“If you really believe in what you’re doing – you should not be de-levering in a drawdown.” – Kim Bang (Tweet)
In our second part of this conversation, we dive into the details of the program that Kim runs. We discuss why it’s important to stick with your models, even in times of severe drawdowns, and how Kim views risk. He also talks about the books and people that have inspired him in his career.
Thanks for listening and please welcome back Kim Bang.
In This Episode, You’ll Learn:
- The details behind the program Kim runs.
“I think there’s a direct correlation between the simpler a model is, the more robust it is.” – Kim Bang (Tweet)
- What he feels about larger drawdowns.
“If you believe in your models, never ever deviate from them.” – Kim Bang (Tweet)
- What kind of indicators goes into the models that he uses.
“We are probably out of the markets about a third of the time.” – Kim Bang (Tweet)
- Why he uses a non-cyclical approach to creating trading models.
- How he looks at volatility.
- What kind of mean reversion strategies he uses.
- How he responded to the Swiss Franc move.
“In general – these are the kinds of events that we live for.” – Kim Bang (Tweet)
- How Kim quantifies risk.
- How do you prepare for a drawdown when you haven’t gone through a significant one?
- How his firm comes up with new ideas and goes about doing research.
- The positives and challenges with working as a father and son team in the same firm.
- How he knows when a model has stopped working.
- What he is doing to launch his first fund.
“We have an inverse problem, where it’s costing us to keep money in cash.” – Kim Bang (Tweet)
- What questions investors should be asking in due diligence conversations with him.
“Most of the conversations we’ve had so far are with these early adopters, and they are very savvy.” – Kim Bang (Tweet)
- What books have impacted Kim’s career.
- How he sees the firm in the future.
Resources & Links Mentioned in this Episode:
This episode was sponsored by Swiss Financial Services:
Connect with Prolific Capital Markets:
E-Mail Prolific Capital Markets : email@example.com
Follow Kim Bang on Linkedin
“If you want to be in this space at the cutting edge and you want to compete, you have to be somewhat original. Not completely from scratch, but somewhat original in the way you extract earnings.” – Kim Bang (Tweet)
Kim: There're so many problems with these kinds of drawdowns, because I think a lot of investors, they really can't stomach probably once you start going much beyond 15% drawdown, certainly 20%. I think a lot of people get very uncomfortable. I would say that I would also get rather uncomfortable because it's not really what I would be expecting. You've got to have a certain type of investor and a certain type of manager who is able and willing to keep going even if you are experiencing 25%, 30%, maybe even 50% drawdown. I'm not sure how many managers and how many investors really can pull that off successfully.
Niels: Most new CTAs benchmark themselves against the BTOP50 Index, comprised of the some 20 firms in possession of more than 50% of the industry assets. These are the largest and most successful firms in this particular category. The question is, how does an emerging manager compete and break-in knowing 80% or more of the available investor capital flows to these few firms. The good news is that all of these firms at one stage was also an emerging manager. Of the 20 firms in the Index, 19 of them started with AUM of less than five million dollars. Interestingly, each of these firms had their best-annualized returns for the first three to five years in business, returning on average 20% per year. Once their AUM grew beyond 500 million dollars, their annualized returns settled around 10% to 12% and their correlation converged.
Introduction: Welcome back to Top Traders Unplugged where the best traders in the world come to share their experiences, their successes and their failures. Let's rejoin the conversation with your host, veteran hedge fund manager, Niels Kaastrup-Larsen.
Kim: ...Currently, we're allocating 50/50% across the two time frames where we want to move to an allocation of about 1/3 in each time frame. So again, sort of a philosophical perspective of an even weighting. The reality is, in our case anyway, the shorter term time frames actually have better risk-adjusted returns, higher sharp ratio. We are making a deliberate decision that, no, we want an evenly balanced portfolio. We also look at it from a scalability point of view, it's harder to scale on the very short term than the long, that kind of thing.
