What's Wrong With Traditional Technical Analysis?
- Aidan Lee-Wen
- Jul 31, 2021
- 8 min read
Updated: Jul 31, 2021
I. Introduction
II. The History Technical Analysis
III. Random Walk
IV. Cognitive Bias
V. Thinking Deeply
VI. Conclusion
I. Introduction
Technical analysis is more popular than ever as people are drawn to trading and investing in financial markets. There are more trading education courses available than ever before, but it doesn't take long to realize that most of the concepts and strategies they teach simply don't work.
Here I will present information that will cause many to question the validity of what they've been taught about trading markets and technical analysis. I may even provoke feelings of defensiveness or even anger as many have relied on methods or indicators that I will deconstruct. The goal is to make you question, challenge, and therefore think deeply about what you do and why you do it when trading financial markets. First, I will present a brief history of technical analysis and its development.
II. The History of Technical Analysis
Most of the key icons of technical analysis expanded on theories formulated and taught by Charles Dow - the creator of The Wall Street Journal and the Dow Jones Industrial Index in the late 1800's. Ideas like volume confirming trends and practices of drawing on price charts are over a century old.
These theories and practices evolved until the revolutionary bible of technical analysis was published by John Magee in 1948, Technical Analysis of Stock Trends. Magee's teachings and methodologies still persist, forming what we know as modern technical analysis. Magee taught the classic chart patterns that we still use today, such as the "head and shoulders" and triangles. He charted everything from prices to volumes, trying to find patterns in everything possible.
On the other side of the world in the 18th century, price candlesticks were invented in Japan as a way to track the price action of rice. Rice traders were then able to identify patterns in its price action to supposedly become successful trader. Out of this, candlestick patterns were born.
Through empirical analysis, Japanese rice traders developed a list of price patterns they found to be significant in predicting future price movements. The harami cross, hanging man, and shooting star are just a few names they gave these patterns. The Western world continued using bar charts until candlesticks were introduced to them in 1991 when Steven Nison published Japanese Candlestick Charting Techniques. Wall Street was quick to adapt candlestick charting and almost all traditional trading education courses, both retail and institutional, continue to teach a blend of John Magee's price patterns and the ancient candlestick patterns of the Japanese rice traders. But these aren't the only evolutions of technical analysis that are still popular today. Other, even more questionable theories and practices still persist as superstition crept in.
In 1938, Ralph Nelson Elliott published The Wave Principle which detailed his theories on how equities markets must follow the mystical "natural laws of the universe." He taught that measuring these natural forces can be used with Fibonacci numbers and specific "waves" that the market follows. Hence, Elliott Wave Theory and Fibonacci indicators were born. In the early 1900's, William D Gann created a variety of indicators based on geometry and astrology. He published his forecasts of commodities markets based on astrological phenomena, believing that planet alignments and zodiacs would signal price tops or bottoms in some markets. Gann was and is still regarded as a supposed trading legend, even though Alexander Elder revealed the following in his book in 1993:
"I interviewed W.D. Gann's son, an analyst for a Boston bank. He told me that his famous father could not support his family by trading but earned his living by writing and selling instructional courses."
The technical analysis taught today still teaches many of these theories and practices out of tradition. Unfortunately, not much forethought is given to their upbringing and level of practicality. These theories and practices were never statistically proven strategies, but ideas that derived from observation. From Magee's price patterns to Gann's astrological indicators, these were nothing more than early, primitive attempts to find patterns and understand financial markets.
Perhaps some of the ancient strategies may have or at least had merit in the past. The Japanese rice traders may have used their candlestick patterns successfully to earn great wealth. But markets evolve quickly and vary across asset classes, so it cannot be assumed that any old theorized trading strategies would automatically carry over.
Not long ago, prop-firm traders were very successful using basic strategies that almost a child could understand. However, prop-firm trading quickly died due to emergence of quantitative and algorithmic trading, which came as a result of the development of formulas such as the Black-Scholes Model and the creation of the internet. This led to the quick obsolescence of those same basic, prop-firm technical strategies as markets are evolving faster than ever. Now, this doesn't mean that all patterns are false or outdated, but those that are real and significant probably won't be found in a book from the 1920's. In fact, the real patterns with a significant edge are likely hidden in the cloud of randomness.
III. Random Walk
In 1973, a Princeton University professor named Burton Malkiel published a popular book, A Random Walk Down Wall Street. This book popularized the Random Walk Hypothesis, which states that changes in equity prices are random and therefore market participants cannot predict price movements. He argued that most of what we see in stock charts is not meaningful. While the content of Malkiel's book introduced subjects of great controversy to the finance world, there is great value found in exploring its claims. Are most price movements random? Are we really just finding patterns in randomness?
(hint: the answer is, 'definitely more than most would like to accept')
Let's take a look. Below is a price chart of 100% randomly generated currency price data:
Patterns can be found within this randomly generated data. Below is a charting where many of John Magee's technical patterns can be found:
Here we can draw support and resistance lines, trend lines, triangles, and all sorts of patterns that are taught in traditional trading courses. This begs the question, how much of what we see in charts is random? How do we determine if a pattern is valid or just a product of randomness?
