The Striker Report - gold standard for objective trading research

Murray A. Ruggiero, Jr.

Consultant: Tuttle Tactical Management / Vice President: R&D Trades Studio Inc

Program(s) Developed: Mar Long Bond 2017 Trading System

Interviewed by John F. Gallwas - Founder of Striker Securities, Inc.

Since the early 1990's, Murray Ruggiero Jr., after earning a degrees in physics and computer science, has pioneered artificial intelligence to develop trading systems for the financial markets. He is a consultant for "Tuttle Tactical Management" as well as Vice President of Research and Development at "TradersStudio Inc" and UsingEasyLanguage.com, where he designs trading systems and testing platforms.

John F. Gallwas: As one of the pioneers of trading systems business, what led you to the development of computerize trading systems?

Murray Ruggiero Jr.: When I was young, my dad worked at Yale University as an associate in research. He use to take me to work during the summer and some of the grad students and professors would babysit when he was running dangerous experiments. They would explain the research they were doing to me, this was when I was 5 to 6 years old. In fact I use to hang out in the meteorology department and was reading a college text book on meteorology was I was 7 or 8.

When I went to college I received an undergraduate degree in physics and an undergraduate degree in computer science, and got involved in trading in a backwards way due to my work in artificial intelligence.

While I was finishing my second degree in computer science, I started working at Olin Chemical, and I was helping the researchers use artificial intelligence to mine the databases of test results they had. I had a budget to try out all kinds of early machine learning products, it did not matter if the product cost five grand because this was a multi-billion dollar corporation. One of the scientists there threw the manuals back at me and said "Isn't there any damn thing I can use in my spreadsheet?" And that is what gave me the seed to develop a product called Braincel, which I developed in the late 1980s at Promised Land Technology, which I cofounded. The original version of this was embedded in a Lotus clone we developed , but when we saw Lotus fading we rereleased that product as an Excel add in, and I also received a patent on the concept of embedding a neural network into a spreadsheet.

Now we were at the Excel 3 and Windows 3 launch in Boston in 1991, and, I mean, we got a lot of press during that early time. We were in Business Week, we were in Wall Street Journal, and a lot of people that were involved in the financial markets wanted to use our neural net product to develop trading systems. And that's how I got involved in trading, was because I had to do support for these clients as well as consulting for these clients.

Well, what ended up happening, there were some issues with Promised Land and my partner and I'm not going to go into, but I ended up leaving Promised Land in '93. And in December of '93 or January of '94 I ended up calling Ginger Szala at Futures Magazine and offered to write some neural net-based trading articles for her, because I knew the person she was having write them was no longer working for the magazine.

I met Larry Williams in early 1995 and worked for him on retainer until the late 1990's I was his research hired gun. My job was to develop new trading concepts, Larry was great to me and we would divide my work into three bins. (1) Only for Larry (2) I could sell and Larry could use,and (3) I could use for articles and newsletters. He was very good to me and guided my early career and has help me become who I am.

After working for Larry, I continued to develop product for myself. And at that point, through the mid-to-late 1990s, I was one of the top TradeStation developers. In fact I was featured in Omega Magazine as their fourth cover. At the conferences I would chair with Bill Cruz, and he would bring me out in the dog-and-pony show as a TradeStation developer that was making a living selling trade station add-ons, which was part of the dream back then of people that bought TradeStation, was that they could buy a copy of TradeStation, write some code, quit their job and be able to play with trading full time. So that's how I got to where I did.

My strong scientific background gave me an edge over other people who were in our industry because I understood how to do scientific research. I understand developing a hypothesis before actually designing your tests. These are things I still use today in developing trading systems.

John F. Gallwas: Is there any specific technology you have developed that you are most proud and think has been important to the field in general.

Murray Ruggiero Jr.: One of the most important concepts I invented in my career is the concept of "intermarket divergence", which I first coined in 1998. This is a simple but powerful method for developing trading signals based on intermarket analysis that are both robust and very profitable. This method uses simple differences or moving average to define trend and then applies this simple concept. The first market I applied this to was the bond market.

Intermarket Divergence occurs when a traded market moves in an opposite direction to what is expected. For example, if we trade the S&P 500, 30-year Treasury bonds rising and the S&P 500 falling would be divergence since these are positively correlated. If we were trading the 30-year Treasury bond, both bonds and gold rising would classify as divergence since they are negatively correlated.

