Niels
Welcome to another episode of Top Traders Unplugged, and thanks so much for tuning in today. I know how valuable your time is, so I appreciate you spending some of it with me here today. Now, on today's show I'm talking to Peter Kambolin, Co-founder and CEO of Systematic Alpha Management. His firm has recently received the 2014 Pinnacle Award for Best Diversified CTA. This comes as their futures program celebrates its 10th anniversary, but it has not been an easy path for Peter and his team. In 2011, following a drawdown, they lost 90% of their assets due to investor redemptions. Having come back from this drawdown, Peter feels that today their firm is stronger than ever, and he shares fascinating insights to a CTA strategy that is completely different to the traditional trend based managers.
For those who are new to the show, I just want to let you know that you can find all of the show notes, including a full transcript of today's episode on the TOPTRADERSUNPLUGGED.COM website. Now let's get started with Part 1 of my conversation. I hope you will enjoy it.
Peter, thank you so much for being with us today. I really appreciate it.
Peter
Sure.
Niels
Now what's quite interesting about your firm is that you started back in 2004, when not many CTAs were doing short term trading. What's quite funny is that my previous guest, Karsten Schroeder, from Amplitude, also founded his firm in 2004. So, let me start by one, congratulating you on your 10th anniversary which I believe was at the end of last month, and perhaps start by asking you what was so special about that time, when clearly some firms, including yourself, were beginning to look outside the classical medium to long term trend following trading style?
Peter
Yes, well actually, our track record goes even beyond 2004. We started managing individually managed accounts in 2001, and in 2004 we decided to consolidate these small individual managed accounts we had, and we launched a fund. So, before the fund we do have an audited track record that goes back to 2001. The roots of our firm come...we have to go back to the late 1990s. In 1998 I established a broker dealer in New York City. It was an introducing broker dealer that was trading equities. It had nothing to do with futures at all. At that time markets were good, most stocks were going up, and we were basically trading stocks for individual investors. Then in the year 2000, the internet bubble burst, and NASDAQ had a major collapse, and that was the time when we wanted to come up with a strategy that would perform well in any market environment. We did not want to rely on the direction of the equity markets, and that was the period when I met my current partner Alexei Chekhlov, was in the year 2000. He, by that time, already worked for a long term trend following CTA, one of the original ones called TrendLogic Associates, and he also worked for a couple of other firms - a big hedge fund out of Connecticut, and a big French bank. So he had all the experience building the models, and I had the initial setup infrastructure and initial clientele. So it took us about a year, from 2000 to 2001, to develop the strategies and we started offering these strategies to our individual investors at that time.
The foundation of our current program, which is a market mutual program, comes from a period when markets were going down, and we had to have a strategy that would perform well in any market environment. So some of the ideas that people applied to long and short equity trading, to set statistical arbitrage trading, to spread trading in fixed income markets for example, we applied to major global equity indices and currency futures, and this is the main part of what we do. As I said, from 2001 through 2004 we were managing these small individual accounts, polishing our strategy, and in 2004 we launched our fund which recently celebrated 10 years of traffic. So that is our background, if we go back to the late 1990s, early 2000.
Niels
Sure, and as you point out, 10 years latter your firm has recently been awarded, by CTA Intelligence with its US performance award for 2014, which must be a great way to celebrate this milestone that you have just passed. Let me also ask you, what does it mean for you to win such an award, and also what does it mean for your business development, do you think?
Peter
Well most recently we actually won another award, which is even more prestigious, I would say, which is a Pinnacle Award, as the best ever certified CTA for 2013. This is our 4th major award. In 2009 and in 2012 we won similar awards from HFMWeek. I would say the current environment for CTAs, for futures traders in general, is quite difficult to raise capital. Over the last 3, 4, 5, years, most of the capital that was raised was going to big multi-billion dollar firms. So, on one hand it's very nice to win all these awards and it gives us a great name recognition, on the other hand we also understand that getting a client requires a lot more than just a good award. We need to continue doing a good job for the clients from the performance side. We have to explain our story better, because what we offer is something very unique, uncorrelated, in fact our correlation to CTAs in general is close to zero, which means that what we're doing is very, very different from traditional CTAs, and sometimes explaining it to investors, and even some professional investors that are supposed to understand and recognize what we do, is sometimes the challenge. Awards are nice, but that's not the only thing that we need to do to build the business.
