Creation of a High Performance Stock Market ModelThomas Nogales Financial's Market Alert |
|
WWW.ThomasNogales.COMPeople have searched for the key to financial profits ever since they first began trading with each other. The question is: what special knowledge will give me an edge over others in my chosen market. Today, economists, stock analysts, and just regular investors follow technical charts and economic releases trying to anticipate the market's next move. There are indeed correlations but translating that into a working predictive model of stock market prices is really tough. I call the process "financial science" because, like a scientist, the model builder must examine the data, back test it, and create real-time experiments that confirm the theory with enough iterations so that professional peers are convinced of its validity. This scientific process and the required peer review works great in the hard sciences but is difficult for the financial model builder. Trouble is, you can't easily share a good idea because everyone else will start to use it. For example, the Dogs of the Dow model bought Dow stocks with high dividend yields and regularly beat the index until a book explained the technique and a multitude of people started using it. A successful timing model is rare and the number of users must be kept limited. Otherwise, its returns in the market place will decline. So-called experts who continuously and erroneously proclaim that market timing doesn't work would be far more accurate to say that a true winning method probably won't work forever. The TNF Market Alert model was created in 1999 based on prior observation and experimentation. After many failed attempts and dead ends, the idea for the model literally flashed into being in a financial epiphany. It combined unrelated but separately known ideas in a non-obvious way. My first thought was "that can't possibly work" but in only two hours the core concept was computer tested and validated. I was vigilant about not creating a biased experiment that fits the past data but won't work well in the future. The first draft beat buy and hold by well over 2:1 on a 20 year back test. That was especially impressive because buy and hold had the advantage of a roaring bull market. The model outperformed by selling as the bull stumbled and then quickly buying back in. A white paper was written and given to a mathematician for critical analysis. He reviewed the data, the logic, and the results and pronounced the concept "brilliant". Over the next several months I added some tweaks to improve performance and reduce risk. Then came the hard part - investing money based solely on the model and ignoring the market mood. It worked. As the stock market collapsed in 2001 the model caught opportunity on the way down and mechanically made profits in a seemingly impossible bear market. I've been asked if the model uses "technical analysis" and the answer is no. I've tested too many other models to have much faith in subjective methods. Technical tools may have value in the very short term but projecting historical trends into the future often causes dramatic failures that devastate the investor. Risk is everything and must be carefully controlled over the entire investment timeline. For that reason, it's critical for investors to allocate their money over various assets classes and not to bet everything on my models. I encourage users of this site to work with a personal financial advisor who exhibits skill and integrity and to utilize index funds or very low cost mutual funds. Properly used within an asset allocation program, I believe the Market Alert timing model can greatly enhance returns by signaling periods of danger and opportunity. This web site is a form of peer review with the site visitors being evaluators. I use MARKET ALERT for my personal investing and will share it's signals with you on this site. Thomas Nogales Financial, LLC |