Trading System Lab

TSL comments

TSL user references may be available for "serious", interested parties since numerous TSL clients have agreed to be product and company references for TSL. Contact TSL for additional information.


I've been using TSL for a year and am very pleased with it. TSL is now an integral part of my CTA business and has helped me manage the significant task of quickly designing and deploying trading systems during a time of rapid changes in the futures markets. Instead of taking on the associated risks of hiring additional staff, and hoping that they can actually create robust systems, I was able to quickly and affordably employ TSL to address the current and future development needs of my trading business. What follows is a summary of why I chose TSL and how I use it in my business.

I have spent many years developing trading systems using my own methods and understand the time and effort required to create and maintain robust systems. Done properly, it is not a trivial process even in the best of times with relatively "cooperative" markets. At the time I was evaluating TSL, I was in the process of deciding how to best adjust to the rapidly changing futures market environment. Whatever the potential solution being evaluated, the three questions to be answered were: 1) What will be the quality of the systems (robustness over time), 2) how long will it take to develop and deploy, and 3) what will be the total cost for the life of the solution.

I explored a number of options and products but ultimately I chose to license TSL for the following reasons: 1) Robustness / Quality of output --- Based on live demos and other data, TSL appeared to truly be able to create robust systems on many markets, using various styles and time-frames that hold up over time. Without output of robust systems, you've got nothing! (See Mike's systems tracked by Futures Truth and other user comments.)

2) Time required to develop and deploy systems --- It appeared from the TSL videos and live demos that I would be able to develop and deploy systems at a very rapid rate relative to my typical manual methods. Somewhere between a few minutes and a few days. And as a key part of that process, if a market seems like it's tough to trade, you can abandon your effort sooner and move on to other opportunities... System development is an iterative process with many, many dead ends interspersed with some successes. The developer uses that experience to learn and decide what logic to keep or discard, and what to do next while designing system logic. TSL performs this process at an absolutely amazing speed. 3) Total cost over time --- I concluded that TSL was likely to produce robust systems very quickly and that the cost over time was very reasonable. Computed as a monthly expense (license fee is actually lump-sum), I found the amortized cost to be in line with the costs of my other critical functions, such as trading platform and data fees, and managed account administration service (performance, accounting, reporting, etc.). One should consider that for essentially the price of a fully functional seat of CQG products you can have your own R&D department. Contemplate the costs and various risks of hiring one or more people to do the same work, and TSL looks like an incredible bargain. Further, it is a force multiplier in that it delivers what I need quickly in large quantities without using up huge amounts of my time. This leaves more time to operate and manage the rest of the CTA business.

There are a couple of other issues. As a long-time system developer, letting a machine build the logic felt uncomfortable and seemed like a large leap of faith. After all, I have thousands of hours of developing, testing and trading that created my skill sets and shaped my biases. How could I agree to let a machine take over the design function when I have invested so much time in developing my craft? The way forward for me was first to realize that any "secret sauce" elements that I have coded could be integrated into the building blocks or "DNA" that TSL uses. Secondly, as I studied how TSL functions, I realized that it's system construction process was very similar to my manual process, except infinitely faster, unflinchingly systematic, and not constrained by the biases that I have built up over the years. It turns out that TSL is in effect more "open minded" to strategy innovation than I have been. An initially humbling, but ultimately liberating realization. My conclusion was that TSL makes me the supervisor of an almost unimaginably fast system developer, or a "Super Minion" as I like to call it. As I continue to use TSL, I realized that I still use my market knowledge and many of my system development decision skills in exploiting new markets, assessing TSL's results and implementing the systems.

TSL is the trading system design and manufacturing facility for the trading system designer looking to make the leap into the future. After using TSL, I'm not nostalgic about the old manual process days.

Conclusion --- Since I've been using TSL I've made significant positive changes to my trading systems and overall business. So far, the strategies have been making money during a variety of market conditions. Client account balances are up, and I have peace of mind that I can quickly and effectively address any market and trading system challenges that arise.

A CTA/Fund Manager and Trading System Developer


I applaud TSL's efforts in pushing the state of the art in Machine Designed Strategies. Just as, on "Jeopardy", Watson showed that programs can understand language, your research is showing that programs can help uncover strategies too complex for boundedly rational humans to discover on our own.