I would encourage everyone to research and question the idea of 'randomness' further. There is incredible depth to this topic that will not be covered here. But whatever we do, we cannot escape the fact that randomness impedes the practice of visual pattern recognition and therefore technical analysis. Finance has made this same mistake of finding patterns where none exist for years.
Now, this doesn't mean that everything is random. If this were the case, the selection of winners and losers of markets would also be random. Valid patterns do exist, but are difficult to find and are even short-lived as markets quickly evolve. Most new technical traders entering the industry fail to understand that the majority of what they see in the short-run is really just randomness. Consistency and profitability cannot be built on randomness, but cognitive bias keeps poor beliefs, strategies, and indicators alive.
IV. Cognitive Bias
Imagine you are playing a coin flip game where you earn $1 for every heads and lose $1 for every tails. You play the game ten times and win 8/10 times. Congratulations, you made a profit!
Does this mean that you should continue to play the game? Is the expected value of this game really positive and in your favor? Not at all! In fact, your net profit/loss will be completely random. In theory, the Expected Value of this game is $0 and we would realistically expect some fluctuation for net gain/loss around $0. But in reality, the results can deviate much further than expected.
Below are a two simulated outputs for playing this coin flip game 10,000 times. Take a look at the results:
Both random outputs yield extreme and opposite results. Once again, it is easy to find patterns within these completely random outputs. If John Magee searched for patterns in almost everything possible, it is highly probable that he unknowingly charted random, meaningless information. But cognitive bias allocates more psychological weight to winning trades and provides confirmation, while losers are dismissed as outliers. Someone once said,
"With any charting technique, there’s a human tendency to look at the chart and only see cases where the rule holds. We ignore the multitude of cases where the rule doesn’t hold."
People still use bad strategies and indicators today for this reason. They will swear by moon-phase indicators, Fibonacci Retracements, or other superstitious ideas because they appear to have significance. But observation fails us and patterns can be found in randomness, so cognitive bias keeps bad beliefs alive.
How can we ensure that our strategies and indicators are valid and significant? How can we ensure that we're not overfitting to a random past and finding patterns where none exist? Making smart decisions and using meaningful strategies and indicators starts with thinking deeply about what we do and why we do it.
V. Thinking Deeply
Rather than reverse engineering a belief, strategy, or indicator out of a vacuum, they should be based on some fundamental thesis that we have reason to believe to be true. This is similar to finding the truth in order to create a belief, rather than creating a belief and then cherry-picking reasons to confirm that belief. This requires thinking deeply about our strategies, indicators, and beliefs and decisions in general.
Why should this strategy or indicator work? How is this indicator calculated and why should it apply to this symbol, asset class, or markets in general? What is the fundamental thesis behind this strategy or indicator? These are all questions that are often overlooked.
Indicators like RSI and MACD are statistically shown to not work when trading currency pairs the same way they do in equities. Why? Because currency pairs don't mean revert like equities do. Why? Because it's a pairs trade and with little-to-no mean rate of growth to revert back to.
Asking this "why" question is imperative to building a correct fundamental thesis. Why do I believe that stock ABC will pivot here? Why do I believe that there is a high probability of XYZ to happen right now? And why do I believe that the answers to those questions are significant and meaningful? If we follow the "why" question, it'll bring us back to our original core belief and will easily shine a spotlight on any holes in our beliefs and methodologies.
Here's an example:
Q: Why could the MACD indicator's calculation be meaningful?
A: The MACD indicator takes the difference between two moving averages, one long and one short. The shorter moving average will be more reactive to current price movements while the longer will lag behind. By taking the difference between the two, I get a clever way of measuring the 2nd derivative - or acceleration - of a price trend.
The issue is if we asked most people what the MACD is, how it's calculated, and why it can be meaningful, very few would be able to provide an answer beyond "that trading course told me to use it."
If everyone thought deeply about what they do and why they do it, the only users of Gann indicators or Fibonacci Retracements would be those who believed in astrology or superstition. If we asked, "why would the Fibonacci level of 23.6% be meaningful?" the only answer we could deduce would rely on superstition and a belief in universal magical ratios. Yet, so many non-superstitious people continue to use a superstitious indicator because they fail to think deeply about what they are using and why they are using it while cognitive bias continues to tell them it works. This is a common mistake that everyone makes - including myself - because we are all human.
VI. Conclusion
The key points should be taken away are:
- Most of technical analysis comes from early 1900's ideas. While there is truth to some ideas and methods, these should be regarded as nothing more than early, primitive attempts to find patterns and understand markets.
- Patterns can be found in complete random data. It is easy to find patterns where none exist.
- Cognitive bias keeps false beliefs, strategies, and indicators alive. The way to combat this is to think deeply about what we are doing/using/believe and why we are doing/using/believe it.
- Strategies and indicators should be based on some fundamental thesis that we have reason to believe to be true.
Technical analysis can be incredibly useful if done correctly. The most successful technical traders I have met are those who pair their technical analysis with some other, validating edge. This can be a variety of things such as economics, company and financial statement analysis, or quantitative analysis. Whatever you choose to do, you need to find an edge and you should be able to explain that edge (if you can't explain it, then you probably don't have one).
I hope this post brings insight into your trading and decision making. Markets are more complex and competitive than ever and it takes more than finding patterns in a price chart to be successful. Continue working hard and good luck!
Comments