Niels
No, that is true, and actually, in a sense, that's exactly why we are having our conversation today, because what I found over the years is that you're right, it is a difficult strategy for people to comprehend, and I think by documenting it this way, where people can go back and listen to you telling the story, you explaining the strategy many times is needed, really is why hopefully people will get a much better grasp of what you are doing after listening to our podcast. Before we go into the details about the company today, I'd like you to go even further back. I know you just touch upon it a little bit before in the introduction about roughly when you met your partner and so on and so forth, but I'd actually like to go back even further, and if you would share a little bit about your background from even before you started your introducing broker, just for people to get a feel for you as a person, and your partner as a person, and then how it evolved. Because at the end of the day, although we are talking about systematic trading, systems, and models, it's all based on individuals and people in the background, so I think it's important that we try to give people as much as we can in that respect as well, so if you wouldn't mind, just take us back to where it all started and how you ended up in the systematic trading world which is not where most people end up now a days.
Peter
Well, Alexei and I, we come from the same city, which is Moscow, Russia. I came to New York on September 11, 1989. That was such a special day to come to work. Alexei came, I believe, a year later in 1990. When I came I was a kid. I went to high school in New York, it was a public school, then I went to more or less a public college, so my whole education probably cost me less than $10,000, which is next to nothing. Then after school I graduated with a degree in finance, magna cum laude - I was one of the best in the class. The truth of the matter was that it was very hard to find a job. That was 1995, and there were very few financial jobs available. The very first job I took was with a broker dealer. I passed Series 7, which is the license someone needs to have to work in a broker dealer environment, and I just wanted to be close to the financial world, even though that job did not pay any money. I was getting $250 per week for working there. After doing it for about four months, and passing the Series 7, I decided to leave the broker dealer and I was looking for other opportunities, and it was very tough to get a job and later I ended up in a firm that was a small boutique firm that did a lot of private placements. It worked with small cap stocks, raising capital, and it was a small firm, as I said, maybe 10 people all together, and I introduced to the owner some of the high net worth Russians that I knew at that time. It was funny, after my former boss made some nice money off of my introductions and he chose not to pay me a bonus, as a result of that, that's when I realized that maybe I should go and start my own firm if I had these connections to these high network individuals, and if I can't make money even in a small boutique firm, why don't I do it on my own. That's how I started the broker dealer. At that time I was twenty-one years old. I was one of the youngest, if not the youngest owner of the district 10 Manhattan broker dealer in New York. As I explained before, from 1998 through 2000, things were going well. We were making good money for our clients. They gave us digression over their accounts, and then the market started to collapse, and we needed to have a absolute return oriented strategy.
Niels
So these were equity investments predominantly, at the time?
Peter
Yeah, we were just trading stocks, nothing to do with futures at all. Alexei 's bedrock is very different. He is the quant ( quantitative analyst) at our firm. He has PHD from Princeton University. He worked in academia for a couple of years, he wrote a number of articles - scientific articles and finance related articles that are very well sited in the media. He had a different trajectory. He worked at big firms. He had a lot of experience, so, I would say, if you look at us today, we have, on one hand, a somewhat similar background and on the other hand our personalities are very different. He is the quant, I am the common sense guy. I'm the business guy, he is the brains and blood of the company. So I think our tandem together is very good. I remember one investor told me once that if both of us were professors he would never allocate to us. Sometimes it makes sense to have a common sense guy next to a PHD guy. I think we have that combination. We've been through a lot over the last 14 years we have been together - good and bad times, and we know each other well. We know each other's weaknesses and strengths and we're trying to make sure that everyone is doing his job, and of course, at the end of the day, we have to have a common decision on any things that we do, so we, both of us, have to basically approve or disapprove certain decisions. Some research related decisions driven by Alexei , and some business related decisions introduced by me. This is how we work, and that's where we come from.
Niels
Now, you mentioned that Alexei was with TrendLogic, and I remember the name, but I have to admit I don't remember the specifics. The name suggests that it was some kind of trend based strategy.
Peter
Right. If we go back to the 1970s,they were one of the original trend following CTAs and Alexei started working with them, I believe in 1996, 1997, and they hired him (even at that time) to diversify and build some of the strategies that are not necessarily trend following strategies.
