Benefits of custom software vs. Trading Blox

Questions and discussion of Trading Blox and other platforms for non Trading Blox customers. Trading Blox customers should use the Trading Blox Support forum.
Post Reply
sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Benefits of custom software vs. Trading Blox

Post by sunyata » Fri Jun 12, 2009 9:13 pm

Hi everyone,

I have been working very diligently on developing a custom simulation engine in C# language and it is taking a very long time. I estimate that to program software with as many features as Blox Builder--and one that I would feel is accurate--might take me another 8 months to a year. Thus, economically Blox makes a lot of sense: if I establish that my labor is worth $12/hr to me (I am cheap), at 55 hrs/wk for 8 mos, that is over $21,000. And there's no way I would be able to devote that much time to developing the software, given I will have to find a job sometime soon. Thus, I am seriously considering buying Blox and have already ordered the trial version.

The only reason I began developing my own software to begin with was because Seykota recommends it on his Trading Tribe website. However, I think he may have been ignorant of Blox at that time, which was in early 2006. His main gripe was that software cannot "simulate interactive portfolios with multiple systems, multiple instruments, multiple entry and exit points - that also respond to upswings and drawdowns in equity." However, I believe Blox can do this, except maybe for responses to changes in equity. Before I buy Blox though, I want to put out the feelers and see if anyone here can comment on the benefits of custom software over and beyond Blox Builder. If you could have custom software built for free and ready for you by tomorrow morning, would it be significantly different from Blox? What would it include and what wouldn't it include? Would there be things that you could do that you couldn't do in Blox Builder? What would they be?

Also, it seems that custom software might be important to have if one was going to run a professional fund. Do you also think this is the case? I feel as though potential investors might not take someone who uses retail software as seriously.

Finally, the answer might be not to abandon my coding entirely, but to use it as a powerful supplement to the Blox software.

Thanks everyone!![/i]

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Fri Jun 12, 2009 10:53 pm

There are discussions and recommendations on this very topic, right here in the Forum. Look in "Custom C++ Platforms" and elsewhere.

Besides what's already been said, I would add one more thing:

Remember that you'd be paying $21,000 for the first copy of your software. All subsequent copies would be free. If you can envision needing to run numerous copies of backtesting software (for example, on several machines) all at the same time, then at some point it's cheaper to pay $21,000 once instead of $3000 many times. This same thinking is what leads many organizations to choose "make" in the "make versus buy" decision: (1 x $big) is cheaper than (N x $small) for sufficiently large values of N. Intel and IBM wrote and use their own C.A.D. software (rather than buying from vendors) for this very reason.

Asamat
Roundtable Knight
Roundtable Knight
Posts: 175
Joined: Fri Jun 03, 2005 7:50 am
Location: Walldorf, Germany

Post by Asamat » Sat Jun 13, 2009 6:18 am

I have another thought to add, having gone down a somewhat similar road. The thought is that the choice does not need to be one or the other.

You can use TB for what it does best, use other software as well, and use your own code for what you need additionally, and to stitch it all together. In a corporate environment this is called best-of-breed approach. TB is the best for backtesting, but not best for some other tasks.

I did write my own code for part of my analysis, while using TB in parallel. From the beginning it was clear to me that I would not attempt to reinvent what other have already done (and better than I'm able to). So I use open office spreadsheet as an output medium (not wanting to write my own spreadsheet). I use the (free) cernlibs for graphical display of my analysis and optimization calculations (not wanting to code my own display). I use ta-lib for trading-related functions (not wanting to spend the time testing my own implementation). I use Amibroker for graphical display of charts and indicators (sorry, Tim, but it's better than TB in this area). I use TB for backtesting (and seriously doubt that you, sunyata, could replicate on your own all the experience built into it, whatever the time frame). I export a lot of data from TB to csv files, which I then use in my own code, to analyse futher, for checks, for graphical output, for documentation and order display.

sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Post by sunyata » Sat Jun 13, 2009 10:23 am

Asamat and Sluggo, thank you so much for your responses!

I have found answers to a few of my questions within other conversations. For example, Trading Blox can reference the equity curve in generating new Blox. Also, a thread on Designing Your Dream Software from 2004 lists many wish-list features, most of which Trading Blox seems to have incorporated since then.