We will define an uptrend as when prices are above a moving average and a downtrend as when they are below the moving average. Now we can predict with some reliability the future direction of bonds, stocks, gold, crude and even currencies using this simple intermarket divergence model. Pseudo code for this basic model is as follows:

Price relative to a simple moving average
Let InterInd = Close of Intermarket - Average (Close of Intermarket,N)
Let MarkInd = Close Traded Market - Average (Close of Traded Market,M)

Positive correlation
If InterInd > 0 and MarkInd < 0 then buy at next bars open
If InterInd< 0 and MarkInd >0 then sell at next bars open

Negative correlation
If InterInd< 0 and MarkInd< 0 then at buy at next bars open
If InterInd > 0 and MarkInd >0 then sell at next bars open

This simple concept represented above has proven to be a robust methodology for predicting future price action using intermarket analysis.

I originally published this concept in Futures magazine in April 1998 using a simple bond system traded using utility stocks. This systems rules were published in that magazine over 20 years ago and this system still works well today. The story of this system and how I discovered it actually was the inspiration for my paper in 2012 at the CIFEr Conference, which is the IEEE.

Computational Finance Conference, and I revisited an intermarket system that I've published in 1998. It was "Trading the US Treasury Bond Using Utility Stocks." And I was originally using the New York Stock Exchange utility average, and when that was discontinued, I just substituted the Philadelphia Electrical utility average. And I did not change the parameters of the system since 1998 when the article was first published. And the system has continued to perform very, very well over all this time.

Now one of the things I did when I built the system was that I actually did a three-dimensional plot of my parameters versus profit, and looked at that three-dimensional surface. Now, if you look at the three-dimensional surface, the surface was very, very robust, but if I looked at the surface from an inflation measure, and, let's say, how did the CRB Index predict bonds, even though the performance of that looked very good, it made almost as much as the utility stocks did, I rejected it because the underlying three-dimensional space was too rough.

Looking at a three-dimensional plot of different parameters versus profit will give you an idea about how robust the underlying relationships are, and you could predict which relationship will hold up well into the future.

John F. Gallwas: Our last Striker Report interview with you was in 2014, when you started www.UsingEasyLanguage.com, can you tell me a bit about that?

Murray Ruggiero Jr.: This site has dual purpose, first is to be educational, and second is to sell tools and trading strategies which are useful to help traders become more profitable. One of the big things I believe in is it's easier to trade a system when you understand all its rules, that why the systems and tools I sell on my site are fully disclosed, I require you to sign a NDA for most of my products but I believe in open code and selling at a reasonable price so that people can learn and get the confidence they need to trade.

John F. Gallwas: Which brings us to your newest trading system traded with live results at Striker called, "Mar Long Term Bond Intermarket 2017 Trading System". What are the features of this system?

Murray Ruggiero Jr.: This system uses two different inter-markets, and most important it is different from my earlier systems in that, one of the two inter-markets is based on new research because one of the original market I normally use no longer works well over the past 2-3 years. Things have seemed to change and this market does not represent inflation risk like it did for the last 20 years. This required me to find a new intermarket which represented inflation risk and was predictive. I found this and released my newest bond system in June 2017. Another change in this system is it�s symmetrical, it�s as easy to go long as short. This was not true of some of my earlier bond systems developed during the great bull market in bonds from 1996-2015. I always had an upward bias in these systems, but now bonds are a two way market.

John F. Gallwas: What is the current status of using artificial intelligence to trade futures?

Murray Ruggiero Jr.: There two different areas I am working on, one is advance modeling the other is machine learning. One example of advanced modeling methods is the Arima/Garch Hybrid model. I use this to predict log difference of price one to three bars into the future. We can predict futures using ratio adjusted data and using the predictions in a systems based on back adjusted data. This methodology also works on ETF's and stocks. It's very computationally expensive and we have developed research tools which use cluster computing on the AWS cloud.

Another branch is machine learning, more classic things like deep learning, xgboost and other learning algorithms which can mine relationships in data. I think over the next five to ten years, simple system will no longer have an edge and without access to this type of technology the average trader will not be able to compete.

John F. Gallwas: What are your risk-reward expectation for this bond trading systems?

Murray Ruggiero Jr.: We have a winning percentage of 57% on this system with a good win/loss ratio of 1.47 as of the date of this interview. Since Striker's site updates daily readers can view current performance by going to your site.

John F. Gallwas: Is there anything in the "what's new" department you can share with us?

Murray Ruggiero Jr.: Yes, I am working on several new areas. I am working on modeling markets with not only arima/garch but also deposing market cycles and creating a prediction model. I am also looking into many areas of machine learning and modeling, these include deep learning, hidden markov models and several other areas. Markov models are good to detect regime switching. I am also looking at building cryptocurrencies models based on internals of their block chain both bitcoin and Ethereum.

This interview is for informational purposes only and is not intended to be a solicitation of any kind. Trade only with risk capital. The risk of trading can be substantial and each investor and/or trader must consider whether trading systems are a suitable investment.