If you give the increasingly-common phrase “artificial intelligence” a quick search on Google, you may notice a common theme: 2017 is TheYearOfAI.
But for Louis Mendelsohn, the proliferation of AI isn’t news. In fact he’s been using AI in his trading for over 30 years. That alone makes him something of a legend on Wall Street.
Mendelsohn started his first company, Market Technologies, in 1979, and in 1991 launched VantagePoint, the first commercially successful trading software that utilized AI. Today VantagePoint has about 60 employees and customers in over 120 countries.
We wanted to know more about the expansion of AI in financial services, so Benzinga spoke with Mendelsohn about his company, what makes it different, and how AI has changed in his lifetime. The interview has been edited for length and clarity.
Benzinga: You were one of the first people, if not the first, to connect AI to the financial markets. How exactly did that come about?
Louis Mendelsohn: Well, I was a trader. I got involved in the commodity and futures areas, which was my main focus prior to developing the software. My interest was taking what was already being done in the 70’s by institutional investors and bring those abilities down to the PC level. Rather than look at the PC as a glorified calculator, I wanted to introduce a disruptive capability to technical analysis.
Benzinga: So how did that start?
Mendelsohn: We started in 1979. My interest was to develop the software for my own personal use, but I quickly realized the cost of software development could be shared. I started creating a following for myself among traders around the world, and that kind of is what led to the creation of the company. I introduced the capability known as strategy backtesting in PC software in 1983.
At that point in time, of course, everything was focused on single market analysis—looking at one market, whatever market that is, and looking at past price data, past volume data. That set off pretty much an industry focused around backtesting. I had numerous customers of mine who were traders themselves and after using my software they went on to become competitors of mine. There were others in the industry that followed suit.
Benzinga: So how does artificial intelligence come into play in all of this?
Mendelsohn: The crash of 1987 was kind of a milestone in my mind. To me it represented the first instance of what I would consider to be a global market crash. It was really the first clear indication that the markets were becoming more integrated and globalized. I realized at that point that continuing to focus on individual markets one at a time was becoming obsolete. So I began developing what I called intermarket analysis software.
In 1988 I introduced a program that implemented that in a very primitive way. I really started to delve into other areas of math, and I was fortunate in that I had gone to Carnegie Mellon University in the 1960’s. Interestingly enough artificial intelligence isn’t a new concept. It was taught at Carnegie Mellon in the 60’s. There were some people there that were leaders in application of AI and decision making at the time. I began looking into AI and starting to research it, and realized it really was the right tool that I was looking for that could successfully implement an intermarket analysis approach. The specific area I was working in is known as neural networks.
Benzinga: So you started Market Technologies in 1979, and then VantagePoint as a subsidiary in 1991. What was the biggest challenge of using AI back then?
Mendelsohn: Well the problem that I had at the time was that the computing power that would be necessary to do the trading and allow the AI to learn just wasn't there at the PC level. So I rented time on minicomputers, rented time on accelerator boards, in order to beef up the computing power.
AI is an old concept. It’s been around for decades. It’s just that the computing power wasn't there. Particularly not at the level of the individual PC.
Benzinga: Obviously you’ve seen how AI has become a hotter and hotter topic over the last few years. Has this helped or hurt your business as more competitors have sprouted up?
Mendelsohn: It helps. When I left hospital administration I had a wife and a newborn son. Everybody including my father-in-law said ‘You’re crazy.’ His argument was ‘Nobody has computers.’ He didn’t see the vision.
We’ve been operating for the last 25 years with headwinds affecting our company. Why? Because almost no one had any knowledge of, interest in, or didn’t even hear of the word AI.
Now, recently, as the words AI and artificial intelligence and deep learning have started to infuse themselves into the mass media, that’s beginning to shift headwinds to tailwinds. We’re still early in the adoption of the technology and embracing of AI by the public. As far as I’m concerned no there hasn’t been any negative impact of the recent hoopla.
Benzinga: Aside from computing power, what’s the biggest difference between 1991 and now in terms of AI?
Mendelsohn: Now, our prospective customers have heard of AI. AI is going to disrupt probably every industry that exists. It won’t happen today, it won’t happen necessarily tomorrow, it may not even be for 10 years. And many of the companies that are getting into it, many of them will go by the wayside because they don’t know what they’re doing.
In 1991 it was just not even being applied in this arena by anyone other than myself. I didn’t go into it for the purposes of coming up with some kind of buzzword from a marketing perspective. We’ve been doing this for decades. Nowadays people are just jumping on the bandwagon. Everyone’s got AI. I’m sure that if you look into what people have it’s probably more similar to what I was doing in the late 80’s when I was first getting my introduction of how AI could be applied to this area.
So you have a lot of people that are jumping on the bandwagon, they see it, they hear it, it’s a buzzword, they think it can be something that can help them get exposure to the market and they don’t really know what they’re doing. That’s happened a lot over the last 30 years.
Benzinga: You’ve got a lot of competitors now. What do you say to prospective customers to sell them on VantagePoint as a trading solution?
Mendelsohn: We have absolutely been able to demonstrate that our software can make short term forecasts of the financial markets with up to 86% predictive accuracy. That’s real. That’s where we’re at. But it’s taken us 30 years to accomplish that. If you can’t forecast with a level of accuracy with at least 70%, we just don't even want to bother with that market at this point in time.
We’ve been at this for decades. This isn’t new. We’re not the new kid on the block. We've been doing this for so long. We've gone down all the rabbit holes. We’ve hit all the roadblocks. That was decades ago. Now for the last 10 years we've simply automated the process.
We have been refining Vantage Point since it first came out in 1991. We don't need to go out and reinvent the wheel. We have a really well-polished software program at this point. What we're doing is continuing to tweak it continuing to make it more user friendly.
Benzinga: Who’s your target audience at this point?
Mendelsohn: We have no interest in dealing with institutional traders. The institutions, they’ve gotten on the bandwagon, they have the resources to do this kind of work. They don’t need Market Technologies. But the individual traders do. The people who don’t have the ability to build this technology on their own, they’re the ones that can most benefit. That's our bread and butter and that’s what we do.