Then we look across the models, and we really have multiple models running right now but they fit into two broad categories I would say: directional volatility, and mean reversion models. Now, our core philosophical viewpoint, or perspective is that we believe that the markets have directionality and that there are trends, and that's our cornerstone. We believe that that's the place for us to be, but we have done pretty extensive research, and we've found that the mean reversion models actually can, on a standard long basis, actually make money and have a good risk-adjusted returns and perhaps, more importantly, they provide a more steady return stream and an inversely correlated return stream to the core directional volatility models. So, we weighted those 70/30… 70 to the directional volatility, 30 to the mean reversion strategies.
So, now when you look across portfolio risk you have the situation of how do you want to manage your portfolio as the assets grow and the assets decline? Let's just say it has nothing to do with withdrawals or additions, it's just has to do with the underlying performance of the portfolio.
In our case the research that we've done and the decision that we've made is that we will increase our capital on allocation and risk proportionately as the assets grow, and we will stay static as you go through a drawdown, so, that's a decision. In terms of position risking, we look at underlying volatility in the markets, so we will position size relative to the underlying volatility of the market at any particular point in time. In higher volatility environments, we actually look to reduce our positions, to scale out our positions, and in lower volatility we look to increase our position sizing.
Niels: So let me just… I just want to make sure I fully understand this. So, are you saying that, as an example: you have 10 million dollars in the management and when you go through the allocation, let's just say that coffee gets a 10 lot allocation based on current volatility. Now, if you drawdown say 10% or 20%, and volatility is unchanged, so all things being equal, you would still trade coffee at ten lots? Is that what you're saying? Meaning that you keep the same asset allocation during drawdowns?
Kim: Okay, so you're talking about two things here, on a portfolio level and then on an individual market level. So on the individual market level, we will look to use a constant risk, and you're right, relative to the overall asset on under management number. But to use your example, if the volatility in the coffee market is the same, and in your example, at an AUM of 10 million we were trading 10 contracts, and now AUM goes down to 9 million, and the underlying volatility of the coffee market is the same, then yes we would be trading 10 contracts at that time as well.
Niels: Sure, sure. It's very interesting because I've seen this before, and to some degree I like that. I mean I know a lot of people who say logically if you have a smaller account size, you should trade smaller. On the other hand, we all know that if you believe in your system and then...
Kim: That's what it boils down to!
Niels: Exactly! Then you know that the best time to invest is when you're in a drawdown. So I can see the logic in what you're doing by keeping it… although I would say personally, this is just a personal observation, there probably should be a limit, meaning that there is a certain… Because otherwise you could end up having a disproportional higher, internal leverage than when you started out, and maybe there is a limit in your system as well.
Kim: Niels, your observation is spot on, and I think you're exactly right, it comes down to your inner belief. If you really believe in what you're doing and your models and your portfolio construction, etc., etc., then you should not be de-levering, right? Because the best opportunity is during the drawdown, and you want the recovery time to be quick, right? You don't want it to be dragged on, so that's why you don't de-lever.
However, there is a point where you have to say, "Well there's something wrong here." For us, that is about this profiling. So that's where the historical simulation and portfolio construction, and the correlation matrix, and everything, all of that research comes in. Based on when you look at that and you go back over 30, 40 years, and maybe more years, and you have that data, and you simulated it, and you built these portfolios, and you have a high degree of confidence in sort of that inter-relationship and the behaviors of all these models, right? And your expectations are that the worst case scenario over those 30, 40 years, is a 20% drawdown or something, right? Then I think you have to say that at some point if you're exceeding those parameters, I think you have to then say that something is wrong. There's something wrong, I made a mistake somewhere, or maybe the world has changed so much it's possible? But there's something wrong so that you have to pull the plug at maybe 25% drawdown, you just have to pull the plug. And that's sort of what I've been saying to investors and prospective investors. I've said to them, "Listen, we're going to pull the plug. If that scenario occurs, it's either that there's something wrong with all of our research and our models and the portfolio allocation construction, etc., etc. Or, that the world has just changed so much, and it's not working the way we thought it was going to work.
Niels: But when you say that Kim, and to a large extent I agree with you, but I also have to base my own 25 years of experience in this business and say, "Well hang on a minute." Because we've just been through a period of time where managers who've been around for 20, 30, 40 years, some of them saw their worst drawdown get exceeded by 50, some even 100%. Meaning if they were historically down max 15%, suddenly they found themselves in a 30%, plus, drawdown. And so by any standard I think the question would be: Is this still working, and so on and so forth?