In regards to Sluggo's comment, if I purchase Trading Blox, may I only run it on one machine? I ask because I have both a laptop and desktop computer.

Asamat, I agree with you wholeheartedly that it would be very difficult for me to write a backtesting engine that would rival TB, especially given that I have no experience trading futures. Also, can you talk a little more about ta-lib? I am not quite sure what you are using it for. Does "ta-lib" mean "Technical Analysis Library"?

Thanks again.

Roger Rines
Roundtable Knight
Roundtable Knight
Posts: 1946
Joined: Wed Oct 06, 2004 10:52 am
Location: San Jose, CA

Post by Roger Rines » Sat Jun 13, 2009 12:09 pm

sunyata wrote:In regards to Sluggo's comment, if I purchase Trading Blox, may I only run it on one machine? I ask because I have both a laptop and desktop computer.
You are allowed to install Trading Blox on two computers. You also can have more than 1 Trading Blox installation on a computer. I have several TB installations on my development computer, so I can keep special projects undisturbed on their planned efforts.
sunyata wrote:Asamat, I agree with you wholeheartedly that it would be very difficult for me to write a back testing engine that would rival TB, especially given that I have no experience trading futures.
I was surprised to see how much I agreed with Asamat. His described approach is a close reflection of how trading operates here. Even his use of other software based upon its perceived strengths according to its abilities.

If you, or others following behind you have this same question, take away anything from what is shared above, the point is to remember the ability of Trading Blox, and other programs like TradeStation, AMIBroker, TradersStudio, and many others didn’t happen by accident, or a lot of expense and heartache. In each program is an enormous amount of time that has already been spent to get it to its current performance level and reliability. While all the above programs are great in their own ways, that development time spent can't be made up by a small development team, and still have a new platform ready for reliable testing and signal generation for a long, long time out into the futures. More than likely, it will never be as good as any of the above, for what the above are especially good at doing, and you’ll spend so much energy making your effort work that you’ll then be in the software business with the hope of a future goal to learn how to trade.

Another important point to consider is the peer groups that are in place around each platform. That support and advice for a new platform takes time to attract people who will be supportive and competent with a platform, and who also are successful with trading. What will be the value of that kind of community for someone at your stage of learning?

What is listed here is from my experience, and it has shaped a belief you ask yourself the question of whether you want to be a software developer who is playing a low volume competitive catch-up game, or be a systems trader who is discovering and applying winning solutions, of which you have yet to discover. In simple terms, which effort path will generate the best risk adjusted returns for meeting your most important goal?

As a programmer, I have learned it is best to save programming prowess for the specialized tools that nobody offers, or at least doesn't offer them in a fashion with how they are needed to function. That strategy minimizes distractions from learning the ways of successful trading, and it prevents me from spending enormous amounts of time on software that is easy to buy.

What strategy will ensure you are focused on discovering what you need to find success in your trading accounts?

sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Post by sunyata » Sat Jun 13, 2009 12:31 pm

What is listed here is from my experience, and it has shaped a belief you ask yourself the question of whether you want to be a software developer who is playing a low volume competitive catch-up game, or be a systems trader who is discovering and applying winning solutions, of which you have yet to discover.
Roger, I am very appreciative for your response. The feeling that I was becoming more of a software developer than a trader started to set in and was a main motivation in considering purchasing capable off-the-shelf software. I am thankful for the programming skills I have acquired in the past few months because I think they will allow me even more flexibility in devising solutions, however I don't want to be spending countless hours (and/or months) debugging code to, for example, simulate contract roll-overs when I could be designing systems. Especially when there is software available that is accurate and sophisticated enough to achieve what I want to achieve and yet is still within a reasonable price range.

Also, how do you feel about the future viability of Trading Blox? Do you think it is hear to stay?

Thanks again.

Asamat
Roundtable Knight
Roundtable Knight
Posts: 175
Joined: Fri Jun 03, 2005 7:50 am
Location: Walldorf, Germany

Post by Asamat » Sat Jun 13, 2009 1:46 pm

sunyata wrote:Also, how do you feel about the future viability of Trading Blox? Do you think it is hear to stay?
Don't worry. More and bigger operations depend on it that you imagine.

