Why Technical Analysis is the Astrology of Investing
June 18, 2009
– Comments (34)
This was originally a response to a comment on Anchak's recent blog. Visit that link if you want to read more. This blog explains some of my reasoning as to why I believe technical analysis (TA) is essentially equivalent to practicing astrology and provides no significant value to investors:
---------------------------------
While I won't deny that many scientists and mathematicians have searched for patterns in the stock market that can lead to accurate predictions of the future, many of the top scientists of the 17th century also tried to turn common metals into gold. "Scientists" aren't infallible. They are capable of the same delusions as all other human beings.
[Anchak's] critique of "fundamental analysis" is based on a bit of a straw man. You say that fundamental analysis relies solely on past data. This isn't strictly true. Certainly, most of the variables and the foundation come from financial statements (i.e. past data), but my valuations are never strictly derived from that info.
Rather, I run future projections to see how different future scenarios could affect the valuation of a company. I also analyze a company's underlying business and try to come up with future projections of growth/decline. I tend to buy in with a large margin for error due to my own conservatism with money.
While, it's impossible to predict the future, the past is your best guide. If you look at seemingly worst-case scenarios and buy in with a large margin of safety, you can outperform the market over time by a significant margin.
The Difference With Technical Analysis
How does this differ from technical analysis? It may seem similiar; after all, you are relying on past data for both FA and TA. But with FA, you have some level of protection. Technical analysis isn't related to the underlying company you are investing in. Rather, one who invested strictly on TA would be solely trying to predict human psychological patterns for a large set of underlying transactions.
The main guide for TA'ers is past charts. From the past, TA'ers try to predict the future. But of what? Of a company's performance? Not really. TA'ers are trying to predict the future actions of an undefined group of traders based on the premise that all human psychology will be relatively similiar over time and can be fully understood, explained, and predicted by charts. That underlined part is key. It suggests that all of human psychology can be understood by mathematical models that are based on charts that are based on pricing of individual transactions of many individuals over a period of time.
How do TA'ers Fare?
Certainly some TA'ers fare better than others. The ones who understand portfolio strategies and have some comprehension of risk-reward will do better than the ones who don't over a long period of time. The TA'ers who do not base their analysis solely on technicals, but also look at the fundamentals will fare better than those who don't over the long-term.
But here's the real question --- which group of TA'ers would fare better over time: the ones who dogmatically use charting in order to predict future market behavior? Or the ones who inject their own common sense and their own understanding of their own emotional reactions to their analysis?
My issue is this --- unless a computer can analyze the charts and predict with an extremely high rate of accuracy the patterns of the future, how do you know that TA is actually working? How do you know that this isn't basically the case of the brilliant scientist who thinks he can turn a common metal into gold? Or that is to say, how do you know the individual using TA isn't simply convincing their self it works when in actuality, it plays virtually no positive role in their performance?
You can obviously argue that FA creates the same conundrum. You can say that the patterns of a stock are so random, that you would not know whether or not your FA actually played a part in your performance. But there's an issue with that argument.
Fundamental Analysis in Perspective
On a very basic level, when you buy a stock, you are buying a company. You actually own a share of that company's assets and earnings. If you purchased more than 50% of the stock in a company, you could take it over, take over the board of directions, and pay out 100% of earnings as dividends to yourself (and the other shareholders). So in actuality, you be entitled you to those earnings.
Why is this relevant? Because it puts things in perspective. It suggests that stocks are not "random". While human psychology might make stocks drift upwards and downwards due to the uncertainty surrounding future earnings and various other factors (e.g. the honesty of management), stock pricing is not completely random. You are buying a share of a company and you are entitled to your share of the results.
If you buy a company with $0 net assets that earns $1 with a 3% growth rate for the rest of eternity, how much is that future revenue stream worth to you? While the answer might differ from one investor to another, it's probably safe to say that not too many people would pay $200 for that future revenue stream and most people would be willing to pay $1 for it (assuming they had the cash).
The Underlying Foundation for TA?
Would you be so confident in buying out an entire company based on the signals a chart gives you? Probably not --- I don't think any TA'ers would buy out an entire company just because the chart said the stock was going up. So in a very real sense, TA has no real solid footing on which to stand.
You can convince yourself it works. On some basic level, I think nearly all of us can glance at the past charts, take our own understanding of human psychology, and apply it forward. I'll even argue that this can be beneficial to one's own performance. But I wouldn't argue that by merely coming up with elaborate methods for studying charts, that you can somehow accurately project the future.
TA and Occam's Razor
Occam's razor is a scientific principle that the explanation of a phenemonon should involve as few assumptions as possible. That the simplest explanation is probably the best explanation. That the simpler model is probably the better model. The more assumptions and intracacies you add, the more likely your explanations are inaccurate.