Niels
OK, because that was my question exactly, then the transition from being with a trend based firm, for you then to go and look at short term trading, so that explains the direction that you took even at that time.
Peter
Yeah, it was a combination of things. Alexei worked with them on some mean reverting strategies, number one, at the Bank Paribas where he also worked, he was doing fixed income spread trading which somewhat resembles what we do, but we do it via the equity index and currency futures. We end up trading fixed income markets, but the idea is similar - we would go long and short, in highly, highly related markets, expecting a reversal with some kind of a mean of the spread. So these ideas came from how Alexei 's experience at the French bank, and, of course, as I mentioned before, we wanted to have a strategy that will have no dependence on the direction of the equity markets. Our data to S&P, to other hedge funds, and to CTAs is what's interesting, is close to zero.
Niels
Now, obviously we'll go into much more details, but I just wanted to ask, again, one of these sort of introductory questions before we launch into that. Obviously running Systematic Alpha Management today, is a big part of your life, but when you are not working, what do you like doing Peter?
Peter
Well, I live in two cities. I live in New York and that's where I work, and I also live in Miami, that's where my family lives, so every weekend I go to Miami, more or less, unless I travel for business. I think it helps a lot in my business endeavors as well, because when you are away from the city, when you are away from the office, and when you walk on the beach, sometimes some of the greatest ideas can come unexpectedly. You might play with your kids and all of a sudden you're realizing, "why don't I do this, or that?" So this has been my lifestyle for the last 8 or 9 years. I work in New York and I live in Miami.
Niels
Yeah, that's very interesting...that's excellent.
Peter
Of course I love sports. I love to ski. I love to play golf. I love to play tennis. I like to go out with my friends. In Miami there's some very interesting people now-a-days living there. Most of my friends are Russian for reasons, just easier for me to communicate I guess with them or to understand each other. There are some very wealthy Russians in Miami now-a-days, but again, for whatever reason, I never talk business with them. I don't try to recruit them or get their money under management. I prefer to stay just friends. I did have one experience with one of my friends. He allocated a million dollars and six months later his return was flat and he was complaining big time, "why don't I make money for him?" I told him, "look, just take your money back and let's just continue being friends."
Niels
That's probably a good idea. Now launching a little bit into the business side of things, maybe you could start by just giving a brief overview of the programs that you run today, when they started, and how the assets under management in each program, where they stand today?
Peter
So, at the moment we're running two programs: our original Systematic Alpha Futures program was started, as I mentioned before, in late 2001 and the fund started in 2004, and that is the program that won four awards, two from HFMWeek, one from CTA Intelligence and one, most recently, from the Pinnacle Award. That program...the evolution of AUM was like that in 2001, more or less, we started with very little assets, maybe 3, 4, 5 million under management. When we started the fund in 2004, we ended the year, I think at around 5 million. So our assets from 2001 through 2004 ranged from 5 to 10 million dollars. and then in 2005, we already had over 20 million in the fund. In 2006 the number grew to 80 million and then we were more or less doubling our assets under management and we peaked in February of 2011 at over 700 million dollars in the program. In 2011 we were one of the largest truly alternative CTAs, not the largest, but...you know...QIM was there with a lot of assets at that time, and three or four more firms, but I would say that we were in the top 10 in terms of AUM. 2011 was a tough year for us. It was the only negative year on our track record. In fact we had six consecutive negative months, starting in March of 2011 through August of 2011, and most of the capital that we had was in direct managed accounts - well over 500 million actually, was in direct managed accounts that had daily liquidity, and most of the clients redeemed. So our assets went from 721 million at the highest point, went down to as low as 50 million a year later. The low of AUM was in March or April of 2012. Since that time we have been growing assets, but not at the rate that I would have expected given our performance over the last two years. In 2012 and 2013, we generated very solid returns, and we were ranked among the...you know...very high compared to all CTAs. So, some of the capital is coming back, some of the former clients came back, which is a good sign, but very slowly and hesitantly, I would say. So, at the moment we are running a total of one hundred million under management, and we hope to grow it in the coming months with our new project - maybe we can talk about it later, we are opening a UCITS fund soon with investors. In 2011 we introduced the second program that we have, which is called Systematic Alpha Multi-Strategy fund or program. That program came about after the experiences that we had in 2011. We wanted to come up with a program that would have more diversification and this program has a 50/50 split allocation to our...50% is allocated to our original spread trading program, and the other 50% is allocated to directional, you can call it short term trend following or momentum models. These models are a lot more similar in style to what most other CTAs are trading - directional positions, expecting a continuation of a price move, up and down. So the model strategy program was introduced in 2011, and at the moment we have about 30 - 35 million in it under management. Most of the capital in that program that we have is via Deutsche Bank relationship with our on the DB Select platform, both of our programs are available there, and we have a number of different clients in the rest via Deutsche Bank.
