“2014 may be viewed as the year of the machines, or the start of this phase of machines.” – Dave Sanderson (Tweet)
In this year-end-review, we take a look at how 2014 might be seen many years into the future. Is this when “trader-less” systematic trading caught on? We take you into the workings of a 1-year-old trading technology company, how they succeeded, and where they are going in the future.
Thanks for listening and please welcome back Dave Sanderson.
In This Episode, You’ll Learn:
- It was their first year of trading under the public eye.
- How consistent the statistics were the entire year.
- How the backtest matched up with live data in 2014.
- What led to their profits and where they lost money.
- Why it’s hard to draw conclusions that tie trades back to world events.
- How they made a lot of money in the first half of the year when other trend followers were having a difficult time.
“I would rather be in the hot sector with a good story, than a standout performer in a sector that people are trying to allocate away from.” – Dave Sanderson (Tweet)
- About their one-year birthday party.
- The improvements they made during the year.
- How 2015 will see a “refined” interest in trend following CTAs.
Resources & Links Mentioned in this Episode:
This episode was sponsored by Swiss Financial Services:
Connect with KFL Capital Management:
Visit the Website: www.kflcapital.com
Call KFL Capital Management: +1 (416) 849-1925 x212
E-Mail KFL Capital Management: Info@KFLCapital.com
Follow KFL Capital Management on Linkedin
Dave: The driverless car, the traderless fund, I guess is the analogy. It is coming and whether it's a million miles on a Google car or whether it's the Jeopardy contest. It's now three years - February 14th since Watson beat Ken Jennings and Brad Rutter in that contest. Things are happening and you never quite know when you're in the midst of it that it's happening, or when the turning point comes, But let's keep an eye out, because perhaps 2014 as we look back in the great span of time, was when all this started to happen, at least in respect of the hedge fund industry.
Hi this is Dave Sanderson, Founder and CEO of KFL Capital Management, and you're listening to my year in review on Top Traders Unplugged.
Introduction: Imagine spending an hour with the world's greatest traders. Imagine learning from their experiences their successes and their failures, imagine no more. Welcome to Top Traders Unplugged. The place where you can learn from the best hedge fund managers in the world, so you can take your manager due diligence, or investment career to the next level. Here's your host, veteran hedge fund manager, Niels Kaastrup-Larsen.
Niels: Welcome back Dave, for this review of 2014 where we look at the big events from the point of view of your trading strategy. Now I want to explore the ups and the downs, as well as the big takeaway, from what can only be described as a great year for systematic trading strategies, in general. But as we know, just because you're systematic in your trading, it doesn't mean that necessarily your strategy will deal with the market events in the same way. And clearly in your case it probably won't, because you do things very differently to many of my previous guests. So let's just jump right into it. Tell me about 2014 from your perspective. How did the year evolve both from the strategies point of view but also from your firm's point of view?
Dave: Well as you know Niels, this was our first year of trading under the public eye. So it's not if I can compare this year with many other years, but it certainly was a year where it was a great test for the strategy. Now I say that, and since you and I talked last time, it's really caused me to think about the broader perspective of systematic traders. And you're so right to characterize them as having a great year. You asked me a good question last time about the books I was reading. It strikes me that 2014 may be viewed as the year of the machines or the start of this phase of the machines. I think about the innovators, and the second machine age and Jim Mellon just wrote a book called Fast Forward, which I think is fabulous, and Zero To One. And then we see who's really done well in 2014, and it is, as you say, the systematic traders, and more interesting to us those who use fairly complex machine learning algorithms to trade these markets. So it's a really… I think you know, it's nice to have a deep science underneath the firm, but it's nice to get lucky too. And I think we're lucky in the sense that we're in a phase where people are going to be talking about this kind of trading. Folks like Two Sigma, with their returns last year being off the charts are going to do us all a huge favor. So, sorry for rambling… what was the original question?
Niels: No, I was just… Well, its fine Dave, and I appreciate that. Yeah… no, I was sort of curious to a little bit about how your strategy dealt with some of the events, really just from a thirty thousand foot point of view at this stage. Despite it being the first year, it's obviously a very important year. So was there anything that you felt when you look back, and you look back at the numbers that sort of surprised you, in general? Or was it really as expected?