As you just described, however, as we also saw in 2014, the very same managers, some of them had their best annual returns ever in their long history, and I guess by now, all of these big managers and the small ones as well, are making new all time highs. So, it's a tricky question because things can certainly get pushed out of whack, and you very elegantly described the convergent decisions by central banks a few years ago, which clearly created an unusual environment for these strategies causing larger than normal drawdowns. But, as a study, I think Aspect sponsored people at Bath University, if you go back 100 years or however long they did, when you've seen periods of unusually bad performance if you put it that way, usually after that you get a period of above average performance. So, it's a tricky one, but I think at least there should be some steps written in the system that says if we go beyond our max drawdown then we certainly can't have additional leverage built into it, or at least it needs to be some mechanism of deleveraging I guess.
Kim: Niels, its… look, I hear where you're coming from and you are indeed correct in the observations. There're so many problems with these kinds of drawdowns because I think that a lot of investors, they really can't stomach… I don't know, probably once you start going much beyond 15% drawdown, and certainly 20%. I think a lot of people get very uncomfortable. I would say that I would also get rather uncomfortable because it's not really what I would be expecting, right? So, you have to have a certain type of investor and a certain type of manager who's able and willing to keep going, even if you're experiencing 25%, 30%, maybe even 50% drawdown. I'm not sure how many managers and how many investors really can pull that off successfully.
I think what we've seen… I haven't studied it in great detail, but I think what occurred over those years is a tremendous amount of assets left. Even these larger managers, really a lot of volatility and assets under management, and I think some of them probably also went out of business. But also you are right, that the ones that did stick it out, they've been able to come back. But never the less I think they came back with much less assets under management than what they had before.
Niels: Sure, and I think that is a very, very good observation and comment, and as you know and probably as many of the listeners know, today I work for one of the most successful managers in the world who have been around for 40 years and who have stomached a 50% drawdown along the way, but who have also outperformed most managers in the last 5 years, coming back and coming out of this. But at the same time, my good friend Jerry Parker at Chesapeake, alerted me to a study, or a comment at least, about systems, in general. I know it wasn't his own research, but something he had read. It escapes me right now what it is, or who wrote it, but the conclusion was, and I actually tend to agree with that very much indeed even though it's counter-intuitive, and that is that the most robust systems are actually the ones with most volatility and big drawdowns. And that's a little counter-intuitive because everybody thinks that if you can create something that looks steady and for all intents and purposes is steady, that that must be very robust. But actually their argument, or the argument in this piece of research is that no, you want a system that has been through rough times and came out of it and survived over a long period of time because they are themselves or have indeed proven to be robust.
So, it's a super interesting conversation, love to talk to you more about that, but right now I know that a lot of people are dying to learn more about your particular strategy, so let me get back on track on that. Give me… you know one thing I've found in doing this podcast, and that is people love examples, and if you're game, give me some examples of what is a trend following model for you? What is a mean reversion model for you? And what do they look like? You don't have to be super specific, but what kind of indicators or whatever you might use goes into creating these models?
Kim: Yeah, great Niels, I think you teed it up very interestingly because there is something very powerful in simplicity and the simplicity of a model. I think there is a direct correlation between the simpler a model is and the more robust it is. Also if the simplicity of the model that cannot be applied, and in our case we insist on that, that it can be applied across multiple sectors and markets, and it has to be able to show that sort of robustness and consistency. I think all of those things lend themselves to giving a higher confidence level as opposed to a model that's very intricate and very targeted to a particular sector or market, right? So, I think there's a lot of truth to that.
Now, the conundrum is if you make it very simple, and I believe that the earlier models were probably very simple, and they worked, and they continue to work. But I think that the problem with these very simple models is that they have a lot of volatility in them. And so as we spoke about, I think the industry has just become much more competitive. I think the investor is more demanding, and they want something that has better risk-adjusted returns.