Niels
Great stuff, great stuff. Now before we jump to the first real topic, I just wanted to ask one thing, so I think for the purpose of this conversation we're going to be focusing on the original futures program, but in terms of overall objective for that program, are you focusing on a particular environment? You mentioned that the experience came about from the equity market going down, but the futures program, is that designed to make money in all conditions, or are there certain environments that you in particular are looking to make money in that program?
Peter
Well of course, certain environments are better than others. For example, in 2008 the environment for the program was perfect. That was a time when equity markets were collapsing on the one hand, on the other hand correlation between the equity markets was very strong. Everything was going down together and our market mutual long/short spreads did perform extremely well during that market environment. The reason for the difficulty in 2011, for us, was related to the fact that European equity indices were going down a lot faster, with a lot greater volatility, compared to the US markets, because of the European debt crisis. This type of an environment was a very challenging environment for us. I would say, similarly in 2013, we made money when the markets were going up, so from the standpoint of markets going up or down, I would say our correlational data is close to zero. What we are not immune to is the volatilities, or volatility in the marketplace, and more importantly relative volatility of major indices to each other: so volatility in the European markets versus volatility in the US markets. When we construct our long/short trades...spreads...we would often trade US indices against European indices. In 2011 volatility in Europe was a lot higher compared to the US and that hurt us. On average we are benefiting when volatility is elevated and when volatility is high, but we are not outright long volatility. We are long and short volatility at the same time, but in different markets. We would prefer bear markets, because that is the time when volatility is higher, but overall, correlation to volatility is positive, but it's not very high.
Niels
Now, normally I would go in and ask certain questions, but I think you bring up a topic that I just want to dive into a little bit more if that's OK, and that's the thing about your looking to...or you prefer, in a sense, conditions where the volatility, if I understand it correctly, of say the US equity markets and European equity markets is somewhat similar and their not diverging too much.
Peter
It could be high or low, but somewhat similar.
Niels
Yes, yes, and of course, in a sense, we've lived in a world where central banks and the introduction of monetary unions and so on and so forth, in my view at least, have meant that to a large extent economic cycles have become more coordinated, if I can use that word, especially in the western world. But we also see a lot of pressure building up inside these economies. We've used different methods of solving the crisis' and not all of them have been the same, and not all of them have been working equally well. We certainly see now, again, that Europe has certain issues that maybe the US hasn't. Now, in a situation where the world starts to become more fragmented, and maybe not so coordinated, what does that do to a relative value or spread strategy that you're running? What are the things that you might be alerted to where you would say, "hmm, something is going on here which is slightly different to what we would like to see?"
Peter
Well, yeah, on the one hand we are hoping that the volatility in the market place will increase. What we've seen over the last two or three months is unprecedented. If you look at the NAFTA implied volatility which is a fixed index, which is implied volatility of the S&P, but realized volatility of the S&P 500, has been running at the rate of 6%, 7% annualized compared to the normal average long term of about 18% - you can argue 16%, but still it's at least 3 times lower than it's suppose to be, if it could trade, so this coordinated assistance of governments around the globe has clearly helped to reduce the volatility to such an extreme levels that it's very difficult for systematic strategies to produce returns. So, on one hand we are looking for some kind of a crisis. We're hoping that there is going to be some kind of a correction, because the best investment one could make over the last five years would be long on the S&P 500. We know from the history that these periods do not last. We also know from the history that the previous time volatility hit such a low point were the times when, soon after, some crisis would develop. The last time we saw volatility at this level (implied, not even realized) was back in late 2006, early 2007, and we know what happened a year later. Yeah, so, to answer your question on the one hand we are sensitive to differences in volatility around the globe. On the other hand, we do want volatility to go higher. Our models are very adjusted, so if the environment changes, we would adjust quickly to the new regime. For example, the lookback window to calculate our hedge ratios, how many contracts to go long and short - that's S&P,that's FTSE, and the British Pound. That would be an example of what we would trade. Sometimes we would use a 30 day lookback window, only, which is a fairly short window. This means that if market conditions are changing rapidly, it would be reflected in our hedge ratios quite quickly. So, the reason for the very tough period in 2011 was that it was a very unprecedented scenario where credit risk of certain countries was under question. I remember some months when European indices were down 12% a month while S&P was flat. So divergences on the monthly level were just unprecedented. When the environment is like that, even our dynamic models can't handle that. Most of the other times we are doing a good job and that's why in 10 years we only had the 1 down year, which was 2011. All other years were either had one flat year or all positive.