Dave: So, what surprised us was the consistency of the statistics after the entire year was over. When you peel it back, and we're always very keen to see what's happening in the portfolio and what did we expect to happen. So as you know, our backtest is five years long across fifteen different assets. So we have twelve thousand… more than that actually trades in our backtest. We're always judging the live record in reference to that backtest. And it was incredibly close, and that's very comforting to us.
So, there were a lot of economic events this past year as there seems to always be. You can never quite anticipate them at the beginning of the year, but despite all of that, the core statistics we look at were incredibly similar to our backtest and in fact a little bit more positive. So for example, we had sixty-seven percent of our days were positive days, and seventy-five percent of our months were positive, and that's very consistent with the backtest. Also, the key metric that we use, that I shared with you last time - this idea of a coin toss ratio: this combined accuracy number. Our backtest was fifty-four percent, and then this year of trading was fifty-four percent, so we love that.
The Sharp was two and a half, which is a little higher than our backtest. And the Sortino was 3.3, which again is a little higher than our backtest. So in general we step back, and we say, "Okay, regardless of all the exogenous events, as long as we get enough trades, and in 2014, it was eighteen hundred and twenty-six trades." As long as we get enough trades, our edge should reveal itself, and it did, so that's very comforting.
Niels: Sure, and I guess that's actually really important, especially when people want to understand a strategy in a firm like you, that it really is the statistics that matter, and dealing with different events, some of them very big events, but actually coming out with numbers that are so consistent with what you expect, I can imagine that that must be very pleasing. Now I do want to ask you about, even though you may not look at this as being important… I do want to ask you a little bit about which markets were the biggest contributors, both on the positive side, but also on the negative side? If you remember.
Dave: Sure, yeah I do, in fact. I can share with you exactly what those were. And you know, you make a great point there about the statistics and how we focus on them. But I must make an admission to you, as I look at the portfolio too often; I still have those human reactions that the scientists don't have. Our scientific team doesn't look at the portfolio very often. It's the portfolio management team that does, and unfortunately we react like most humans do. When we're up it feels good, and when we're down it feels five times as much pain as when we're up. So our leading markets this year were the NASDAQ and the S&P 500. By far they were seventy percent of net profits let's say. But right behind that was oil, and gold, and soy oil, copper, and so on. The losing markets were corn, silver, and soy meal. So again it was, you know, we trade twelve assets at this point. Nine of them were positive, and three of them were negative.
Niels: Yeah, and when you look at the statistics, do you drill down to the level of saying, actually we would have expected to do a little bit differently in this market, given what happened? Or as long as the overall statistics are in line, you're okay with that?
Dave: Yeah, well, what we look at is the coin toss ratio for each market. And I can tell you that the range for the year, at the high end was 63.4%, that was the S&P 500. And at the low end, there were two underneath fifty, and that's what we get concerned about. If something has a coin toss ratio over a significant number of trades under fifty percent, then we have to analyze why that's happening. So there were two under fifty, one of them was silver, and one of them was soy, one of the soy contracts. We trade three. In those cases, the only sort of rationalization for us (and I hope it's not a rationalization) is there just weren't enough trades to come into that fifty-four percent coin toss ratio.
So in the case of one of the soy's, it was two hundred and six trades, and in the case of silver, it was one hundred and seventeen. So, we won't make any changes based on that. We fully expect, over a greater sample size, that those are going to evolve to the fifty-four. You can sort of see as you aggregate all those trades to eighteen hundred, that the coin toss ratio becomes very, very close to that fifty-four percent.
Niels: And remind me Dave, is it one thousand trades that you ideally need in order to say, "right, as long as we're at fifty-four over one thousand trades," or what is the number that these numbers you mentioned just before, has to relate to?
Dave: I should know that answer off the top of my head Niels. I think the reason you picked one thousand is that when we went to validate the technology, we accumulated one thousand trades. Now that was just sort of the investment guys saying, "Please get us one thousand trades because the industry will leave us." So I don't think that is the magic number, and I'd have to get back to you.
