Options

Search

Subscribe to Thread

Bookmark Thread

First Page First Unread Last Page Last Post

Printable Version

Similar Threads

Psychology is nothing 195 replies

Investor Psychology 1 reply

Psychology in trading 26 replies

Trading Psychology is BS 52 replies

Trader Psychology 3 replies

changing psychology

I've been developing a trading strategy based on trends and news announcment since May of 2006. I've come up with a REALLY good strategy. It is in a constant development still but I keep improving it and theoretically I could make over 1000 pips certain months. Of course theory and practice rarely match up. But here is what bothers me. I tested the strategy all the way back to January 2006 and everything lays out perfectly. I rarely lose. On average maybe 1.5 trades. My loses are limited to 60 pips. I let my winners run. I trade GBP, CHF and EUR vs USD. I have already funded my live account and ready to trade live. But a few days ago I decided to test it abit further just to make sure. I went back all the way to July of 2005. And I noticed that my strategy pretty much fails every month. I had to augment it quite abit to make theoretical profit in a month that I could have made in a week in 2006. How could this be? How could one strategy work so well for 11 months but then just stop?
Well, if you "curve-fitted" or "over-optimized" your strategy, then yes it is very possible it could only work for that time period. There was a great post here about developing a system and proper backtesting. If someone could find the link I would appreciate it.

andersenws
"Youth is the trustee of prosperity." - Benjamin Disraeli
Fooled by Randomness could that be the reason?


Lou
Quoting andersenws
Well, if you "curve-fitted" or "over-optimized" your strategy, then yes it is very possible it could only work for that time period. There was a great post here about developing a system and proper backtesting. If someone could find the link I would appreciate it.

andersenws

Yeah I always thought that would be a huge problem if my system was based on pure technical analysis. I do involve it abit with SR and trends but most of how I'd earn is though the price action. What ends up happening for me before 2006 is that I break even on 90% of the trades I make. And the funny thing is, there is no transitional period. I always thought that a change in the market would just slowly fail the system more and more often until you optimize it again. Especially true with mechanical systems. But by using the price action you shouldn't need to do any big adjuctments. Yet, the conditions in 2005 are so different. Has anyone else found this?
Quoting Lou
Fooled by Randomness could that be the reason?


Lou
Just finished that book! Awesome read.

andersenws
"Youth is the trustee of prosperity." - Benjamin Disraeli
the reason is that market dynamics change, its just change from being stuck in a range for the last six months


have a look at what sort of market your in and make it a trading factor, hope this helps
This isn't necessarily true. Imagine, for example, that you've developed an incredible strategy for a ranging market. How does a market break out of an extended range? With powerful moves. Look at what happened Thanksgiving week to the EURUSD and GBPUSD. Hundreds of pips while the market wasn't even looking and a big breakout of an extended range. Market dynamics can truly change very quickly.

I'd suggest you go figure out why it's so successful during that time period. Then, figure out if there's something that can indicate you to those conditions and trade the strategy when you get that indication.

Quoting mmnoname
Yeah I always thought that would be a huge problem if my system was based on pure technical analysis. I do involve it abit with SR and trends but most of how I'd earn is though the price action. What ends up happening for me before 2006 is that I break even on 90% of the trades I make. And the funny thing is, there is no transitional period. I always thought that a change in the market would just slowly fail the system more and more often until you optimize it again. Especially true with mechanical systems. But by using the price action you shouldn't need to do any big adjuctments. Yet, the conditions in 2005 are so different. Has anyone else found this?
Thread Tools Search this Thread
Search this Thread:

Advanced Search

0 traders viewing now