So to your point, can you produce models that are indeed superior, have better risk-adjusted returns, but have the same underlying robustness of the very simple models? I think that's an excellent question and a very interesting debate. So look, we tried to tackle that. It's a fine balance because there's no doubt that the more intricate you get, the better hypothetical results you can show, but the question is how well are they going to hold up going forward? So, part of the things that we try to do is that whatever ideas that we introduce into these models, again they should hold true across broad sectors and markets, and that the underlying idea should be… should we say, sort of a value-added component that shows its value again across multiple sectors and time frames, and markets. Because then the conclusion is, at least from our point of view is, the conclusion is, okay there's something here that appears to be robust and valid, and there for it earns the right to be part of the model.
So, I'll give you an example. So, you can look at market sectors and markets as having certain underlying cyclical tendencies, right? And you can, therefore, build the quantitative models and indicators that reflect the underlying cyclicality in each sector in each market, and you may be very comfortable with that, right? Now, we do not do that. It's partly I think that the underlying cyclicality of the commodity markets has somewhat diminished over time. I think that maybe historically there was more underlying cyclicality as the markets were more local and more concentrated, but many of the markets have become more global. So the demand and supply is more spread out than it was in the past and hence, perhaps some of the underlying historical cyclicality that existed particularly in the commodity markets is not as prevalent as it once was.
So, we actually subscribe to applying our models on a non-cyclical approach across these various different markets and sectors. Now, another question that comes up is, "Okay, so you trade across multiple time frames, well, how do you pick the time frames?" Should you optimize these various time frames or should you not optimize them? You know, what's your approach? Right? So our approach was again not to optimize the time frames and actually we are pretty indifferent as to these time frames, but where we are most particular is in the spacing between the time frames, right? Because if you align one-time frame very closely to another time frame then you're really not getting all that much diversification between one or the other because they're so closely aligned.
So for us the most important thing was the spacing of these time frames, and because we don't believe as much in the underlying cyclicality of the markets, and we don't want to differentiate one market from the other anyway. We're completely sort of arbitrary about the time frames that we use in the models. That time frame is the same that we use across all of the various sectors. Now, the other thing is that when you look at these components and these mathematical indicators, you can have them be static in nature and moving average could be a 50 bar moving average as an example, right? Maybe a moving average can be parabolic, and it can have a different type of characteristics, but the number of bars is pretty stagnant. I think that is probably true for most of the models out there, and we like to think that's true because this is a point where we think we have an element of differentiation. In other words, we will split our models across these relatively arbitrary time frames, but within the time frame we have an underlying dynamic component so that this 50 bar example I gave you, could be 50 bar sometimes, maybe could be 40 bars, or 60 bars, it has varied ability. So, we have underlying components that help make these primary signals dynamic in nature.
Niels: So let me try and understand that. Sorry to interrupt here, let me try and understand that. So, if you're trading today, you're models are choosing a 50 bar breakout model, just to use that as an example, that would then apply across all markets. But there is some element to that model, and maybe it looks at historic profitability as an example so that two weeks from now, instead of using 50 bars, it might only use 45 bars.
Kim: I'll give you more specifics in terms of how does it determine whether it should be 45 or 60 bars? For us it depends on the underlying volatility of the markets, and that volatility, of course, can be expressed in many, many different ways, but we have an approach to how we determine and calibrate volatility in the most basic concept. We look at near-term volatility versus some historical volatility, and that gives us a gauge as to how to adjust this underlying time frame. The objective for us is to try to get somewhat better traction and better accuracy in our primary signals, and we think that we're able to achieve that.
Niels: And so changes in volatility clearly can change the number of bars you change, and I guess… or you look at. And I guess there might be a range so that the longer term models cannot go below 40 bars for example, or I guess in order to keep the distance between the time frames in check, I guess? But, do they not look at historic profitability meaning, if 40 bars were really a bad, profitable sort of time parameter to use then it wouldn't use it? Or how do you balance between changes in volatility and profitability over time?
