Greetings from Baker City, Oregon! I'm currently traveling between Idaho and Oregon for work but manage to keep an eye on my stock positions in the evenings, provided the hotel I'm staying at has Internet!
My positions in WIT and GOLD are steaming higher, especially GOLD. In the evenings I just check my trend models and update my stops as required. So far things have worked out pretty nicely and I'm looking for another set of wins in WIT and GOLD when my models tell me to SELL.
Trend following systems work great provided you have sound risk management and a good idea when to get OUT. You can set it, go about your daily business, and then update the model and stops as required.
May 2009 Archives
Sorry guys but I’ve been busy with work and doing some light stock trading. My first live trade in YZC, using my Neuro Classification system was a success. I nabbed over 7% in that trade over April. I’m currently long WIT using the same system and sitting on a small gain.
On the downside, I’m finally capitulating on my C trade. I’m selling my losing positions after sitting on a -80% loss. This will probably mark the bottom in equities. LOL.
You can follow my twitter stream to see what I’m doing in the markets daily. Twitter is killing my blog.
Paul Wilmott, financial mathematical modeler and all around quant genius, was recently interviewed about how banks relied to much on their (wrong) models to value their assets and how it led to their undoing. Of particular interest is how he says they overlooked the human side of the model equation.
"You can model electromagnetic waves: a mathematical model that shows molecules of air moving around a plane, making it fly," he says. "But in a financial model, you need more than numbers. The models in finance are not very good. In this field, it matters if you're not psychologically synchronised; people don't behave rationally. You can't rely on people following equations. It's half maths and half human."Â [by Elena Moya]The moral of this story is that fear and greed will make you do stupid things.
I'm recently came across some Python Code snippets for Machine Learning maintained by Stephen Marsland. I'm not a programmer or coder but I know several of my readers are. Follow the link and enjoy.
