Saturday, June 29, 2002 1:28:22 PM
Shotgun approach is fine.
There is nothing wrong with the shotgun approach to optimizing as there isn't really a standard method.
I personally eyeball the chart and try to place the bands at the very top and bottom of the extreme price moves to start. This is preferable it the stock has exhibited a fairly even sideways trend in the timeframe I am backtesting. If the stock is showing a classic downtrend (ie, brcm, emc, brcd, and just about every other friggin tech stock...gg), I'll lower the top band so that it "sells the rally". My guess is going forward, the former method (bands at extreme highs and lows) will be preferable as I think the markets are within reach of setting their lows. May take another month or 2 (or more) but I have to believe the werst is behind us. Just one guys opinion though, so take it for what it's worth. I have been known to be an optimyst...g.
Next I'll shuffle the buy and sell deviation multiple's. Again, the general rule of thumb is a higher sell multiple if the stock is in a downtrend (to sell the rally faster) or a higher buy multiple in a strong uptrend (to buy the dips faster). The multiples will be close to even with a stock that exhibits a sideways rolling trend. Adjustments to the multiples will also effect your risk ratio. The higher the numbers the faster you will sell all shares and go to cash, or the faster you will burn all your cash and get fully vested.
You can almost tune it to the point where very few multiple purchases or sells are initiated and instead you are almost getting an "all-or-nothing" method. There is nothing wrong with this as long as you understand the risk profile of your set up. As stated many times, I developed X_DEV to be as flexible as possible and suit as many different trading styles as possible.
During the optimization I will adjust the minimum buy or sell shares at any time. I personally like to trade a minimum dollar amount of anywhere from $600 to $1000 or more on partial positions, so I'll try to set the minimum shares accordingly. If I find that I could achieve greater performance from fewer or more shares, I'll weigh it out and decide which setting is best. Almost always, I'll put more weight towards X_DEV's signals than mine, and accept it's settings.
After those 3 groups of settings are complete, I'll try a 50%, 66% or 75% Initial Purchase percentage to see which would have led to a greater return. I'll weigh this result against the current market or stock conditions and decide accordingly. For example, on my recent Paypal play I waded in with only a 30% Initial Purchase due to the risk profile of the market and the stock at the time of purchase. I think this setting may be the least critical, as over a longer term time frame X_DEV will balance it all out anyway. The only time it would make a noticeable difference is if you were lucky or unlucky enough to catch the absolute bottom or top and bought in with a Initial Purchase that was very low or very high. That's why it's usually best to chose either the 50, 66 or 75% to start.
Once I have all the rough settings complete, I'll fine tune the bands in 1/2 percentage amounts. I feel anything more than that may be "over-optimizing". The whole purpose of the optimization is just to get a good feel for the natural current rythm of the play so X_DEV can calculate the next best likely targets for a buy or a sell.....and then continue to do so to provide you with a great return over time.
You will find this is X_DEV greatest advantage over other systems that just tell you what to do and allow no flexibility otherwise. You will always be smarter than a "black box" system if you take even a small amount of time to monitor your investments. The human brain is so far ahead of any computer intelligence, that to suggest otherwise is foolish. X_DEV helps the user make better decisions, and doesn't force them upon you. Perhaps much of my personal trading philosophy has been integrated into X__DEV during it's construction. I've have never followed anyone's advice during my entire trading career. I've always just studied the movement of the stock, tried to figure out a recurring pattern, and then tried to take advantage of that pattern. Anyone else's individual opinion never seem to be very important to me. I hope that doesn't seem arrogant, but it has always seemed to work for me.
X_DEV's objectiveness mirror's my own. It finds the pattern that has yielded the best return and then attempts to capitalize on that pattern going forward. If the pattern changes than X_DEV can be adjusted to to keep performing appropriately. I analyze the major markets every day looking for pattern breaks. As soon as one occurs I will consider re-adjusting X_DEV's settings on my individual plays.
Before I forget, the only other thing I will do to optimize a play is try and figure which time frame is best to test. Sometimes a longer term chart is better, sometimes a very short term is better. New user captainjonny just asked about BRCM settings. I would personally use a very short term time frame to judge going forward until a major break in the market occurred. BRCM set up below:
Hope this helps. Any questions, let me know.
~Myst~
There is nothing wrong with the shotgun approach to optimizing as there isn't really a standard method.
