The San Francisco Giants and the Hindenburg share a common trait - they both can be the basis of creative data mining that leads to erroneous investment thinking.
As I watched the Giants win their first World Series since moving to San Francisco in 1958, my mind quickly wondered how the stock market performed in the years following a National League winner vs. the years following an American League winner. I did some research after the game and found that in the years following a National League winner the S&P 500 returned an average of 5.17 percent and only 3.10 percent when an American League team wins the World Series. Since the Giants won, should we borrow as much as the stingy banks are willing to lend us and buy as much stock as we can on margin or using call options? Not so fast.
This is an example of something called "data mining." Data mining is where seemingly smart people armed with gobs of computing power and too much time on their hands sift through historical data to find patterns which they use as the basis for their investment strategy. Also referred to as "backtesting" these strategies fit the historical data well but do nothing to predict the future.
Does it work? Only until the pattern is broken. The pattern can break when there isn't a "causal relationship" or a legitimate cause and effect between the data. Does anyone truly believe that stock market performance is affected by who wins the World Series? Of course not - because there is no causal relationship between the Boys of Summer and the stock market.
Little realities such as this don't slow down the onslaught of marketing hype for new investment products. In addition to data mining, two other factors are used to market investment hype - experts and exposure.
Oftentimes data mining is conducted under the direction of someone with impressive credentials, making the investment appear more credible. We've all heard someone say "It doesn't take a rocket scientist to...[fill in with your favorite rocket scientist saying]." In the investment business I've actually met two people who were rocket scientists (aerospace engineers) in their earlier careers. Both men are geniuses.
Rocket Scientist #1 started an investment company after developing his own secret algorithm which is the basis of his investment strategy. He utilizes a neural network to synthesize all of his data into a model which tells him when to trade in and out of various asset classes. He trades frequently throughout the year. Rocket Scientist #2 utilizes the Fama-French Three Factor Model, based on diversification and minimal trading, as the foundation for his investment strategy. Guess which Rocket Scientist has better investment returns, incurs lower trading costs and generates less taxes? Rocket Scientist #2 handily beats Rocket Scientist #1, neural network and all. Being a genius with a complex computing network can't make up for a bad strategy. This is the reason we utilize the Fama-French Three Factor model at Willow Ridge Capital Advisors.
The final factor - exposure - is what gets the word out. If you can get the media to pick up on your idea, it's almost sure to attract those who are looking for the right opportunity to make a killing in the stock market - a quick buck. They often give their investment strategy a special name to make it sound, well - special.
The Hindenburg Omen, highly publicized in August 2010 gives us a high-profile example of how data mining, supposed experts and press exposure work together to create hype. Jim Miekka, a former physics teacher (implying expertise and credibility), researched historical data and determined that after certain technical criteria are met, triggering the Omen, there is a 77 percent probability of a downward market movement greater than 5 percent within the next forty days, a 41 percent chance of a panic sellout and a 24 percent chance of a major stock market crash. On August 14th The Wall Street Journal published an article "Hindenburg Omen Flashes" and then followed up with articles on the 17th, 18th, 20th, 23rd, and 26th.
How did the stock market perform after the horrendous Hindenburg Omen was triggered on August 12th? Through November 1 the S&P 500 has gained 8.7 percent. If you sold out of the stock market in August based on this great sounding story, you lost big-time. As proven many times over the years, market timing is incredibly tricky and rarely successful. When market timing is successful, it's impossible to determine if it's due to luck or skill.
The next time you hear a new investment strategy, created by a genius and publicized in a credible media source - be careful. It just might be another Hindenburg Omen. If you want a credible investment strategy that simply works, take a look at the Fama-French Three Factor Model, or give us a call to discuss it.
As for the World Series- Go Giants!