It was the steamed glass shower door that was an ideal canvas for drawing up trading ideas. I began looking at stock market data when I was working on my MBA at the University of Michigan and trying to make the connection between company financials and its stock performance. My Aerospace Engineering degree from the University of Florida that gave me the wide range of analytical skills I needed to pick apart the stock market. Give a Rocket Scientist an MBA and this is what you get.
I was on the hunt for a way to predict stock price movements. Everyone was giving out stock tips in the late '90's and there had to be a way to quantifiably figure out the stock market. I've purchased many stock systems, both long and short-term. I never really found anything that could work while still allowing me to work on a "day job" (I worked a Ford Motor Company at the time) or have a life. It did seem like you could pick up just about any "tech" stock at the time and it would just continue to rise. This was back before E-Trade was a web-service and we had to trade stocks through America Online using keyword "etrade". Fun times!
I began a quest to figure out some type of a stock trading system that could generate a consistent positive return over a relatively short term, but not be a day-trading system. This became a passion of mine and I toiled away to develop a system that would work in a bull, bear, or flat market. I surfed the internet and downloaded stock market history at 56k to find my answers with my Intel 486 Laptop and Windows 95.
More than a decade later, after reviewing many systems, reading countless books and actively trading both with real money and on paper, I have finally have a system that works! I had to invent a new indicator to do it, but it has shown a pretty consistent return.
Here's where the shower door came in.
In the late '90's, I watched countless stocks rise very quickly and get to a point where they were unable to hold the momentum and then quickly turn around and fall as fast as housing prices in South Florida; usually after a big story on CNBC. I wanted to find a way to short a stock at the peak of the movement and then ride it down. I was drawing the stock movements on the shower door when I pieced together the idea for calculating how normal or abnormal the stock move was. What if you could calculate the standard deviation of the move of the stock and when it gets to a certain point, short it and watch it fall.
It was about this same time that I learned the trick of a signal and a trigger. Many stock systems were looking for a pattern in the stock which would be a signal, and then looked for a further confirmation in a specific price move that would trigger the purchase or short sale of a stock.
So I calculated the standard deviation of the change in the stock's price over a certain period of time and found the value or the "sigma" when the peaks occurred. Theoretically, these peaks shouldn't happen very often. Stocks that showed a three, four, or five sigma move should be an abnormality, like twenty-somethings convincing seasond VC's that you can monetize a website with eyeballs. When these moves occurred, I had the system issue a signal to watch for weakness in the stock price. When the weakness in price occurred, then the system would short the stock or buy a put option. For high-priced stocks with high volume, this system worked beautifully. What was interesting about this was that lower priced stocks had the tendency to continue their abnormal rise after it reached the three or four sigma gain.
These two systems were launched in the form of StockMachines.com in 1999. StockMachines ran the system for a few years and eventually morphed into FourSigmaTrading.
FourSigmaTrading now searches every stock in the market for specific trading patterns involving its sigma and issues signals when it finds them. If the stocks it signals continue their meteoric rise the next day, then issues a buy signal. The system holds the stocks for a certain period of time and then sells it.
We're going to continue to talk about "sigma" here as well as the trading systems derived from it and the stocks it picks. We're going to win some, but we're also going to lose some. We hope the winners will outpace the losers when it's all said and done. The system currently runs at a little over a 60% win rate with the winners making about twice as much as the losers. Let's see if it holds true in the future.
No comments:
Post a Comment