Niels
I guess that's probably maybe part of the reason why you decided to launch the Multi-Strategy program that builds in certain levels of trend following, which I guess will benefit from divergence.
Peter
The idea for the Multi-Strategy program was to mix, in one portfolio, two streams of returns, that are not only uncorrelated, but in times of stress, for the spread component, it has a negative correlation to each other. So our directional models tend to produce their best returns when spreads are suffering and vice versa. That was the idea. Stand alone, I would say, the directional component that we have in the Multi-Strategy program the quality is not as good, stand alone, as compared to the spreads, but in combination with the spreads, the product makes a lot more sense. The idea was to reduce the likelihood of a large drawdown. At an expense of having maybe a larger correlation to CTAs, having some data maybe to some markets, and at the expense of not having as stable returns as we're used to having in the spreads alone.
Niels
Of course, absolutely, within trend following it's inherently unpredictable in terms of the return stream.
Peter
So that's one or two months per year that would produce your total P&L for the year. The spreads we had runs where we would have 10 consecutive positive months, small gains every month. A different return profile.
Niels
Absolutely, absolutely, great stuff Peter. Now, let me go back to my normal questioning here, and I want to start off talking a little bit about your organization. I wanted to find out about how you set it up. Obviously you mentioned that you had a big change in AUM, and maybe that's also impacted how you do things today, but talk me through a little bit how it's structured, and also how you've come about using technology to your advantage. Not everyone keeps every single function of the company in-house. Certain people choose to outsource some things, but of course, being a short term, you know, with very quick turnover (let's put it that way) in the portfolio, clearly you have certain requirements that the longer term or medium term guys wouldn't have, so talk to us about that.
Peter
Well, even at the current level of AUM, we maintain all the main functions of the firm in house, and we have five main functions. The main one, of course is research, researching the models and programming, the trading and programming, the back office functions etcetera. The trading that we have is fully automated. We have our own back office because we service managed accounts, and a lot of functions of the back office side were also automated by us in house. Our research uses proprietary software that we built to back test the models, not on a daily frequency, but using one minute resolution of data going back many, many years to estimate the parameters that we trade. That back testing tool we've been developing for the last 7, 8, 9 years. It's been upgraded constantly. We have the capability of running various analysis out of sample, in sample, test, etcetera, to make sure that our models are robust. We have compliance, obviously, in house.
Marketing is very important, of course. We do work with a few 3rd party marketers, but the main work is still done in house on the marketing side. And trading - we are open for trading around the clock, 24 hours, because we trade Europe, Asia, and US markets. We don't have traders that decide what to buy and sell. These traders, they are monitoring what the machines are executing in an automatic way, but they are there to make sure that everything is running according to the system - that we have connectivity and all that. At the moment all these tasks are performed by 12 people. Most people that we have, have stayed with us for well over 5 years, and in some cases 10 years. We were larger when we had more assets under management, but at that time I would say that we did not have the amount of automation that we have today. If we were to go back to the same AUM levels, I would say that the only area where we would be looking to hire more would be research, but every other function is covered quite well, so we don't need to go back to the previous head count. I believe we're at 26, 27 people I believe. I think from the quality of the service that we are producing today with a smaller group of guys and girls, we are a better firm compared to 2011 when we had a lot more assets.
Niels
Sure, sure, now the next area I wanted to talk to you about is track record, because, certainly if you look at the longer term CTAs, we know that the environment has been very different looking at their track records before 2009, and after 2009 you can almost draw a line and see quite difficult conditions and different returns. Now, I don't know so much about the short term space, so I wanted to ask you a little bit about your observations, about the market environment, and how should people interpret your track record, meaning have there been some major upgrades of your research and therefore the strategy over time - so that, if you look at the period from 2012 and onwards, perhaps the system is actually significantly stronger than it was before 2011. So just give us a little bit of insight into how should investors or potential investors look at your track record, in your opinion?
