“I thought that trend-following was an amazing part of the story, but maybe it could be complemented by some contrarian approach. And also, taking the trend by cutting it into parts.” - Jean Jacques Duhot (Tweet)
Today on Top Traders Unplugged, I'm joined by my co-host Moritz Seibert to speak with Jean Jacques Duhot, the Chief Investment Officer at Arctic Blue Capital. Jean Jacques's observations of different discretionary trading teams has given him valuable insight into a variety of trading methods and has allowed him to develop a unique multi-part model that utilizes the role of a contrarian mindset to great success. Listen in to today's episode to learn about the beginnings of Arctic Blue Capital, how Jean Jacques developed his multi-part model, and how he tested it using a specifically "agnostic" outlook.
Thanks for listening and please welcome our guest Jean Jacques Duhot.
In This Episode, You'll Learn:
- How Arctic Blue Capital got it's start
- What influenced Jean Jacques in how he developed his unique "contrarian" trend following system
- How Jean Jacques shaped his multi-part trend-following model
“You have your very reactive traders, your very patient traders, and your contrarian traders. And you want to empower them by giving them the same level of risk and being totally independent from each other, and not talking to each other.” - Jean Jacques Duhot (Tweet)
- Who the three different types of traders are in his model
- Why Jean Jacques took a more "agnostic" outlook when he tested his model
- The reason why using a relatively small portfolio of markets benefits Jean Jacques's model
“We found that the VIX was acting as a good warning signal on one side, and a carry generation on the other side.” - Jean Jacques Duhot (Tweet)
- Why Jean Jacques prefers ETF over the futures markets
- How VIX can act as a warning signal
- Why going in full on a new signal works for Jean Jacques
Connect with Arctic Blue Capital:
Visit the Website: Arctic Blue Capital
Call Arctic Blue Capital: +44 207 292 1608
E-Mail Arctic Blue Capital: ABC.Investors@h2o-am.com
Follow Jean Jacques Duhot on Linkedin
“I think that there's amazing gains of production in the United States. The farmers have been the first benefitting from US banks to get cheap money lended to them, and they made tremendous gains in productivity, and therefore the supply is quite ample.” - Jean Jacques Duhot - (Tweet)
The following is a full detailed transcript of this conversion. Click here to subscribe to our mailing list, and get full access to our library of downloadable eBook transcripts!
Hey everyone, and welcome to another edition of Top Traders Unplugged where today, my co-host Moritz Seibert and I are joined by Jean Jacques Duhot, who is the founder and CIO at Arctic Blue, which is part of the Fifteen Billion Dollar Plus H2O Asset Management Group.
Now JJ, thanks so much for coming on the podcast with Moritz and me today. We really appreciate you taking time out to do this, and we’re very excited about our conversation today. Now, there is an interesting story behind your firm and your journey, in particular, and the question of why you created the strategy the way you did. So, why don’t we start with that? Tell us about your journey and what you’ve been up to all these years.
Thank you first for having me, both of you. I am extremely honored to be part of this podcast that I have been listening to for many years now.
The journey of Arctic Blue started in Canada when I moved to the Great North in 2007 to join the Sovereign Pension Fund of Quebec. We developed and implemented systematic strategy with a specific approach, by asset class, taking into account the panel of market participants and their specificities.
In Arctic Blue, the name came from the fact that one day a friend of mine was working for the Administration of Canada and we ended up in Nunavut, which is north of Quebec, in a cargo plane, and the temperature was around minus 43, and we just landed, and it was pretty cold.
The sky was extremely blue, and there was a sign with the distance to the polar circle, and I thought that, if one day I were creating my own fund, I would look at a name like Arctic Blue. At the same time, the logo that we use, which is an Inuksuk, which is that man made of stone, and it’s a directional marker, in the Great North, built by the Inuit population. For us, being directional traders, I thought it was also a good complement to the story.
Yeah, no, absolutely, very interesting. For me, at least, the whole structure and how you’ve used the background that you have in terms of creating your strategy, thinking about what it is you want to achieve, is quite unique. I’ve not really come across that before. Why don’t you take us back a little bit, in terms of your professional background, and how that influenced you in creating the strategy and the philosophy behind it, perhaps.
Well, I started my career at Societe Generale, and I stayed there for a bit more than ten years. Around 1997 I started to manage a group of proprietary traders, taking a risk myself, and analyzing their performance and their P&Ls.
