An Administrative Item
I added some Dakota users to this list that weren't on it.  If you don't want this newsletter coming to this email address, please unsubscribe and resubscribe the correct email using the newsletter box on our homepage.  If you missed the last issue, you can find this newsletter's index at: http://www.biocompsystems.com/newsletters/Bits/

In Review
In the last issue, I discussed BioComp Dakota and did some comparisons between Profit and Dakota, our two financial markets packages.  They differ and compliment each other.  People have been asking us if we would create a Profit/Dakota "bundle" for sale, so we will likely do that.  Also in the last issue, I talked about using an Advance/Decline Pessimist for intraday and end-of-day trading.  In that discussion I asked you to email me if you were interested in trend detection.  Well, I certainly got the emails!  So, I've covered more on that subject in this edition.

Dakota Price Increase
Dakota's price "bumped up" with the release of version 1.2 today, in line with Dakota user recommendations.  Version 1.2 supports unlimited variables to go to bots, enabling a lot of power, including cross-market analysis.  We also made enhancements to the ScriptBot editor.  You can avoid the price increase if you order by January 31st and when placing the order, you put "Newsletter" in the order's special instructions field or you send an email to sales@biocompsystems.com.  Orders can be placed on-line at:
http://www.biocompsystems.com/store/financial/index.htm

BioComp Dakota
BioComp Dakota continues on its successful journey.  In January, the Dakota user community grew 15% and has now spread across 16 countries.  We are approaching a quorum of users sufficient to have both a European and American user group meeting.  The European meeting will likely be in Amsterdam within a few months since Amsterdam is centrally located for that Dakota community, it's relatively affordable and it's easy to get downtown from Schiphol airport (AMS).  There's a lot to do in Amsterdam as well, including some great museums.  If you are a European Dakota user and would like to participate, send me an email at cmcook@biocompsystems.com.

Earlier this month we released Dakota version 1.1 and now this week version 1.2.  Version 1.1 gave us ScriptBots, the ability to write adaptive trading bots within Dakota itself using Visual Basic and version 1.2 gives us the ability to feed our bots unlimited variables enabling a lot of power, including cross-market analysis.

We also enhanced ScriptBot editor and Dakota's web presence including:

• Features List
• Videos Yes, you CAN run Dakota in as few as 4 mouse-clicks!  Watch it done.
• Investor Benefits

With these enhancements we have increased Dakota's prices by another 5%...

• Standard                  $549
  Get version 1, pay for upgrades to versions 2, 3, etc. as you go, if you choose.
   
• Product Sponsor      $1,320
  Receive all Dakota product updates for the life of the product
  Receive all future BioComp-authored bots at no charge

Most Profit users recognize that the Sponsor level is a good deal.  So good that we may discontinue it. 

Avoid the price increase by ordering on or before January 31st and putting "Newsletter" in the order's special instructions field or by sending an email to sales@biocompsystems.com.  Orders can be placed on-line at:
http://www.biocompsystems.com/store/financial/index.htm

P.S.  If you are interested in the Sponsor level, I suggest you do so soon.

Back to top

Trend Detection vs. Pessimism
Last time I said I was going to put a trend detector in my Advance/Decline Pessimist bot and will have the bot switch logic... pessimist during sideways trading, follow the trend during trends. Quite a few of you asked me to write more.  After some thought, I realized why... because trend detection is a critical aspect to trading.

Why is trend detection so difficult, especially when you look right at the screen and see it.  Well, two reasons first come to mind... First, your human eye has the advantage of looking ahead.  If you drag a piece of paper across the screen, covering the future, then you realize how hard it is.  Second, trends have what I consider a "fractal" nature: it's defined on multiple time frames.  What is a trend to one person isn't to another, because one person might be looking over a week's worth of data, while another might be looking at months or years.  The first thing you have to do in detecting a trend is to define what a trend means to you in the particular case you are going to use it...  3 bars in a row up or down?  Average of 15 bars up or down?  It varies by person and application.

There's a number of ways to quantify a trend, including some very common ones: moving average crossovers, momentums, regression slopes, etc.  However, the shorter the timeframe, the less effective these are.  Why?  Because it takes some number of bars to quantify the trend and what if it ends just as you do?  Seems it does all too often [wink].

In my Advance/Decline Pessimist bot, I cheated.  I defined another stop-gap rule around my logic.  My earlier logic was:

    if AdvanceFraction < 0.33 then
      Signal = 1  'Things are bad, so go LONG !
    elseif AdvanceFraction > 0.66 then
      Signal = -1 'Things are too good, so go SHORT !
    end if

and I changed it to this:

if AdvanceFraction <= 0.1 then
  Signal = -1
elseif AdvanceFraction > 0.1 and AdvanceFraction <= 0.33 then
  Signal = 1 'Things are bad, so go LONG !
elseif AdvanceFraction > 0.66 and AdvanceFraction <= 0.9 then
  Signal = -1 'Things are too good, so go SHORT !
elseif AdvanceFraction > 0.9 then
  Signal = 1
end if

BioComp Profit
A good tip for BioComp Profit users is to build systems that contain models from multiple time-frames and allow the equity-based voting scheme to switch them in and out of the system model.  Each model may be tuned into equity performance for a certain period and the equity-based voting discards ones presently not performing well.  This provides an adaptive playing of models for enhanced returns.  For example, on one S&P500 futures system I have 50 models built over numerous time periods.  If I take the average of my signals I get $126,000 in equity gains since Nov 26, 2001 (the date the system was created), but if I use equity-based voting, I get $133,525 in profits and a smoother equity curve.

Closure
If you are interested in getting involved with Dakota, remember that you can avoid the price increase if you order by January 31st and when placing the order, you put "Newsletter" in the order's special instructions field or you send an email to sales@biocompsystems.com.  Orders can be placed on-line at: http://www.biocompsystems.com/store/financial/index.htm

If you find this newsletter interesting, please stay subscribed.  If you think others might like getting it, forward this email to them or post a link to our website in your favorite trading discussion group.  That helps us grow.  We'd appreciate it.  If you have suggestions, please email me at cmcook@biocompsystems.com.

These documents are provided for informational purposes only. The information contained in this document represents the current view of BioComp Systems on the material discussed as of the date of publication.  Materials written should not be interpreted to be a commitment on the part of BioComp and BioComp cannot guarantee the accuracy of any information presented after the date of publication.   INFORMATION PROVIDED IN THIS DOCUMENT IS PROVIDED "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND FREEDOM FROM INFRINGEMENT. Statements of equity performance are hypothetical and have not been substantiated by records of actual trading.  Hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading.  Also, since the trades have not actually been executed, the results may have under- or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity.  Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to those mentioned.

The user assumes the entire risk as to the accuracy and the use of this document. This document may be copied and distributed subject to the following conditions: 1) The entire document must be copied without modification; 2) All copies must contain BioComp's copyright notice and any other notices provided therein; and 3) This document may not be distributed for profit.  All trademarks acknowledged.  (C) Copyright BioComp Systems, Inc. 2006.