Is the Bulk of Technical Analysis “Thinking in Words”?
Until now I thought that the main problem of technical analysis is bias of all kinds. A tweet from N. N. Taleb yesterday got me thinking. Maybe there is a more serious problem than bias: thinking in words.
In my book Fooled By Technical Analysis I discuss the perils of classical charting, indicators and modern quant approaches. Classical charting fails mainly because of clustering illusion and the tendency to see significant patterns in random samples of a population. Post-classical technical analysis fails because indicators are derivatives of price and are neither necessary, nor sufficient conditions of price action. Modern quant approaches suffer from data-mining bias because they rely on backtesting. Essentially, all of the above failures can be attributed to the difference between the true distribution mean, or expectation, and the sample mean, or the average. In finance sufficient samples are large in general and the distributions are governed by power laws. Due to that the true mean may not be known. This means that no analysis can forecast longer-term returns and risk of ruin is certain if you stay long enough in the game. That much I already knew long ago. But yesterday I realized the extent of another major cause of the failure of technical analysis that is not necessarily due to bias but is related to “thinking in words”.
I rebut some word thinking in my book 15 Lies About Trading And Investing but I have not realized that it is major mechanism of generating false ideas and rules. An example of word thinking is when someone takes the words trend and friend and comes up with the rule “the trend is your friend”. This is Lie 1 in my book because of the following empirical facts
- Any point on a trend can mark a trend reversal
- Prices fall faster than they rise
- Profiting from trends still requires an edge
I will add to the above that friendship has no real meaning in trading. This is a frequently quoted rule but in essence it is vacuous and not true at all scales. For example, scalpers, high frequency, swing and arbitrage traders do not care about the direction of the trend. Only trend-followers do but the trend can suddenly reverse and become the enemy. The rule “the trend is your friend” is word thinking.
Another example is when we take the words loss, cut, profits and run. The result is the “cut losses short and let profits run”. This is also word thinking with limited application. As I write in my book in Lie 2, in real trading you never know whether a small loss with reverse to a win and a large win will reverse to a loss. The methodology of dealing with profit targets and stop-losses depends on the characteristics of the particular distribution. In some markets it may be advantageous to let losses run due to mean reversion. By introducing a stop-loss mechanism profitability may be destroyed. This is not a principle that applies at all scales; it is word thinking.
Another example is combining emotions, systematic and trading into “systematic trading removes emotions”. I understand sales departments playing with words but not individuals who pose as experienced systematic traders. The truth is that everyone is aware of account performance, even in real-time. When there is a drawdown everyone is worried. There is no such thing as starting a system and returning back after one year to take a look at performance. Maybe a retail trader could do that but professional traders and funds are more responsible than that. Knowing equity performance induces the same emotions as manual trading. A system could be shut down by clicking on Deactivate button. Then a few good trades can be missed. This is same thing as with manual trading. These rules are due to word thinking. People with no formal training in logic and who have not read Bertrand Russell and Gottlob Frege are usually unaware of the profound implications of the widespread tendency to think in terms of words.
A more recent example is that statement that “Only X months have contributed to S&P 500 gains in last N years.” A specific variation is that “only 6 months have contributed to S&P 500 total return since 2000.” The reason that this is word thinking is that all returns contribute to gains, both positive and negative, and their ordering in time determines the specific equity curve. By selecting 6 returns that when combined generate a return equal to buy and hold and then using irrelevant words, such as responsible, to designate returns, this is word thinking that has no relation to markets and finance. Frequently this type ofreductionist word thinking is used to impress the public and for click-bait.
What is the alternative to the widespread tendency of technical analysts but also people in general to think in terms of words? The answer is mathematics, probability, statistics and hard work. Rules are made to be broken and there are very few universal rules. Those who identify rigorous rules that work at all scales gain fame and recognition. The rest is local rules and mostly thinking in words. In markets you do not need to identify rigorous universal truths. Those that sell them are deceiving the public. All you need is locals rules that work in specific domains of application and markets. But one must avoid word thinking because it may get catastrophic.
This article was originally published in Price Action Lab Blog
If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY
About the author: Michael Harris is a trader and best selling author. He is also the developer of the first commercial software for identifying parameter-less patterns in price action 17 years ago. In the last seven years he has worked on the development of DLPAL, a software program that can be used to identify short-term anomalies in market data for use with fixed and machine learning models. Click here for more.