Peter
Yes, we did make some important modifications to the system in 2011 following the experience that we had. I would say that our returns in 2012, which were very strong, primarily were due to the adjustments that we implemented. We are running the same core idea, it's just the details - how we hedge, which markets we trade, or when precisely we'll buy and sell - these details are slightly more defined. Returns of 2013 were also very good, but I would say had we not made any changes in 2011, returns of 2013 would be quite similar, so we are particularly proud of our returns in 2012, because, if you recollect, the 2012 environment was somewhat similar to 2011, actually. The European debt crisis was continued in 2012, at least that was the case in the first half of 2012, and despite that we generated good returns with the adjustments that we implemented.
If we look at the industry overall, I would say that investors have a lot easier job analyzing and understanding returns of the long term CTAs because the correlation between them is often 60%, 70%, and they can see which ones are good and which ones are not so good. When you analyze returns of the short term CTAs it's a lot more difficult because correlations drop significantly, and there are some good short term guys, and some not so good short term guys. It's more difficult to understand the source of the returns. It's more difficult to understand during which market environments this particular short term program could do well and not do well. I think if investors were to spend some time and understand and analyze the short term players and find the ones that are really good, it would enhance their returns quite a bit. so it's more difficult to find a good short term trader, but if you found one, you should stick with them, even if they are having a local difficult time. Every manager, I would say, with 10 years of track record, at some point will have a drawdown, it's impossible otherwise.
Niels
Absolutely, that is a certainty. Given the fact, if we stay on this subject a little bit, given the fact that it sounds like you have done well during a period where most managers - and I actually would include short term guys as well, because I don't think it's been that easy for the short term traders either, and you seem to have done well. So, would you say...is that driven by the fact that you are relative value, if I can use that word, or spread traders, rather than outright looking for directional bets in the markets, take aside the time frame, but the fact that you do something differently to a trend follower, whether it be short term or medium term or long term, is that the key reason why the environment has not really been that difficult for you?
Peter
Well, even when you make money, you always think that the environment is tricky…
Niels
True.
Peter
Yes, what we do...what we are doing it's very different from 99.9% of CTAs. We speak with a lot of allocators, which ultimately interview a lot of managers, and so the feedback that we get from the investors, from the allocators, is that our programs are very unique, they do not know any other managers that are doing what we are doing, with the exception of maybe one, two names, but even in those cases those other managers will have their own details, their own risk allocation, their own ways of doing things. For example, this one firm that we know is doing somewhat similar to what we are doing, but if you look at the correlation between our returns and theirs, it's going to be close to zero as well.
Niels
Sure.
Peter
It's remarkable. So one of the advantages that we have, I believe, is that we do not have a lot of competition in what we do, because very few if any firms are looking at such short term price changes and not only that, but in a market neutral relative value type of trading. Being a minority helps, and if you are doing a good job, you could stand out, and this is our goal to stand out, and of course not every year we will be the best manager out there. There will be times when long term CTAs will have very nice performance and markets will go up almost every day, and we would underperform, no question about it. But there will be other periods when we will be doing a lot like everyone else. That is why it's very important to have us in a portfolio, because we will most likely have negative returns when other managers could be doing fine, and on the other hand we could generate our best returns when all the other managers are doing poorly. So on the portfolio level, we add a lot of value. We're not only providing absolute returns at the end of the year, that is very nice to have, but on the reduction of the portfolio risk level, we are adding a lot of value as well. I hope investors will recognize that over the long term.
Niels
Sure. Now Peter, before we jump to the next area. I wanted to jump forward a little bit, because this might be sort of a little bit of a research question, but I know from doing my research on your strategy, that you often mention that you don't change the core of the model, but you do a lot of optimization on a regular basis on the parameters. But when I have spoken to other short term managers before, they often tell me that actually model decay is a big issue and that a lot of their models tend to work for two years and then they have to come up with some new models. Is that something that you're kind of (neutral is maybe not the right word), the fact that you do market neutral strategies and so on and so forth, do you think that's what's helping you not necessarily having to come up with new models on a regular basis, and your models being quite robust in terms of longevity?