Dr. Norbert Pierre

Assistant Professor of Finance

Youngstown State University


Trading System Lab has taken our powerful, machine-code, genetic-programming engine and honed it into a focused, well-organized and easy-to-understand automated trading systems development system. In doing so, TSL made technical, design, and conceptual leaps I would not have thought possible-and I'm one of the authors of the leading university textbook on Genetic Programming. The design of the trading-specific fitness functions in TSL is, in my opinion, groundbreaking.

Altogether, a remarkable product.

Frank Francone


RML Technologies, Inc.


We have worked some with TSL and appreciate and acknowledge the efforts and emphasis that TSL has placed on the product and the science. We believe the future is bright for machine designed trading strategies and TSL.

David Andre, Ph.D

Chief Executive Officer

Cerebellum Capital


Mike Barna has created a piece of software that creates trading systems. His TSL software uses artificial intelligence to develop systems that can trade all markets. If you are interested in this "machine" approach to system development I would highly advise you to download Mike's white papers from his website. You must have an open mind when it comes to this type of trading system development paradigm.

Can a computer beat an experienced technical analyst? Some people may not think so. However I have discovered that many analysts are biased toward one certain approach. The computer isn't. It looks at every possible combination of methodologies and leaves no stone unturned. This approach is simply one step above on the evolutionary ladder of mechanical trading systems.

Mike submitted his three systems back in January 2008 and told me to wait 18 months before publishing any details. I did as Mike instructed and the systems have been on my Top Ten tables since I started tracking them. In 2010 he did the same with three more systems and we are now tracking those too. Are they the Holy Grail? No. Like their human counterparts they do lose sometimes and have similar type drawdowns. However, they have performed well during the time I have been monitoring them. In the latest issue of Futures Truth Magazine the TSL systems occupy four slots in the Top Ten table.

The true benefit of Mike's software is that it gives the ability of system development to anybody. Its user doesn't need to know anything about trading systems. You simply plug in your risk parameters and the market that you want to trade and the computer goes to work for you and spits out EasyLanguage code for you to input into your TradeStation. I have probably oversimplified the approach but you get the idea.

Initially I thought maybe the software was simply curve fitting a system to the data and the product would simply fall apart over time. However, after talking to Mike I discovered that he uses Artificial Intelligence constructs to perform out of sample data testing to help develop robustness during development. So far the three systems (six now) that I have been monitoring for over three years have shown robustness in that they have succeeded during a very volatile period of time.

George Pruitt, Futures Truth Go to the Futures Truth website »


Mike and I have worked closely together for over 20 years. Therefore, I know he is an intuitive and insightful system designer. He originated the use of both in-sample and out-of-sample results in system design, giving confidence of robust performance into the future. I am delighted he has enabled the use of MESA DNA in the design of his trading systems."

John Ehlers, Mesa Software, author of 4 books on Technical Analysis and trading including "Rocket Science for Traders" Go to the Mesa Software website »


I have been a TSL user since May 2010, and would like to share my experience with the software thus far.

I'm a professional trader at a top tier investment bank here in NYC. I work in the interest rate sensitive group, where I'm responsible for making a market in various securities as well as proprietary trading principal accounts. I have used TSL to create automated trading systems on a number of different futures traded in Chicago, most notably TY (10 year interest rate future) and US (Long Bond interest rate future). I trade large sizes, anywhere from $50 million ten year treasury equivalents, all the way up to approximately $100 million ten year treasury equivalents.

My experience with TSL has thus far been a very positive one, and more importantly, a PROFITABLE one. I design AND trade a number of different time frames on both TY and US. The time frames have ranged anywhere from 240 minute intervals to Daily bars. I am a NON-PROGRAMMER, and was able to use TSL (which is very user friendly in my opinion) to create systems with outstanding looking equity curves, which have proven to hold up in real time just fine. I'm not sure how many people here are familiar with the rates market over the last few years, but it's been difficult for anyone to trade it, let alone trade it profitably.

There is a learning curve with TSL, and I'm sure this is the case with any other software developed by TSL competitors. If you put in the time and effort to learn how to use TSL, learn how to work with data, and learn how to develop trading systems with favorable statistics (excellent profit factor, large enough trade size, good trade-to-parameter ratio), then in my opinion you have a very high probability of being able to take money out of a market. Being a NON-PROGRAMMER, I was able to learn how to do all of the above in a few weeks time, and was able to master it in a few months. It's been a truly rewarding experience.