Niels: No, that's fine, that's fine. I would rather people call you and find out the answer, probably gives better dialog. Now, there were obviously some events, some themes that played a big role. Obviously, Ukraine, Russia, oil, at certain times during the year. Were there any periods during the year where you saw either above or below average returns that you could influence or tie back to some of these events? Or was it really a year as… I would almost say as usual for you guys?
Dave: Yeah, you know it's difficult for us to do that because of the underlying scientific rigor we have for the whole system. So if I was to say to you for example, just me looking at the results, I'd say, "Gee, we made a lot of money in oil in the second half of the year." You know, that would not surprise anybody, I suppose, from July and August on there was a lot of volatility. And also, a lot of folks who trend trade, that was a perfect trade for them. But the scientific team would be upset with me if I made that connection because really you can't make a robust statistical inference from it. So we have this sort of yin and yang around our place where we sometimes slip into the vernacular of the industry. Then I'm told by the science guys, "You shouldn't say that because that's not sustainable in a way that we have statistical confidence." So as I look at the sheet in front of me, I can tell you that we made lots of money in oil in the second half of the year. We made lots of money in the equities virtually all throughout the year, but it's very hard to draw conclusions that are connected to the events of the day.
Niels: Sure, but what strikes me actually, when I look at the returns (that obviously I was provided before we were having this conversation) was that you had a very, very strong first part of the year. In fact, the first five months were incredibly strong, and actually most trend followers were struggling. Trend followers made money in the second half. You really seemed to do very well in the first few months. Is there anything in that that makes a good conversation point, so to speak?
Dave: Yeah, it's a great point. Some of our investors are wondering the same thing. Particularly whereas you and I chatted last time, we would suspect that higher volatility is good for us because if you make money on fifty percent of your trades, then in higher vol markets you're going to make a lot more money. Well, the second half of the year was much higher vol than the first half of the year, and yet we didn't make nearly as much money. So if I was able to say… You know, if we were having this conversation twenty years from now, we could make some statistical conclusions about that. But I'm sorry I can't, although I must say that I'm scratching my head about it, saying yeah, just that's odd.
Niels: Sure, but you know, that's the way it is. That's interestingly enough…
Dave: Sorry, I do like the fact that, as you pointed out, there appears to be an absence of correlation between us and other CTA's, or if you want to get more specific, us and other systematic traders, or us and trend followers. I think that's a nice thing to point out to people who are allocating capital. You can make an allocation in this space, and yet not have the same return pattern as others you've allocated to.
Niels: Yeah, well that's very true. When you look back at 2014, what's the highlight for you? On all levels, really… whether it's the firm as a whole or the strategy as a whole… What makes you smile when you look back at 2014?
Dave: I think our one year birthday party is the point at which I kind of felt it all happening in front of me. So, it was a one year birthday party that occurred in December, because at the end of November we had twelve months of trading in our public vehicle, I call it. So we had a little party in Toronto, and Bartt Kellermann was there from the Battle of the Quants, and folks from Jeffery's came up. So it was just a room full of people that were supporters and watchers of us, and then our chief scientist, Garry Lee, gave an incredible… not a speech, but just a talk.
And it was very apparent to everybody in the room that this was something deeper, more profound, more meaningful perhaps than the hedge fund business venire that's on this thing. It was very much the case that Gary was sharing his gratitude about the ability to work on such a difficult project. And he even talked about the support of his wife, because here's a guy on the prime of his career that was being asked by Dr. Wong, the founder of The Pattern Analysis Machine Intelligence Lab, five years ago, to spend his precious career on a data set that had confounded most people.
So he was nervous about going down this road, and after now, twenty-eight hundred live trades, he certainly feels like he's made some really profound breakthroughs. So it was great to see it from that perspective, and there were a couple people who at one point in the talk, Gary choked up. It was apparent to him that it was very emotional, and you could hear a pin drop in that place, Niels. It was not a normal situation where people are checking their blackberries, and distracted because here's his sort of "Quant guy" and he's going to talk about numbers. Here was the Quant guy, and he had everybody fighting back some tears. It was pretty cool.