So, I started to sit close to them taking notes about their behavior. I was interested in human reaction regarding risk-taking: risk-taking in different trading environments, and different volatility regimes. [I asked] is this trader patient, impatient; is he rigorous in terms of risk; is he reactive, is he opportunistic?
We learned great lessons by people making money on a consistent basis, but also the ones that were losing money on a consistent basis. At a point, I developed some frustration around some seasonal patterns in banks where you started to be asked to reduce your risk exposure (around the October/November [timeframe, which were] quite famous bonus times in those days) where a lot of opportunities in the marketplace could not be captured by people like us. So, I start to read a lot about a firm I think you know well, Niels, Dunn Capital Corp., and John Henry, and really, the totals, and all the legend of amazing individuals that gave birth to our industry.
I thought that trend following was an amazing part of the story, but maybe it could be complemented by some contrarian approach, and also, by taking the trend and cutting it in two parts - break out at the beginning, then trend, then contrarian. The chain of those different phases in the momentum cycle were relating around the impact of the new information hitting the price and creating different types of positioning by market participants.
I thought that when new information is hitting the price, it tends to create a surprise effect and the market goes from a very precarious equilibrium to an unbalanced distribution in the price action, to then reaching a point where it could be a maximum favorable excursion, or it could just mean revert, and that first part being very impulsive.
I guessed that what was needed was to have develop a systematic trader that would be very reactive at entry and very opportunistic at a money management level. You know very quickly if you’re wrong and therefore you have tight stop losses, and you have a multiple areas of profit taking depending on the shape of the expansion.
Then the market tends to pause, and you end up having a second type of crowd coming in. [There is] a second layer of human beings keen to take risk as the information is more analyzed, accepted, and you observe more of a convergence type of price action where that information is becoming really consensual.
You are moving from a game of reactivity, in the first phase, to a game of patience. As human beings, those parts of the trend tend to last much longer than expected. When we think about US equities, who would have told us, in 2012 or 2013 that we would have been in a nine-year-long expansion, I would have been the first to doubt it.
At a point, therefore, we don’t want to be too reactive in terms of money management. We want to stay in for the longest possible part of that convergent phase to the point where that information is becoming so well known in public media and completely over-consensual that you need to then think about introducing a third class of market players that are the contrarians and playing on the exhaustion of that momentum.
You end up having three groups: you have your very reactive traders; you have your very patient traders; you have your contrarian traders. You want to empower them by giving them the same level of risk and being totally independent of each other and not talking to each other. Which I think was also an interesting point. In some firms, where I’ve been operating, you end up having quite a significant correlation risk between discretionary traders as they tend to talk to each other, feel good about it, and you end up having some potential risk of max-peak to valley drawdown on that front.
A firm where I’ve had the chance to manage capital for was Millennium Capital Management. At Millennium we were not very encouraged to talk to each other in order to be sure that their local relations between the teams was maintained. I think today they have more than two hundred investment teams and it’s a very successful firm on that front.
So, really keeping that independence for each model, with each of them marrying their own risk, and you end up back to our beginning of the journey - creating that desk of automated traders, with their independence, under the supervision of another set of algorithms that are the risk managers in order to avoid correlation and concentration risk. So, that’s really more of an empirical analysis, at the beginning, of human behavior, regarding risk taking and market price action compared with some other systematic funds that have more of a purely statistical analysis that we use, but at a later stage.
I’m interested. So, when you did the initial analysis, and you were making your notes about these traders, were there more of a certain kind of trader? Meaning, are discretionary traders, do they tend to be more contrarian or mean reverting? I’m thinking that finding discretionary traders that have the mindset of being a trend follower is, perhaps, rarer than it is for someone who will take the profits quicker and so on and so forth.
You’re absolutely right. On that point, I could not add anything else because I agree one hundred percent. But, back to the early part of your question, what was striking was the inability for a very large percentage of them not to take their loses very quickly but to more try to average by being incremental when momentum was against their initial position. That was what some trend followers are doing, which is adding when the moment tends to be in favor of the position.
There is this famous line by Paul Tudor Jones which says, “Losers average losers.” I think that’s what you’re describing. So you’re in a loss-making position, and then you average your price by buying more on the way down. It’s not a good idea, in general.
Another thing that is striking to me, when I was preparing for our conversation today is that, obviously, you thought about the system design very differently, which is intriguing, but I also thought your choice of playground, so to speak, the markets you decided to trade, I’d love to hear the background of that because it is an unusual portfolio, to some extent, in our space.