Roger Rines wrote:Even his use of other software based upon its perceived strengths according to its abilities.
Thanks, good to hear that you agree. Do you have anything to add to the list?

sunyata wrote:Also, can you talk a little more about ta-lib? I am not quite sure what you are using it for. Does "ta-lib" mean "Technical Analysis Library"?
Yes, spot on. I code in java, the library can be included there. If you use other languages than the one offered, you need to search the web for something similar.

I use it for calculating trading-related functions in my code. The problem with writing functions like these yourself is not so much that they are complicated and therefore difficult to code. They are not. The testing effort, however, to convince yourself that it's sufficiently error free to bet you money on it in the market, is what prohibits doing it yourself and alone. I found the ta-lib results identical to TBs (and to my own implementation, where I did it anyway), over many instruments, many years of calculation, and down to more than 6 significant digits. That gives a good and reliable feeling in all directions.

ta-lib is not very well documented, it needs a bit of figuring and experimenting. That's the price for getting it for free. Since the source code is given, it's possible to so it, however.

babelproofreader
Roundtable Knight
Roundtable Knight
Posts: 138
Joined: Wed Nov 10, 2004 4:36 pm

Post by babelproofreader » Sat Jun 13, 2009 2:01 pm

You might consider the software packages "R" and "Octave." Both are free and open source and writing scripts to run any imaginable test would quite straight forward if you are already familiar with programming. "R" is statistical testing software and "Octave" is almost exactly equivalent to "Matlab" in its functionality. Charts of the results can be obtained using "Gnuplot" software, again free and open source. At the moment I use all three exclusively for my back testing and system development.

sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Post by sunyata » Sat Jun 13, 2009 3:51 pm

Thanks. I looked into ta-lib briefly and it looks like there's support for C#/.Net, which is the language I've been using. Thanks for the tip. I'll be checking it out. Also, someone in another thread mentioned "R" too. I didn't realize there were so many open source solutions available. That's fantastic!

I wonder if anyone has ever thought to open source backtesting software...? That'd be pretty nifty.

Peace.

Roger Rines
Roundtable Knight
Roundtable Knight
Posts: 1946
Joined: Wed Oct 06, 2004 10:52 am
Location: San Jose, CA

Post by Roger Rines » Sat Jun 13, 2009 4:26 pm

sunyata wrote:Also, how do you feel about the future viability of Trading Blox? Do you think it is hear to stay?
I sure hope so. Of all the platforms I've used and seen over the years, Trading Blox is the best platform for building a realistic trading plan. Trading Blox management also has a plan in place to allow users to keep the software working should development end unexpectedly.
Asamat wrote:Thanks, good to hear that you agree.
Yes I agree. Our strategy of selecting the best tool for the job has been very effective. However, it took me a while to let go of the need of "Rolling my Own" software for everything.

We do differ in that I don't use Tomasz's software, but instead my other platform is TradeStation. I've been using Omega’s software since early in 1991 when System Writer was discontinued. Still, I have followed Tomasz's progress and in my discussions with him years ago I found him committed to his product and caring about his customers. For certain I've been impressed with how broad a base of users he has developed, and how active and generous those user have been in their postings.
Asamat wrote:Do you have anything to add to the list?
This question is so broad that I don't know how to answer, given how much I post in the Trading Blox section of this forum.

sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Post by sunyata » Sat Jun 13, 2009 6:56 pm

Amibroker is "Tomasz's software?" Also, the list of Amibroker's features is extensive and seems to match up with most of TB's, but yet it is 1/10 the price. What am I missing?

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Sat Jun 13, 2009 7:50 pm

Let "P" represent the probability that you will find AmiBroker completely suitable. Suppose you buy AmiBroker first. If you find it completely suitable, you've only spent 0.1 units of money. You've saved a bundle. On the other hand, if you buy AmiBroker first but you don't find it completely suitable, you then purchase Blox, for a total cost of 1.1 units of money. (0.1 units for AmiBroker + 1.0 units for Blox)

What is the expected value of your total cost?

E{C} = (P x 0.1) + ((1 - P) x 1.1)

Whenever E{C} is less than the cost of Blox (1.0 units of money), it's a smart bet (a positive expectation bet) to buy AmiBroker first. Thus you will be very eager to know: What is the break-even value of P which sets E{C} exactly equal to the cost of Blox?

