claims and how you can test a claim. It does not give trading signals or recommend that you trade.
Take Marcus (an invented learner on the course). He has completed eight modules on reading accounts and valuing businesses. Most weeks someone in his group chat posts a chart with lines drawn across it, usually with a confident caption and a rocket. Investing forums are full of the same thing: a stock "bouncing off support", a "golden cross" on an index, a "cup and handle" forming on a chip stock. Marcus wants an honest answer on whether technical analysis deserves his time.
Technical analysis uses past prices and trading volume to judge where prices may head next. It leaves out the business behind the shares. Suppose you gave a technical analyst Larkspur (an invented company). They would set aside its customers and its margins and study only how its share price and volume have moved.
The most accurate way to describe a chart is as a record of other people's trading. It shows the prices at which buyers and sellers agreed, how keen they were to trade, and how many shares changed hands. All of that is genuine information about the market. Whether any of it predicts future prices is a different question, and the evidence is what settles it.
The sceptical case begins with the efficient market idea, which Eugene Fama laid out in a 1970 review. In its weakest form it is about price history alone: any dependable, exploitable pattern in past prices would be traded on until it vanished. Studies since then have mostly confirmed this, with momentum as the notable exception.
Momentum is the tendency of recent winners to go on winning for a time. Narasimhan Jegadeesh and Sheridan Titman reported it for US shares in the Journal of Finance in 1993. They looked at returns over the previous three to twelve months and found that the best performers tended to keep beating the worst performers in the months after. Researchers have since found momentum in many other stock markets and in other asset classes. In 2012 Tobias Moskowitz, Yao Hua Ooi and Lasse Pedersen reported a related effect across futures markets, where an asset's own return over the past twelve months tended to predict its return in the next month.
Of everything in technical analysis, only momentum has made it into mainstream academic finance and into the factor models fund managers use. It still has serious weaknesses: it can crash sharply, and the trading it needs is expensive. Lesson 10.5, "Momentum and relative strength: the part with academic backing", goes through those weaknesses in detail.
The rest of the field has a thinner record. In 1992 William Brock, Josef Lakonishok and Blake LeBaron ran simple moving average rules over almost a century of Dow Jones Industrial Average data, and in that data the rules showed predictive power.
Ryan Sullivan, Allan Timmermann and Halbert White went back to the question in 1999. Thousands of rules had been tried on the same data, and when that many are tried, some will look good purely by chance. After adjusting for this, they found the results much weaker, and the strongest rules failed to hold up in the years that followed.
Researchers see this again and again: a rule performs well on the data it was found in and then fades. Chart patterns such as head and shoulders and triangles have produced mixed results when studied. The ones that seem profitable tend to shrink once you count trading costs and the number of patterns that were tested. Lesson 10.6, "Data mining, overfitting and transaction costs", explains why this happens and how to guard against it.
A chart of a single stock over a stretch when a rule worked is an example of the rule working once. It tells you nothing about whether the rule works.
Give a chart claim the same scepticism you would give a fund manager's brochure, no more and no less. You applied this standard to fund returns in Module 1, lesson 1.5, "Annualise, compare periods and spot cherry-picked start dates", and to smooth track records in lesson 2.3, "Sharpe and Sortino ratios, and when each misleads". Before you believe a chart claim, you need all of the following:
a rule stated precisely in advance a test on years of data that played no part in designing the rule every trading cost deducted a comparison with simply buying and holding over the same period the worst drawdown of each approach included in that comparison
One week a message in Marcus's chat says: "When a stock closes above its 200-day moving average, it keeps rising. Buy the cross." (The rule is an invented example.) Before making up his mind, Marcus writes down the five requirements. The post meets none of them. All it offers is one chart of one stock during the months when the rule happened to work. It shows no test on other years, no costs, no buy-and-hold comparison and no drawdown.
A chart is still useful because price and volume describe the market you are about to trade in. They can show whether a stock is in a calm or a violent phase, where many orders sat in the recent past, and whether a move came on heavy or light trading.
First, decide on the trade for reasons that have nothing to do with the chart, using the fundamentals from the earlier modules on reading accounts and valuing businesses. Next, read current conditions from price and volume. Then use those conditions to decide when and where to place the order you have already chosen. Used this way, a chart does everything the evidence supports. Lesson 10.7, "Use a chart to time entry within a decision made on fundamentals", walks through the method.
Think of a chart claim you've seen online recently, in a forum, a video or a chat, and get ready to write out the evidence that would make you believe it.
Pick one chart claim you have seen online. Write it down, then write the evidence you would need before you believed it.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).