Yes, so as I joined the Sovereign Pension Fund of Quebec, in Canada, due to the very large portfolio and multi-assets that they were managing, I thought that an approach by asset class made sense. Digging into that early observation of their portfolio, I found that the commodity traders tend to be very different from the market participants in equity markets.
When you look at commodities you have, on one side, the producer and the consumers - the Starbucks versus the Brazilian coffee producers, on one side - that come to the marketplace to guarantee themselves a price to buy or sell their stuff they need. On the other side, you tend to have people like us, with some financial techniques, with potential asymmetry in our favor, to harvest those insurance premium rode by the commercials.
But, when you look at the dynamic between commodities and the overall asset class, you tend (most of the time) to pay to wait, i.e. you cannot apply traditional cash flow analysis such as in real estate, or equity, or fixed income where you get paid to wait via dividend, rents, or coupons. Therefore, I think it was important to take a very agnostic type of view of that playground and equally allocated risk for breakout, trend, and contrarian models as to whether supplied disruption, geopolitics tend to send very little warning for abrupt [change] in the direction of prices in the commodity space.
So, that was really, that approach for commodities, and looking at what were the most liquid markets and the correlation between them, between different complex and intra-complexes, that was really the approach for commodities. Then we applied the same for fixed income. So, an approach by silo wherein fixed income you still have a momentum engine, but the curve is a very important component and therefore deploying models specifically designed for curves.
For currencies, there is the same momentum engine, but with specific models developed for carry on thirty pairs of currencies including emerging. The carry model has money management around it, in a dynamic way, instead of having those classic carry models that tend to just benefit from the interest rates differential, but then, all of a sudden, try to have significant max peak to value drawdown.
Then, in the equity space, you end up having a bias to be longer in equities. We see that distribution of momentum that tends to last longer than in other asset classes when they are free to evolve, I would say. This is a bit different than in 2008 when you look at fixed income with QE analysis. You look at equity from corporate buy-back to 401K, to a large range of pension funds of Nordic or Middle Eastern countries, or Japan, that are very large buyers of equities, making the overall float less and less ample, I would say. [And then there’s] all the merger acquisitions also, on top of this.
Therefore, if you want to bring that typical type of proposal in the equity space, we have been taking an approach where we would overweight the contrarian bias in order to bring cover, if I may use the term, to the long equity or long/short types of portfolios in case you are exposed to dislocation, such as in August of 2007 for example. So it was really taking into account the nature of the panel of market participants in each asset class by not tweaking the model, but just applying the model with the same type of input but with different weights, based on the nature of the underlying momentum and what the purpose is of your strategy in order to avoid purely delivering beta in phase as we’ve known over the past few years.
Yeah, how many markets to you actually have in the portfolio?
So, in the commodity market we have twenty markets, and that’s a specific product that is available in different types of formats. In equity, we have sixty different markets, and so on, so it’s really segregated. We do not offer, like most firms, four asset class products.
Right, right, interesting.
I saw in your presentation that you’re focused on commodities, and I saw the energies in there and a few others, but it’s a relatively small portfolio, correct me if I’m wrong, but maybe ten markets or something like that. Is there a reason for that and what impact does that have on your diversification which I guess you’re seeking to maximize?
Sure, sure, so we’re trading twenty different markets in the commodity space with significant representation, in terms of numbers in Ags.
So that’s expanded, over time, the number of markets? Have you always done that?
Yes, because we basically used to trade Ags as a basket; and we trade grains, softs, meats, and cotton individually. Then, in the energy space, we’re trading the main markets. What has been, for us, important is taking into account the correlation risk. We tend to find that we extract very little diversification from the barrel.
When you look at the different products, and in terms of risk/reward, mainly with the Sortino ratio, we thought that it was a less attractive proposal for us. We tend to be mid to long-term, lower frequency, and selective into our participation. If you want to go even further about that, we may be doing something different, but I guess everybody your…
They will always say that they are doing something different, that’s true.
We are binary, so we really look at being in or out, but we are not incremental when momentum is positive.
That is different.
So, the full amount of risk would be deployed at inception of the position when the signal has been cleared by some signal-to-noise ratio filter. Also, when we are looking at how differently we can be looked at, we have that specific approach by asset class, which I think will also make us slightly different. Those independent models have the ability to participate at different locations through the momentum cycle, in terms of trade positioning, so that gives us a low correlation with most of our peers.