TA fails occam's razor. Have you ever read a GoodVibe chart? GV's blogs are enjoyable reading, but those charts are otherworldly in complexity. Sure, I understand the principle of "what goes up must come down", and that's true to a limited extent, but stock market investing involves no gravity. A rapid growth company can go up for twenty years if management is sound and stays ahead of the game. Sure, it will have dips here and there, how do you explain the future performance of the stock via very complex mathematical analysis of a chart that is based on trading patterns of individual investors who are buying into a company?
There is no direct line with TA. You are merely gauging market psychology and there's no realistic mathematical model to predict human psychology on a mass level when you can't read the future and don't know what stimuli humans will even interact with.
What if I use some TA that measures volume indicators. It tells me the stock is overbought. The next day, the company releases earnings and absolutely blows away expectations, earning $1 per share rather than $0.25. It's pretty easy for me to guess the psychological reaction to this news --- model or no model. But how would I have predicted that stimuli to begin with? The answer is, you wouldn't have if you were simply staring away at a chart.
Time is Money and TA is a Waste of Time
Don't get me wrong. I'm not claiming there's nothing useful about technical analysis. If I thought it was completely worthless, I wouldn't read GoodVibe's posts. I'm not even suggesting that I don't use some level of TA when investing. But I use TA at a very, very basic level. I look at the past stock charts for companies I'm interested in. I try to gauge market sentiment. I might even look at volume for a stock. But it takes me less than a minute to examine and analyze these items.
If I had to quantify it, I'd say that stock picks are based 98% on analysis of the fundamentals and 2% on technical analysis. I'm not saying that's a magic ratio. I'm simply saying that on a basic level, there's a foundation with which to stand on with fundamental analysis. My guess on the psychological element is simply a guess, which is why you have no foundation with TA. You are basically sticking your money on the roulette wheel and hoping for the best.
More importantly, I'm probably not going to do a significantly better job analyzing those technical trends with extremely complex mathematical models and charts than I would in one minute from examining the technical data. I can see when the price chart has a line rocketing upwards or one plummetting downards. I can see when volume is high. I don't need any complex analysis to uncover that.
The bottom line however is that if I buy a company at a certain price and I'm "wrong" about its prospects, I still might make a reasonable return if I bought in with a margin of safety based on the fundamentals. If I buy a stock based solely on a prediction of whether it will go up or down based on a reading of a chart, there is no protection. For all I know, that stock go lose 100% of its value or gain 500% in the next few months.
For this reason, complex TA like EWP is simply a waste of time. The chart has no relation to the underlying fundamentals driving the company. Your time would be better spent analyzing those fundamentals rather than trying to guess at which direction the stock will move. Time is money and complex TA is a waste of time.
TA is Astrology
I'll admit that TA might have a more solid footing than astrology. But not by much. And the two seem function remarkably similiarly.
Humans are very good at convincing themselves things work sometimes, even when there's no real rationale as to why it would and basic logic would suggest that it doesn't. You can't accurately predict the future of a stock any more by reading a chart than you can predict the future of your own life from a deck of tarot cards. If you could, you'd be a billionaire by now.
If a "model" beats the market 60% of the time, what's to suggest that the practitioner didn't throw some fundamental analysis or plain ole' common sense into the analysis? How could one show that the extra 10% was solely due to the success of the TA. If TA really works, shouldn't there be someone who has near 90% accuracy with it?
Ok, I realize that there aren't fundamental-based investors with 90% accuracy either, but once again, this goes back to the foundation. You *know* the fundamentals provide a look at the foundation of a company with a reasonable level of certainty. There is of course, fraud, but even that can be better predicted by finding "red flags" in the financial statements. How do you know with any level of certainty that (a) charts provide realistic models of human behavior, (b) the charts themselves aren't distorted by poor fundamental analysis by investors, and (c) complex human behavior can be predicted by mathematical analysis of the charts?
With TA, sometimes people win over the short-term, but over the long-term, TA does not allow one to consistently outperform. Instead, practitioners simply convince themselves that their readings were correct (even if it takes weeks for them to play out) and they gloss over the failures. In that sense, TA is very much like astrology. The best practitioners of TA happen to be the most ambiguous.
Attack Away
For whatever reason, I've noticed that TA'ers seem to be more emotional about criticisms of their models. This has actually solidified my belief that it is basically astrology. If someone criticizes my analysis of a stock respectfully, I don't get offended. Oftentimes, I'm happy they gave me additional perspective. So based on my past indictators of reactions of TA'ers, I fully expect to be attacked here. So attack away. ;)
RSS Headlines
Fool UK