Peter
Yeah, well, the core essence of the strategy is, on one hand very simple, and on the other hand is very persistent. If we just step back and try to understand what we're doing.
Niels
That would be great actually.
Peter
Looking at most liquid and highly related or correlated global equity markets, again, let's look at the example. FTSE 100 and S&P 500, correlation between the two indices on a data level historically ranges from 70%, 75% up to 90%, 95%, and it's very stable over time. What helps drive our returns is a very simple fact that these equity indices, they open for trading in different time zones, and hence their liquidity intra-day is not the same at any given hour, it's shifted in time. When Europe opens, S&P futures trade, but they're not as liquid as they will become latter on. When US closes, 4 o'clock New York time, European futures trade, but they're not as liquid as the S&P at that time. So what happens is that, while on the daily level, on average, if you average across many, many days, going back, correlation stays high intra-day. These markets could temporarily diverge from one another for liquidity reasons. These are not fundamental divergences, these are liquidity driven mis-pricings, and that's why, when volitility is high, this is a good environment for us, because these divergences tend to be larger and more often. When volatility is low, people are not afraid, they're not scared, and they do not make these little mistakes - they have time to react to whatever news is coming out. When volatility is very high, and news is coming out, sometimes people don't have enough time to make the rational decision and they tend to over buy or over sell certain markets, and most people are concerned about a directional or particular market. They are not concerned how FTSE is traded in relation to the S&P. Most people don't even care about that. On the other hand, that's all that we do. So, if you think about it, these time zone differences will always be there, the remainder of time. We wake up, let's say the UK and US, 5 to 6 hours apart. If we talk about Asia the time difference is more than that, 12 to 14 hours. So, that type of structuring event, or fact will stay forever.
Markets, now a days, on the other hand experience a lot of correlation because markets and economies are becoming more and more global every day. So, what we do when we re-back test and re-optimize parameters, we... see in the short term trading space, it's very important when you buy and when you sell, because if you are a long term guy, and you hold a position for a month or five months, if you buy tomorrow or yesterday your P&L at the end of the trade will be more or less the same. If you are holding positions for hours - 6 hours, 5 hours, 12 hours, or a day and a half, the timing of your entry and exit plays a crucial role, because your P&L could be totally different if you are one hour late or one hour early. So, while in our case the main concept stays the same for all these almost 14 years now, we have to adjust to the every changing market environment. We have to try to incur better, when to place these trades, how to hedge, how many contracts to go long and short, which markets to cover. Let's say that after 2011 we stopped trading certain European markets like Italy and Spain. We added some new markets this year which we think are overlooked by most players, but at the same time have enough liquidity in them. So these details of how we execute strategy, they are adjusted, but the core concept stays the same and this phenomena doesn't disappear after two years, thankfully. We have ways of monitoring that. We have certain mean reversion statistical tests that we do before we optimize our model, before we fund these hedge ratios. We try to look at just simple data without any trading signals at all. We try to simply understand if this time series, or spread, has mean reversion in it, or it doesn't. We're doing these tests constantly. For example in 2012 and 2013 these tests looked as strong as ever, and that's why the returns were quite strong for us. There are different ways of how we are monitoring the situation, but my main point is that the main source of the arbitrage that we are trying to exploit, stays, it was there, and it will be there, unless billions and billions of dollars will come into this market segment and start to arbitrage that away. It's always possible, but we're not seeing it so far.
Niels
Sure. Obviously I'd like to go into a little more detail, but just to make sure that I kind of understand what you are saying is that, essentially that one of your strategies is to buy or sell a certain spread between the FTSE and the S&P that normally would take place, as an example, more or less at the same time, in the morning where liquidity is somewhat different between the FTSE and the S&P.
Peter
We execute our long and shorts at exactly the same time.
Niels
Exactly, so that's the key.
Peter
What we see, in the early morning hours, Europe is the leader, S&P futures could be the laggard. I'm talking about 2, 3AM New York time, in the early morning hours. At around 3, 4PM New York time, it's the opposite. Futures on FTSE and CAC and DAX and SMI are traded, but cash markets are closed, and that's the time when S&P is leading the way and the European indices are following the example. Sometimes if they under react to the moves in the S&P, that will produce a move in the spread in a certain direction, and what's interesting is that the next day when European markets reopen, cash markets reopen, they often would price in whatever was underpriced the day before. If it doesn't happen, S&P futures could react and adjust to the European level. So that produces, often, moves in the spread in a certain direction and then reversal of the spread that we would trade, and that's exactly what we are looking for.