Very interesting, thank you. One other thing, if I may, that I found interesting was that most of the markets you’re trading are futures, then there have been a couple of products where, for instance, don’t trade the Gulf Futures, you trade GLD, at least we found that in one of the presentations. Is there a reason why you’re doing that? Why you prefer the ETF as opposed to the futures markets?
Well, the ETF tends to have two specificities. So, we are trading futures in one program, we’re trading in ETF in the other program. ETF is clearly not that efficient in terms of funding, so that’s the first negative point. But, on the positive side, we’ve found that the ETF market is reflecting participation of investors in gold without the noise of the short-term high-frequency trading during night session mainly.
In a very noisy market, such as gold, and mainly since 2013, we did not observe the same type of distribution of directional opportunities than in twenty years prior to 2013. We found that the quality of the signal was quite higher through the ETF. Also, you think that the ETF is being backed by physical, which is quite a bit different than the futures where there’s always this mystery about the number of claims versus the effective amount of gold available, but I don't want to sound like a gold bug here.
Interesting, and then there were two other interesting markets that I found in the portfolio, next to the commodities, which was the VIX index and the US dollar index.
So, the dollar index obviously is nearly fifty-eight percent of Euros as we speak, tends to capture the dynamic of the correlation between the dollar and commodities which, through the time, is not always as stable as we think it is. So, it acts as an alpha generation via a currency hedge, even if our index is traded in the same way as the other markets. So, that’s the presence of the dollar index.
Now the VIX, well, the VIX has been there because we could not identify the liquid future volatility of commodities. We looked at what is the largest asset class in the world with a liquid volatility instrument listed, and we found that the VIX was acting as a good warning signal on one side, a carry generation on the other side, but as our input is taking into account price, implied volatility, realized volatility, and we do not trade options, we do feel that the VIX was a good compliment by, in a way, trading volatility as a sub-asset class.
Have you seen changes in the VIX, actually, after XIV disappeared? Have you noticed changes?
Yes, because we have two types of models. We have the models that we take position with, that were described before; and on the other side, we have filters. Those filters are analyzing the relationships between different types of volatility. We’ve been more and more filtered out of the VIX market.
It’s interesting because when you look at when that episode of dislocation happened at the beginning of the year, you looked at the VIX as having bottomed in November of 2017, just with a “print” of nine and it then started to really increase until that quite volatile episode. Since then it never revisited those low levels.
So, for us, on November 17 there's an inflection point in the market. So you could look at the VIX correlation with the five-year, thirty-year spread in US treasury. You could look at where we are in the cycle with the Fed. You could look at the trade wars today. You could look at geopolitical tensions, but there’s clearly an inflection point in market behavior around that date that has been reflected by the VIX.
Yeah, my observations seem to be that, not only has the volume dropped since XIV lost all their money, so to speak, but also there are just fewer people who are willing to sell at the very low level. So, as you say, the bottom of the VIX seems to have gone up a little bit compared to last year. Interesting.
I wanted to ask you a little bit more about the commodity side. So, traditional trend following systems, and it may not apply to the way that you do it, but my experience is that we tend to make more money on the upside. So, we tend to make more money on the long-side trades compared to the short-side trades.
I’m interested in finding out whether you have the same experience with the way you trade commodities. Also, whether maybe part of the struggle for CTAs, in recent years, may come from the fact that we’ve been in generally down markets or bear markets for many commodities. So, perhaps opportunities have been harder to capture. Have you thought about this? Have you noticed this? Have you…
I think we’ve made more money being short commodities than being long on our side, mainly due to the contrarian models.
Right, of course, which makes sense.
What we’ve been observing is lower participation in the market, very large fundamental discretionary hedge funds that have exited the playground. Also, inflows that were behind commodity arise from early 2000 to 2008, 2010 was pension fund money and commodity index money that was protecting against inflation and also benefiting from the thematic of Chinese infrastructure development.
Since we’ve observed those very low and sticky, low inflation levels, that money is not as aggressive and is not a part of the asset allocation as used to be the case pre-2008. So, therefore, we’ve seen some markets offering better diversification at market level. I think that when you look at the price action of indices such as the Blumberg Commodity Index, or the GSCI, they are trending quite less. There’s less correlation between commodities than when the commodity index was dominating the price action in the first part of the 2000 years, and therefore the markets are more back to their fundamentals.