Did anyone happen to mention that a TSL generated system (TSL_CEL_NG_1.1) currently claims the #1 spot on Futures Truth in the category 'Top 10 Systems Since Their Release Date'. Please see the below link for proof: Side Note: As of this post, TSL also holds the #6 AND #9 spots in this same category. It should also be noted that TSL also holds THE TOP 2 SPOTS for the Futures Truth category 'Top 10 S&P Systems'. Again, see for yourself:

Personally, what I find most valuable about TSL is it's ability to FIND a profitable trading technique for a particular set of data. A recent TSL update has included this feature in the software. For example I can load E-Mini S&P 500 data into TSL, and have TSL 'test' different trade types, and have it locate the most profitable one based on my criteria, such as profit factor or average trade size, etc. TSL may find that entering the market at the open is more profitable than entering at the close, for example. This is INVALUABLE, as this could take MONTHS to develop by hand, whereas TSL can generate the system in a few minutes time. Why not have TSL FIND the most profitable way to trade a market, instead of me TELLING TSL HOW to trade the market. TSL will not only find the most profitable ENTRY/EXIT for a market, it will find the IDEAL time frame on which to trade the market.

As a TSL client I've also been working with other clients to develop more active trading systems on SPY, E-Mini S&P 500 (ES) and Ten year interest rate futures (TY). The OMS (Order Management System) we're currently working with has been an excellent fit with TSLs capabilities. So far the more active systems have proven to be VERY PROFITABLE, ideal for smaller accounts who can't handle drawdowns.

In conclusion, I would say TSL is second to none, and as far as building algorithms to trade a market, it's proven itself in REAL TIME, which is all that really matters. The systems I built OVER 18 months ago are still running, and profitably so. I'm sure this leaves many forum members asking if TSL is worth the cost. In my opinion, it is ABSOLUTELY worth the cost. Rates Trader

A Top Tier NYC Investment Bank Fixed Income Trader


I am a TSL user and have designed a couple of high to medium frequency systems that are profitable in a variety of futures markets and stocks. I spent a lot of time trying different settings from bar size, entry tactic, etc. I did not have much luck in medium to longer term systems. I have tried other products doing similar things. None of these came close to what TSL can do, still it is a lot of work and research to find a robust strategy. The developer of TSL is often updating the software with new features to add value and to speed up processing. The newest enhancement called EVORUN sounds like it a very valuable tool that does have the potential to find robust systems across a variety of bar sizes and trade types. Prior to this feature you had to try all the entry tactics manually. This will speed up things quite a bit, I haven't had a chance to try this new feature yet, so I cannot say with certainty how valuable it will be. TSL requires a lot of hours of work to get to something viable to trade in today's markets. This is not for the part-time trader in my opinion.

Proprietary Trader, Chicago


TSL was one of the best investments I have made for my company since its inception. TSL is not for the ordinary client who is looking to create 1 or 2 systems; it is for a client who needs to run many systems. Once you get familiar with TSL you are able to create extremely robust systems on almost anything that trades. One of the best things I have found to be true with TSL is that every run you do to create a strategy is different than the last. So instead of just running 5 lots on one strategy I will make 5 systems that are all slightly different and run it on the same symbol. This has actually decreased my drawdown per symbol as well. Another huge PLUS about TSL is that the creator, Mike, is extremely dedicated to his clients and you can get support at almost any hour of the day or night. I think for any fund looking to create robust systems on almost any market TSL is the best thing out there. I agree 100% these are only tools that help you get where you want to be easier and there still a lot of human interaction, but the results with using these tools (using them properly) is far Superior to anything you can get without the tools. I can speak from personal results using TSL.

Proprietary Trader, Florida


I have been an Easylanguage programmer supporting TradeStation clients around the world for more than 20 years and on more than one occasion had the opportunity to support clients using Mike Barna's TSL product. The preprocessor is a strategy written in Easylanguage that is simple to modify to generate custom user input into the TSL product. Anyone capable of writing the Easylanguage for a trading strategy or an indicator should be able to modify the TSL preprocessor code without a hitch. In addition Mike has always been a ready source of support to his clients and has the highest reputation in the industry.

William Brower, Inside Edge Systems, Author of "EasyLanguage Learning by Example Workbook 2012" Go to the Inside Edge Systems website »


This is my first grid with Evorun. This thing is amazing. It will have a huge impact on our systems. Mike the features you have added to TSL 1.3 are huge, I think this will be a real game changer.

Private Trader, California


I've been working with TSL for a few weeks, and it is fascinating to watch the product at work. The resulting systems appear to be much more robust than those created using traditional system development methods. TSL seems almost too good to be true, but everything checks out under the covers. It's definitely a product I will recommend to clients.

Darryl Tremelling, Voyager Trading Solutions