Niels
Right, OK. From memory, I seem to recall that your average holding period is about a day or so, so it's not so much the difference between the liquidity from the morning where you say the FTSE is the leader and to the afternoon where the S&P might be the leader that you are looking for, or did I misunderstand that?
Peter
The average holding pattern is traded about one day, or 8 to 10 trading hours - that's the way to describe it. If we see a reversal within an hour we will take it of course. If we have a position that is not reversing, we could be stuck in the position for two or three days, potentially, and there are two ways we can exit a losing trade: we have a hard stop - if markets are totally diverging from one another, and we very rarely hit the hard stops, less than 1% of trades are stopped that way; the other way we can get out of a losing trade is in a small reversal, eventually the spread would reverse, but not to the level where we took the trade originally, but if we hold the position for longer than 1 day, we want to get out of that trade as soon as possible. On the small reversals we would often, in a fully automated way - systematized way - we would get out of the trade with a loss, looking for another opportunity in the future. After one day our predictive power is getting weak. We can predict direction of the spread with very good accuracy within one day. In fact, our hit ratio - percent of profitable trades is on average about 65% profitable. So roughly 2/3s of the trades are positive, and only 1/3 is negative. That hit ratio is strong only because we have very high predictability within one day. Beyond one day markets are very efficient, if markets are diverging for two, three, four days, that means they're diverging for fundamental reasons, not liquidity reasons, and if that is the case, there's no reason for us to hold a position for another day. We're trying to basically cut the losses if we have the losses and wait for another opportunity.
Niels
Sure. Let's try and put some numbers on so it makes it easier for everyone to understand. Let's just say that the S&P was trading at a price of $1,900, and the FTSE was trading at a price of $5,900, so a difference in price of 4,000 points, if you take it that way, so is it correctly understood that what you might be looking for is that if that price spread goes to, say 4,050, and you think that within the next 1 day of trading that the spread really should go back to 4,000, is that the kind of difference you are looking for?
Peter
Well, our spreads are not just pair spreads, we actually have a third leg, the spread is a British pound contract in this particular case. We have to take into account the fact that FTSE futures are denominated in the British pound, while trading the US dollars. So we have a currency leg that hedges the currency exposure that we have. So it's a triangular relationship. Let's assume that currency is not moving, and let's assume S&P is up 1%, and FTSE is flat, at some point it will trigger a trade where we would go short S&P and long FTSE, expecting FTSE to catch up, or S&P to come down, or both to take place. If that reversal could happen via the currency move actually, it's possible that the currency leg will move in reaction to the move in the S&P that will push out P&L onto positive territories. All three legs are very important. Sometimes...I'll give you another example, let's assume that FTSE and S&P are flat, but the British pound is moving let's say 1%, that will potentially trigger a trade which will reverse, ultimately, not because the British pound contract reversed, but because equity markets - FTSE and S&P will start re-pricing themselves in relation to one another, depending how the currency moves. There's some intricate relationships between both equity markets and the currency, and all of them we are taking into account, and we could potentially trade.
Niels
Sure, this 3D way of looking at it, is that required, or could you also do spreads between say markets both denominated in US dollars?
Peter
We have spreads like that. An example would be trading Russell 2000 against S&P, against NASDAQ 100. Here we're exploiting a different type of arbitrage which is not time separated but liquidity separated, but small cap versus large cap stocks, short term divergences. It's a known fact that small cap stocks sometimes react with some delay to the news, to the moves in the large cap stocks, and that can cause a temporary diversion to Russell S&P, which latter on would be correct. We have different types of spreads. There's a third one where we would trade two equity markets, one against another and the commodity market on top. That commodity hedge is used only for those indices that are highly correlated to commodities, for example the Canadian stock market, or Australian stock market, those indices of high correlation to gold, copper, crude oil, and we would use commodity for the particular relationship to trade.
Niels
Fascinating. Tell me Peter, how many markets do you trade all together, and how many combinations of spreads to you have in your portfolio?
Peter
Yes, well, we trade approximately 20 to 25 different markets that include...
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 the 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.