I’m thinking about sugar that had an amazing range (from eleven to twenty-four, back to eleven) now, over the past few days in sugar. It has been a very generous market for us. Some markets have structurally changed.
I’m thinking about natural gas in the US. It’s a market that used to trade in back-month twenty dollars, when you take into account the role, pre-2008, and with the technology based on the extraction of shell gas. We’ve seen a market ranging between high 2s and low 4 dollars. So it’s a price divided by nearly five or even more. That also made the mid to long-term directional reporting tendencies less easy to catch in this market. So, there has been a redistribution, generally, and commodities are exhibiting more of an individual type of behavior.
This is with some successes and some disappointments too.
But more divergence in general in those markets it sounds like?
Definitely, and a very high concentration of volatility in some markets like the meats, for example. On the other side, you have a structural phenomenon of very depressed prices for agriculture. When you look at the past forty years, the overall price of agriculture has been quite mind-boggling, for a lot of people, based on all the long-term positioning, added to demographics, growing demand, and the evolution of food habits in the world. When you look at what happened in terms of prices (at least on US commodity futures in the agricultural sector), it’s been quite the opposite. We saw those prices really going lower and lower and lower. If you look at the DBA, for example, as a basket of agricultural futures, we saw we are trading at all-time lows.
That’s the ETF you mentioned?
Yes. That’s all-time lows. Therefore I think there are amazing gains in production in the US. The farmers are first benefiting from US banks getting cheap money lent to them and they made tremendous gains…
In technology and production.
Yes. Therefore the supply is quite ample.
Yeah, yeah, right. So, it sounds like the concept of your strategy obviously took time to essentially formulate from all of the observations that you talked about and so on and so forth. I know I’m not phrasing it correctly, but it seems to me that you’re sticking with your methodology. You don’t really change it.
There is this debate, which I find interesting, and that is the question, should we adapt our models to the changing environment? Because clearly the environments that you suggest does change. Or should we stay with a relatively unchanged approach to trading? Where do you stand on that and what have you done inside of your own strategy when it comes to adapting it?
So, the approach stays unchanged and the bulk of the research, except for very classic execution tools and their efficiencies. It’s been, really, all about money management and risk approach with an ability to look at the position sizing, the evolution of the position sizing, the allocation through the different markets and the ability to filter ourselves out of some very noisy episodes and always trying to find the right balance.
I think we did a fair job in some markets. I’m thinking about crude oil where the distribution on the upside looks very obvious when you look at the graph, but as a participant, I think it’s been more challenging. Setbacks were quite violent, and in some other markets our presence has naturally been smaller and smaller, and I think one more time about Nat. gas.
So, it’s been really getting a higher reactivity and, at the same time, a higher filtering based on the price distribution - if we are in or if we are out of the market. Therefore, it’s been more the money management part of the algorithm that has been covering the bulk of the research as well as an emphasis put on an additional model brought to the arsenal (if I may use the term), which is the short-term reversal model.
Sure, sure, ok, ok. I know you’ve got some questions, Mort, but I just want to just finish off on this topic because I think when you talk there’s always something new that I want to ask you. One thing that springs to mind is what you mentioned earlier about when you get in you get in fully and you stay invested fully until the signal changes. I find that really interesting, so I’d love for you to explain why you find that to be advantageous for your approach.
Also, I’m wondering, when it comes to capacity, commodities of course in itself have lower capacity than many of the financial markets, but also then adding on top of that trading the full size might actually limit the capacity in some ways because you’re dependent more on liquidity in a certain time when you want to trade without incurring too much slippage compared to getting in slowly, getting out slowly, if I can put it like that. What are your thoughts about that? Why do you find that the way you do it suits your objective?
In terms of the different facilities offered by the exchanges, if we’re going into the nitty-gritty of it, you have a trading asset element, like TAS, which offers a very good amount of liquidity with a very minimum slippage if you are considering that the closing price gives you the validation of a mid-term pattern. That makes your slippage for the more classic model not being an issue. Now, stop-entry on some contrarian models indeed has a very low footprint and that has been part of the research around the implementation of those triggers, taking into account the volume.
Do you spread the volume over several maturities, do you enter in a different time? So, a lot of research has been put on the distribution of the volume of intra-day around market, on open market, on close TAS, block trades, that I think gives you an interesting new dimension versus the very traditional volume and open interest type of analysis.
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 or Sound Cloud 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 on the next episode of Top Traders Unplugged!