|
ULTRA 11 Feature List
Scheduled Release:
01/31/10 System/Data
Organization The first thing I want to say is that all the changes I discuss below have been developed to be compatible with what you're currently doing with ULTRA. So please do not worry about that issue. Even though attaining this goal has proved to be challenging, after being completely ripped-off by Microsoft with the 64-bit Vista compatibility issue, the last thing I'd ever do is abandon what my clients are currently doing and force them to do something "better". One of the biggest weaknesses of ULTRA 10 was simply too much information. The fact that ULTRA has "130+ Systems" has become a significant negative for the program in terms of marketing, learning curve, and usage. The short answer to this problem begins below at "The problem has been solved...". Skip to there for the bottomline. Here's the long answer: As a serious investor but also a researcher, I have great interest in how systems actually perform in real-time, good or bad. Therefore, as a personal "lifelong" research project I want systems that have performed poorly to remain in ULTRA. I also refuse to participate in misleading investors by trying to make ULTRA appear to be the "holy grail" of investing, which would be very easy if I abandoned the poor performers and only left the best performing systems in the product. Keep in mind that every system in ULTRA made at least two significant cuts. First, someone thought it was worthy of being published somewhere. Second, I reviewed the formula, took the time to make sure the historical/real-time results could be replicated, and I determined the system was worthy of inclusion in ULTRA. Lastly, I've personally used at least 80% of the systems in ULTRA with significant amounts of my capital over the years. I've personally been pretty good at knowing when to "give up" on a system. But I've also been pretty lucky too as the end of the great 90's bull market aligned well with my personal situation. Basic personal finance tenets clearly indicated that I'd be wise to become a conservative investor. Bottomline is that in my opinion, being able to see how these systems actually performed in real-time since inclusion in ULTRA is very valuable information. Among many things, It reveals clues as to the correct level of expectation an investor should have with any investment strategy. Having said all that, one of the biggest challenges with ULTRA 11 was solving the problem of how to organize systems and data in a way to make the program most friendly to all types of users from the guy/gal who just wants to follow a few weekly systems with their IRA accounts, all the way to the Economics professor (of which I'm blessed to count many as clients), using ULTRA for research projects. Trust me, I've pondered this challenge for many years and frankly it was always just an "exercise in frustration" because due to the day-to-day responsibilities I had with the ULTRA business (and other projects) I had no time to write/test my ideas. I also had zero interest in extending my responsibilities to that of managing programmers. So with ULTRA 11, solving this problem became a top goal which was more challenging than it appears. Most importantly, the solution must make the program easier to learn and not harder, which is always a challenge when adding software functionality. Well, I'd love to say that I invented some magical method that will be taught in logic classes for decades, but I did not. The problem has been "solved" simply using categorization. In ULTRA 11, users can categorize systems and data any way they desire. The heart of the logic is the CODE:DESC logic used with System Parameter Sets in ULTRA 10 where CODE is one character that represents a System Parameter Set called a "System Variation", and DESC is simply a longer description of that variation. In ULTRA 11 we've extended the CODE to three characters and the CODE:/DESC logic applies to most everything in ULTRA including Historical Analysis Control Parameters (discussed another day). For categorizing Systems, we call the CODE:DESC set a "System Set". A user could have of any/all of the "System Sets" below (or anything else they desire):
ULTRA's treatment of "System Sets" will evolve throughout the ULTRA 11 releases. But in ULTRA 11.00 you'll be able to create/maintain System Sets as well as display/hide those sets in all of the circumstances within ULTRA where applicable. For
example: ULTRA 11 will ship with systems that we've already grouped in many ways from which you can use, modify, delete, or whatever. Warning: (Large, bold red and underlined. How's that for drama...) What
follows next looks complex when written out in words. But when you've
seen it running, it's really very simple. So please don't leave this article
thinking that the logic that follows complicates the use of ULTRA. It
actually simplifies ULTRA while extending its power greatly. And as with
everything in ULTRA, learning/using advanced features is completely optional.
As I alluded to above, many users will be fine with: In ULTRA 11 the definition of "System" and "Data" have been dramatically changed/extended. Data "Data" is anything that is a series of anything associated with any frequency of dates in a file. Since there are no longer limits to what the ULTRA Historical Database can contain (See Unlimited Database Items below), a FastTrack Signal File is actually a database item (DBI) in ULTRA 11. Literally anything that can be represented as a date/data combination can be an ULTRA DBI. An exotic example would be the number of touchdowns a quarterback throws in each game over his career. The dates in the DBI would only be those upon which a game occurred, and the data is the number of touchdowns. That DBI may only have ten date/data combinations but it would still be completely functional as a DBI in ULTRA. Systems "System" has changed to include any logic that optionally takes data (DBI values) as input, and generates signals which "trade" DBI values through a series of dates of any frequency (daily, weekly, random, etc). Some examples of ULTRA 11 "Systems":
Lastly, one VERY key characteristic of an ULTRA 11 system, is that they are no longer restricted to one instance of an internal ULTRA system per analysis. For example, the restriction on only using one VMA system in an ULTRA 10 Composite Strategy Definition and Historical Analysis, no longer exists. New ULTRA 11 Systems as they apply to CODE:DESC System Sets The final "bottomline" of this article is that a user could create a CODE:DESC System Set named: "MA: Moving average Related Systems" which consists of:
If the "MA" System Set is enabled for Historical Analysis, all of the "Systems" above would be available for selection. In a function that operates on multiple systems such as ranking, you'd be able to select any or all of the systems in the "MA" set. .Unlimited
Database Items (DBIs) U11
allows for an unlimited number of database items called DBIs. DBIs can
literally be anything that's a series of date/number combinations over
some period in time.
DBIs can be added to the database by:
Users can categorize and hide DBIs. One of the past weaknesses of ULTRA that we've addressed in ULTRA 11 is simply "Information Overload" for systems. But this could also soon become a problem with data as we expand the ULTRA Historical Database. Exact
details on the method by which we've solved this problem for both systems
and data will be published in this file very soon. But for now in terms
of DBIs, here's an example:
With ULTRA 11, you
can categorize
your data any way you want, such as:
ULTRA
Financial Systems LLC © 2009 ULTRA Financial Systems LLC.
|