Algebra says: break-even P = 0.10.

So, if you believe there's a 10% (or better) chance that you'll find AmiBroker suitable, it's a positive expectation bet to buy AmiBroker before Blox.

kingami
Full Member
Full Member
Posts: 15
Joined: Mon May 04, 2009 12:59 am
Location: Atlanta

Post by kingami » Sat Jun 13, 2009 11:46 pm

sunyata wrote:Amibroker is "Tomasz's software?" Also, the list of Amibroker's features is extensive and seems to match up with most of TB's, but yet it is 1/10 the price. What am I missing?
I have been using Amibroker since 4.0 (latest version is 5.2) and seen its features grow very much. It has a very powerful backtest engine, which I believe can accomplish everything TBB can do if you spend time to program it. However, it is very complicated and low level coding, although would be less than C# since you don't need to do any indicators by yourself, just need to implement the simulation loop as in TBB to support a multiple systems and multiple instruments.

I have done quite a bit programming in Amibroker, but finally decided my time is better spent on trading research not on programming and purchased TBB last month. I am still in the process of learning TBB and find it very powerful in backtesting.

Amibroker is very good in charting and has more features than TBB. But when it comes down to what matters in making money, I believe TBB is the right choice, especially for system trader. I am using both, AB for charting and interface with IB broker and TBB for system building.

kingami
Full Member
Full Member
Posts: 15
Joined: Mon May 04, 2009 12:59 am
Location: Atlanta

Re: Benefits of custom software vs. Trading Blox

Post by kingami » Sun Jun 14, 2009 12:04 am

sunyata wrote: The only reason I began developing my own software to begin with was because Seykota recommends it on his Trading Tribe website. However, I think he may have been ignorant of Blox at that time, which was in early 2006. His main gripe was that software cannot "simulate interactive portfolios with multiple systems, multiple instruments, multiple entry and exit points - that also respond to upswings and drawdowns in equity."
I saw Seykota's note. He developed his own backtest software and thus suggested people do the same in order to get a deep understanding of how it works. However, that's a huge undertaking to say the least, Seykota has been doing it for decades. Choose a system building platform is the way to go IMO.

Choosing a system builder (not a black box) such as TBB does not conflict with Seykota's advice, because you still need to do a lot of development by yourself. But you are on a much higher level. You would be thinking about system rules and other components, not about C# objects or functions.

I have done some research on the platforms and found that there are only a few which can meet Seykota's requirements. Trading Recipes (Mechanica), TBB, Amibroker (to some extent), and maybe Wealth-lab (no longer available in US for no-Fidelity customer) and finally settle on TBB for its great word-of-mouth and simple yet powerful user interface.

kingami
Full Member
Full Member
Posts: 15
Joined: Mon May 04, 2009 12:59 am
Location: Atlanta

Post by kingami » Sun Jun 14, 2009 12:16 am

Roger Rines wrote:
sunyata wrote:Also, how do you feel about the future viability of Trading Blox? Do you think it is hear to stay?
I sure hope so. Of all the platforms I've used and seen over the years, Trading Blox is the best platform for building a realistic trading plan. Trading Blox management also has a plan in place to allow users to keep the software working should development end unexpectedly.
Roger: could you elaborate a bit about the plan if it is convenient? (or PM me). Although I just started to build my systems and not yet in actual trading, that's a concern on the back of my head. TBB is a great platform yet you don't see a lot of publicity or marketing of it. In the world of trading software, most people are attracted to flashy, colorful charts and countless indicators, not the invisible backtest engine.

Roger Rines
Roundtable Knight
Roundtable Knight
Posts: 1946
Joined: Wed Oct 06, 2004 10:52 am
Location: San Jose, CA

Post by Roger Rines » Sun Jun 14, 2009 3:06 am

sunyata wrote:What am I missing?
Experience.

Experience in understanding how the strengths of the two platforms differ, and why someone would choose one over the other, or why someone would want to have both.

At your stage of minimal understanding of the critical aspects needed for successful system development and execution, including how money management, market correlation, risk and portfolio size management controls and many other areas can make, or break a what seems to be a successful entry-exit method.

Your lack of experience isn't a failure, but instead should be warning that there is a lot learning ahead of you. It should also be a beacon on the hill you work your way towards gathering as much information as you can along the way.