I personally eyeball the chart and try to place the bands at the very top and bottom of the extreme price moves to start. This is preferable it the stock has exhibited a fairly even sideways trend in the timeframe I am backtesting. If the stock is showing a classic downtrend (ie, brcm, emc, brcd, and just about every other friggin tech stock...gg), I'll lower the top band so that it "sells the rally". My guess is going forward, the former method (bands at extreme highs and lows) will be preferable as I think the markets are within reach of setting their lows. May take another month or 2 (or more) but I have to believe the werst is behind us. Just one guys opinion though, so take it for what it's worth. I have been known to be an optimyst...g.
Next I'll shuffle the buy and sell deviation multiple's. Again, the general rule of thumb is a higher sell multiple if the stock is in a downtrend (to sell the rally faster) or a higher buy multiple in a strong uptrend (to buy the dips faster). The multiples will be close to even with a stock that exhibits a sideways rolling trend. Adjustments to the multiples will also effect your risk ratio. The higher the numbers the faster you will sell all shares and go to cash, or the faster you will burn all your cash and get fully vested.
You can almost tune it to the point where very few multiple purchases or sells are initiated and instead you are almost getting an "all-or-nothing" method. There is nothing wrong with this as long as you understand the risk profile of your set up. As stated many times, I developed X_DEV to be as flexible as possible and suit as many different trading styles as possible.
During the optimization I will adjust the minimum buy or sell shares at any time. I personally like to trade a minimum dollar amount of anywhere from $600 to $1000 or more on partial positions, so I'll try to set the minimum shares accordingly. If I find that I could achieve greater performance from fewer or more shares, I'll weigh it out and decide which setting is best. Almost always, I'll put more weight towards X_DEV's signals than mine, and accept it's settings.
After those 3 groups of settings are complete, I'll try a 50%, 66% or 75% Initial Purchase percentage to see which would have led to a greater return. I'll weigh this result against the current market or stock conditions and decide accordingly. For example, on my recent Paypal play I waded in with only a 30% Initial Purchase due to the risk profile of the market and the stock at the time of purchase. I think this setting may be the least critical, as over a longer term time frame X_DEV will balance it all out anyway. The only time it would make a noticeable difference is if you were lucky or unlucky enough to catch the absolute bottom or top and bought in with a Initial Purchase that was very low or very high. That's why it's usually best to chose either the 50, 66 or 75% to start.
Once I have all the rough settings complete, I'll fine tune the bands in 1/2 percentage amounts. I feel anything more than that may be "over-optimizing". The whole purpose of the optimization is just to get a good feel for the natural current rythm of the play so X_DEV can calculate the next best likely targets for a buy or a sell.....and then continue to do so to provide you with a great return over time.
You will find this is X_DEV greatest advantage over other systems that just tell you what to do and allow no flexibility otherwise. You will always be smarter than a "black box" system if you take even a small amount of time to monitor your investments. The human brain is so far ahead of any computer intelligence, that to suggest otherwise is foolish. X_DEV helps the user make better decisions, and doesn't force them upon you. Perhaps much of my personal trading philosophy has been integrated into X__DEV during it's construction. I've have never followed anyone's advice during my entire trading career. I've always just studied the movement of the stock, tried to figure out a recurring pattern, and then tried to take advantage of that pattern. Anyone else's individual opinion never seem to be very important to me. I hope that doesn't seem arrogant, but it has always seemed to work for me.
X_DEV's objectiveness mirror's my own. It finds the pattern that has yielded the best return and then attempts to capitalize on that pattern going forward. If the pattern changes than X_DEV can be adjusted to to keep performing appropriately. I analyze the major markets every day looking for pattern breaks. As soon as one occurs I will consider re-adjusting X_DEV's settings on my individual plays.
Before I forget, the only other thing I will do to optimize a play is try and figure which time frame is best to test. Sometimes a longer term chart is better, sometimes a very short term is better. New user captainjonny just asked about BRCM settings. I would personally use a very short term time frame to judge going forward until a major break in the market occurred. BRCM set up below:
Hope this helps. Any questions, let me know.
~Myst~
~Mystic One~
http://stockwerld.com
(What matters)
http://www.investorshub.com/boards/board.asp?board_id=2543
Join the InvestorsHub Community
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
