It is likely no surprise to most people that have studied Warren Buffett (Trades, Portfolio) that he had used technical analysis as an investment strategy for nearly eight years. Luckily for Buffett, he met Ben Graham, studied his writings, studied under him at Columbia and solely committed to the road of fundamental value investing for the rest of his life. The rest is history.
After Buffett came to the realization that technical analysis did not work, he explained years later, "I realized that technical analysis didn't work when I turned the chart upside down and didn't get a different answer."
There are many different indicators available for use by a technician, all of which are based on various weightings and combinations of price and volume. The technician believes price and volume reveal the ability to infer what is likely to happen in the future through trends and indicators.
The price and the volume is from the past and like the ancient saying about performance, it is not indicative of the future. There is a whole host of errors that charlatans (I mean technicians), are susceptible to when analyzing a chart. Two of the largest biases are anchoring and data mining.
Anchoring: Anchoring is a cognitive bias that causes one to become attached to a piece of data, usually the first piece received or a minimum/maximum.
Think when you view a chart showing past price action. If you are like me, after viewing a one-, five-, 10- or 20-year history of the price, I immediately become anchored to the all-time highs and lows of the chart. I begin to think about mean reversion and come to false conclusions about the probability of the future based on past price. This is a bias one must work very hard to overcome and a bias that one can use to exploit, always using a larger number as the first offer when selling and inversely a low number when buying.
Data mining: Give 10 different technicians the same exact chart with the same indicators and averages. They will likely all come to different conclusions because of “top-down processing” or previous experiences, judgments, prejudices and beliefs. The chartist sees what he or she wants to see and infers a narrative that is plausible at the time. The indicators or strategies that are right or "last" are simply a product of survivorship bias and a self-fulfilling forecast. [Here is one from Credit Suisse]
Hindsight bias and the narrative fallacy: This also causes one to believe patterns in the past were “signs” or were easily identifiable at the time. That are other biases one can identify or ask a proclaimed technician to identify (there are a lot of them), the point is the list (of biases and technicians) is endless.
These biases by no means are limited to technicians, but everyone is susceptible to them, especially when dealing with data. Although a major problem arises for the technician that does not for the fundamentalist, this is confidence and a tangible value that is assigned to a business.
The fundamentalist uses revenue, margins, cash flow, earnings, capital structure, operating leverage, country of origin, assets, liabilities, customers, etc., to value a business based on intrinsic value. The value investor then differentiates themself by buying at a margin of safety from the estimated intrinsic value.
The chartist values the business on none of these operating characteristics but only the past price and past volume of the market price. This is a flawed paradox, where a TA is willing to buy an asset under the assumption he or she will receive more in the future based on a given indicator or trend. The paradox is that these past trends and indicators are based on the performance of others, therefore, the crowd is the hand that is guiding your decision.
There are a few studies regarding the efficacy of technical analysis that are at best inconclusive. The point is technical analysis has no statistical or legitimate empirical evidence of adding value in the form of an excess return, unlike value investing, proven by multiple studies, empirical evidence and the Nobel laureate Eugene Fama.
What if it did work or there was a predicative TA strategy?
Some technicians will tell you about which indicators work for them and how you can exploit them to make money (for a subscription fee). What they won’t tell you is, if there was a definitive indicator or trend one could identify and exploit for an excess return.
The market would discount it.
Some say when a golden cross occurs it is predictive of additional gains, a death cross infers bearish times are ahead, a shooting star reveals a short term top is in, and you get the point.
Inverting, one has to ask would the market not discount a definitive pattern, trend, or indicator that can be identified as other speculators would also be aware of the signal? But what about the argument of the self-fulfilling prophecy when most market participants believe in an indicator? Sure it can happen and does, but like Ben Graham said famously many years ago, “In the short term the market is a voting machine, in the long-term it is a weighing machine.” The problem lies in predicting the votes.
In the end, a business is worth the present value of all future cash flows discounted at an appropriate rate.
The following study reveals no value is added from technical analysis.
[Study from Massey University of Technical Analysis on 5000 TAs]
But There Are Rich Technicians…
A common dispute between fundamentalists and technicians is that if technical analysis doesn’t work, why are there rich technical analysts like Paul Tudor Jones (Trades, Portfolio)?
Using a coin-flipping contest and basic probability distribution we realize this is pure chance or luck. If 300 million people all started flipping a coin today, after the first flip we would be left with 150 million winners and 150 million losers.
After the second flip we are down to 75 million winners. We carry on the exercise for 21 flips (total) before we are left with 286 winners.
Now imagine each flip as a year in the market. Each player that flipped tails won and received a 100% return on their cumulative investment. Further imagine the initial investment of all coin flippers was $1,000.
The remaining 286 winners would all have roughly $2.1 billion, rich by any monetary standard today.
The question is not if there are rich proclaimed technicians, but how many consistently outperform the market relative to peers using the same or other strategies.
Of those 286 winner who all flipped tails imagine 26 as being technicians, 80 from the growth category and 180 from Graham and Doddsville or identify as value investors.
Which strategy would you choose?
Unlike technical analysis, fundamental analysis has a proven track record with a disproportionate distribution of successful investors and backed by empirical evidence as well as statistical evidence.
For the still skeptical, try Chart Game and play for 20 to 30 minutes. Write down how well you do each time. You are likely to come to the conclusion using price/volume is a heads and tails game. You may get lucky, you may get unusually unlucky, but over a large sample the empirical evidence will appear.
The best way to make money using technical analysis is to sell it. Sell a subscription, sell a service, sell lessons, write books and collect from others. It is sold to the masses by daily commercials, endorsements, brokers, pundits, banks and entrepreneurs. Most have an incentive to do so, as your commission (or subscription fee) generates their profits.
“Never ask the barber if you need a haircut.”
About the author:
"When you find yourself on the side of the majority, it is time to pause and reflect." - Mark Twain