Also consider that traders who have been at this for a long time think the difference in cost between the two packages, or even the cost of having both, isn't anything more than a small part of what can be lost on a single trade.

As a beginner who doesn't understand much about trading yet, you might find AMIBroker to be a good learning tool so you can get past the stage where you find what seems to be a good looking system in testing, falls apart not long after you start trading. A lot of people never get past that part of the learning curve and for various reasons, include wiping out their accounts they just drift off into their ether as they loose any belief that people can be successful trading.

However, if you can get past the many frustrations that are certainly ahead of you, and can muster the tenacity and motivation of bulldog after a bone, you'll still find the next part of the learning frustrating, but you will then be more experienced to know what questions to explore. As your experience grows, you should be able to see how many people are narrow in their system development methods, and how few people understand that successful trading needs more than a good system, it also needs a well thought out plan that is tested across a lot of data.

I stand by this statement: “Of all the platforms I've used and seen over the years, Trading Blox is the best platform for building a realistic trading plan.â€

sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Post by sunyata » Sun Jun 14, 2009 8:47 am

This is a fine thread. I am very happy to be apart of it.

Thanks kingami and Roger for your insights. They certainly help me clarify my decision process. I would like to minimize the level of programming for basic features and it seems that that's what Trading Blox offers. According to Roger, it also offers a level of realism that is incomparable to other suites and it such a way that is not immediately apparent.

And even though my bank account is dwindling for lack of a job, and is under mild distress, to say the least, the $3000 for Blox Builder does seem worth it, especially given the amount of work I've had to put into my own custom software just to get to my present nascent stage.

Again thanks everyone for their help. I greatly appreciate it!

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Sun Jun 14, 2009 9:19 am

sunyata wrote: And even though my bank account is dwindling for lack of a job, and is under mild distress, to say the least
There is another thread touches on the question of the viability of short term and day trading.


viewtopic.php?t=6610&highlight=

In a most interesting private exchange with one of the contributors, I learnt that day traders can (emphasis on "can") make a good living trading very short term using very little capital. But that this is a job - a real, full time job and a stressful one, glued to a screen all day long.

Trading longer term requires capital, quite a lot of it. Trading longer term can be viewed more in terms of making a return on capital. It is closer to “investingâ€

sunyata
Roundtable Fellow
Roundtable Fellow
Posts: 94
Joined: Sat Jun 06, 2009 4:27 pm
Location: Charlottesville, VA

Post by sunyata » Sun Jun 14, 2009 9:42 am

AFJ,

Thanks for your note. In fact, I was an equities daytrader for 2 years and can attest to the opportunity to make many multiples of one's starting capital. I traded for an arcade firm that required $5,000 initial capital contribution and provided 100% payout of trading profits, but to make its own profit charged commissions between $.002/share to $.008/share. A trader my age, two years out of college, was clearing over $20,000/month regularly and there are traders in their mid-to-late 20s who are earning over $5,000/day trading their own accounts. It is certainly possible to make an excellent living as a daytrader, but it was not for me. I gravitate more towards long-term, systematic strategies--rather than short-term, discretionary strategies--particularly because I have difficulty making quick decisions based on gut feeling, preferring strict rules to guide trading decisions, and also because I need to know via historical back-testing that my trading rules have an edge. It was also a very stressful job (my boss was 30 but looked over 40) and I felt it was unhealthy and to some extent unbearable. To be successful as a daytrader would require that I undertake a radical self-transformation and I would probably run out of money before that could ever happen.

I am going to have to find a job of some sort certainly while I design my trading system/suite of systems. My ultimate goal is to start a fund that utilizes long-term mechanical trading strategies, and I am most likely going to need to seek out seed capital from third parties. I am not very well-connected in the business so that may be difficult. But it may also be that if I can build a track record even with $500,000 seed money from somebody willing to take a shot on me, then in due time I can attract more money. I don't necessarily need to be well-connected although it would probably help (which is why I am trying to get into a topnotch business school to build my Rolodex, but I think the likelihood of that might be pretty slim, given my alma mater and work experience). All in all, if I am going to do this, it probably won't happen as fast or as easily as I had hoped.

Best to you all.

Post Reply