Niels: Sure, sounds like a very important moment for you guys. Did you make any new findings? If I can put it that way… during the year, or has the model pretty much been fixed in stone throughout 2014? And are you looking at something exciting for the beginning of '15? Maybe new markets to add, or how does that side look?
Dave: Yeah, there were a couple of very interesting and exciting movements ahead in the predictor. Not so much in terms of the machine learning and pattern recognition, but, for example, there were a couple of pre-processing improvements that we did. The historical data from Thomson Reuters has… You know, no matter what provider you get it from needs a lot of attention, of course. Anyway, there were a couple of breakthroughs in that category, and then very recently there was a breakthrough in terms of the independent variables that we feed the model. So I think I shared with you last time, we use the target asset and 49 independent variables. Well, we've actually reduced that just recently by four variables and substituted some variables and found that it adds almost 1% to the coin toss ratio, so that is a very big… step movement in our program. It's something that we knew held some opportunity, but with all the things the science team had to deal with, it was only in 2014 that we were able to get to that list and make some principle based improvements to it.
Niels: Yeah. What about from the business side? With systematic traders in general doing better, with you guys coming out with a clearly different strategy, but delivering very solid returns, did the conversations change towards the end of the year, did you feel? Or have you not sort of seen that yet maybe happening? Or how do you feel that the attitude of investors in general towards you and the space is at the moment?
Dave: I did feel it change. It was almost the first half, the second half. So in the first half I would say we were in a segment of the market that was unloved, and not doing fantastic on a performance basis. So our most interesting attribute was the ability to perform in that sector, so, that's one kind of life to live. Or, conversely, would you rather be in a sector that's hot, and you're a good story in a sector that's hot, but you performance doesn't stand out as much as it did. That was the second half of the year, and I would actually prefer the second half. I would rather be in the hot sector with a good story, than a stand-out performer in a sector that people are trying to allocate away from.
Niels: Yeah, no that makes perfect sense. And speaking about that, you know, clearly it is likely that there could be a renewed interest and hopefully some new flows into the space after this 2014... I know you didn't trade money at the time, but obviously were involved in the industry. And you know that subsequent to 2008 there was a big inflow in the industry, and then there was a big outflow because people got disappointed.
Now from your point of view, and clearly you obviously want to take in as many assets as you can. But the quality of the investors that you're looking for, knowing full well that it's not healthy for anyone to get investors in quickly, only to see them leave very early on. How do you expect to try and deal with that? Because there could be some hot money flowing to this sector. You guys might be well in line for that given the fact you're different; given the fact that you had a great year; given the fact that you performed so well at times where other people didn't during the year. Do you have any thoughts about how you're going to try and avoid the hot money in and out?
Dave: Yes, so first of all it'll be a luxurious problem to have, but I do think it's possible. You called it a renewed interest, I think? I would add another word and say it's also going to be a refined interest. By that I mean the first sort of time around that artificial intelligence was talked about, it got hyped up, and then people got disappointed, and that's the normal curve in technology. And this is at least the second go around of that, so I think the interest now is becoming more refined. You can't just say, "Oh, well we're an artificial intelligence firm." People come back with, "Be more specific for me. What are you doing?" So I think the investors are more refined, and I don’t want to say educated because that gives the wrong idea. But they're questions are better, they're crisper, and so perhaps that's going to provide a different kind of investor that stays around for the longer term. But you raised a great point, we have been counseled by firms in this industry that have done very well over the years, they're very giving of their time, and their guidance. The one thing they keep saying to us is don't make that mistake, don't grow too fast. So we want to be in this business for a very long time. Hopefully we'll have the discipline if the challenge does arise, we'll have the discipline to do it properly.
Niels: Sure, absolutely, I appreciate that. Now I only have one sort of last question left, because it is a short review this time, but is there anything that you would like to bring up? And just maybe… it could be a reminder; it could be something that just stood out for you, or just something that is important to you that you want to share with the audience at this stage?
Dave: Yes, and you made me think about it last time with those two interviews, and that is… I believe we are in the second age of the machine, and that there are things happening now that are expressing themselves definitely in the CTA space. Folks who are in the algorithmic based prediction, systematic trading category, are doing some wonderful things. I want to be... and we at KFL want to be part of that. I don't know if it's a resurgence or you know, for the first time… But we want to be part of it. I would say that anybody who's looking around this space, perhaps hold the idea that this is a very interesting time where some very interesting firms are being built. Some technologies are going to do some things that are in the great span of time, seem to be perhaps a turning point. It maybe sounds too dramatic, but there's a lot of really interesting things going on that are expressing themselves in the CTA space. I would encourage everybody when they hear a good story from what's called loosely a scientific or computer based trading firm, to just ask a few more questions and have an open mind. We've met some great people, and it's in all our interests for this sector or this industry to rise to its appropriate place.
Niels: You know I think that's a good point. The other point, just off the cuff, which I wanted to share since... Clearly we know that systematic traders, in general, we never get a warm reception from people because it is a little against their nature for whatever reason that we use computers, even though, we don't… certainly, for the most part, only use them to implement trades. Now you take that even further in your business because you have the computers make the decisions without human intervention as well. But I just wonder, I wonder if 2015 is the year where in California they allow self-driving cars to be freely driven, instead of just the test cars that are running around… and I just wonder what the psychological effect will be once we start seeing cars driving around without a human, whether that will slowly change, hopefully quickly change, people's perception about what we do.
The fact that you don't have to fear technology, you don't have to fear the computer when it comes to investing because we know that people embrace it in other businesses. You know like flying a plane in pitch dark from the US to Europe, you don't want to do that without and autopilot and a computer to guide you. But as soon as it comes to investing, it becomes slightly harder to convince people. So I just wonder, whether we need to get to that point where we see it in our day to day environment and we feel, "Okay, well, if they can drive a car with a computer, maybe they can even pick stocks and futures, I don't know!"
Dave: You know, you're reading my mind. That's a great point. The driverless car, the traderless fund I guess, is the analogy, and Jim Mellon does a great job in his book, Fast Forwarding. He points out that there is a chip coming out with 5.4 billion transistors on it, on a single chip. It's called True North. That is just incredible, so your point is should we fear it? Or what's going to make it more mainstream perhaps? I think you're right, there is a general fear of it with flash crashes, and even… I'm still in conversations where people talk about long term capital management, so there is a general fear out there, and that's a legitimate fear. So it is coming, and whether it's a million miles on a Google car or whether it's the Watson… You know the Jeopardy contest. It's now three years, February 14th, since Watson beat Ken Jennings, and Brad Rutter in that contest. Or even in 2014, we had… Although it's a little bit controversial, the first touring test where a computer fooled greater than thirty-three percent of the audience that it was a human being. Things are happening, and you never quite know when you're in the midst of it that it's happening or when the turning point comes. But, as I was trying to say, but you said more articulately, let's keep an eye out. Because perhaps 2014 as we look back in the great span of time, was when all of this started to happen, at least in respect to the hedge fund industry.
Niels: Sure, absolutely. Dave, unfortunately our time today for this short period is up. But of course for those who want to hear a much longer conversation between ourselves, they can go back and listen to those episodes on Top Traders Unplugged. I do want to thank you though, again, for being on the podcast, for sharing your insights. I certainly want to congratulate you on a very solid year, and your first full year, and I want to wish you and your partners the very best for the coming year, and I look forward to reconnecting later in 2015!
Dave: Well thanks, Niels, and thanks for everything you're doing! You're doing a great service to the industry, and I know it's your passion, so it's not work for you but, it's a great thing you're doing.
Niels: Thank you so much. All the best and take care Dave!
Dave: Bye bye.
Ending: Thanks for listening to Top Traders Unplugged. If you feel you learned something of value from today's episode, the best way to stay updated is to go on over to iTunes and subscribe to the show so that you'll be sure to get all the new episodes as they're released. We have some amazing guests lined up for you, and to ensure our show continues to grow, please leave us an honest rating and review on iTunes. It only takes a minute, and it's the best way to show us you love the podcast. We'll see you next time on Top Traders Unplugged.
Become An Insider
Subscribe for free and be the first to receive new and exclusive interviews with the world's top traders. As an insider we'll also send insightful bonus content direct to your inbox.Free Instant Access »
You might also like:
Date posted: 23 Jan 2015no comments