Kim: So we don't actually look at the historical profitability of the model; we don't. Maybe that's an interesting thing to look at but we have not done that. The whole objective Niels, and I can explain it to you with another example, is that look, if you have a simple moving average model and the market is behaving in a normal way, meaning it's trending in a normal, contained way, and you have a moving average that has a pretty decent fit to this kind of activity, then it's very hard to beat that model in that environment. Okay, it gets you in a little bit late, it gets you out a little bit late, but it really captures the move, and it captures it very nicely. So, that's an environment that's a normal trending ordinary environment.
The problem is, with the market, it's not in an ordinary trending environment, and so you can kind of sort of classify the various types of environments it might be. So, one environment is an environment where it is trending, it's directional, but the market tends to have a parabolic shape. So, it's accelerating on the upside and it's going vertical, and the problem with that market is that if it reverses quickly, the problem is, is that the open P&L evaporates, and the impact on the underlying volatility of the portfolio is enormous. So, you're building up this huge P&L, and then it evaporates. It's a very bad feeling, right?
Niels: Sure, so you need to reduce the number of bars that your system is looking at in that environment?
Kim: Exactly, right. So, ideally you want to be able to say, "Okay, when the market is very normal and behaving beautifully and then you have a certain amount of bars and it captures it very nicely, and then when the thing goes parabolic then we want to have a more snug fit. So, essentially we attempt to do that and at the same time what we also attempt to do is we attempt… or this is not an attempt, it's more of a money management thing really, is that we reduce our open positions, we reduce our position sizing. So, we'll look to scale out, we'll look to scale out and take profits, and it's not because we know how to pick tops or bottoms, but it is really to try and deal with this underlying problem of… it's really to try to mitigate the enormous volatility you can have on the portfolio in those kinds of situations.
Niels: Sure. Can I ask you a question? Is the same true that if you get into range trading environment for these kinds of models, that actually the best thing to do then is to increase the number of bars in order not to be whipped around too much?
Kim: I would say yes, there's an element of an attempt, right? So, yes, we try to do that. The other thing we try to do is, we're actually flat, our capital utilization is very low. We run an average of around 15%, which I think is pretty low, and the reason for that is that we're out of the markets a fair amount of the time. We are probably out of the markets about 1/3 of the time. So, part of it is sort of the sort of scenario that I just explained, that the inverse of it is what you teed up is that what you do in sort of sideways trending, sort of choppy environment, and the holy grail is to stand aside and not do anything. Niels: Sure, no absolutely, and you can do that by combining say, if we stay in the moving average space, you can combine 3 or 4 different oscillators together and obviously some will be long, some will be short, but net net you might be flat overall, and obviously that's one way of doing it. But that's fascinating, I appreciate that. I hope people are paying attention and making notes because that's a real good insight from your side. How do you… I don't want to talk risk management just now, but I do want to ask you whether you use models that have hard stops or whether you only do as we just discussed, sort of change the parameter sets dynamically in order to perhaps contain, control the risk better?
Kim: So Niels, I was saying on the directional models, they all have pre-determined stops, stop loss. So really the only ones that don't are the mean reversion strategies; they don't run stops.
Niels: Okay. So tell me a little bit about mean reversion strategies in the context of what you do. What have you found that tends to work better? Because a lot of obviously mean reverting strategies are really known for the ones where you pick up pennies but then one day you get run over by a train and then you know, you're out so to speak. So how… and it's against the philosophy of trend following which is interesting as well. So, how do you marry the two?
Kim: Well I think that's the attraction, right, is that they are really opposite?
Niels: Yeah, opposites attract, as they say!
Kim: Right, and they behave completely opposite which is the beautiful thing because the directional sort of trend following approach is one where you see the spikes on the equity curve. You see it sort of meanders around, and then you get the moves, and you see the equity curve increases sharply, you see the beautiful pure Asymmetrical style of the directional volatility markets, right? And that's why we are CTA's, we love the Sortino ratio, instead of the sharp. And the mean reversion are exactly the opposite, you make, you make, you make, you make, you make, you make, you make, and then boomp! You get hurt, and then hopefully you make, you make, and you make again. So, marrying the two has a real attraction.
Now, what we just talked about, sort of the reducing positions sizing and scaling in and scaling out of the markets, for us that is a baked in mean reversion strategy. It doesn't fight the trend, but it reduces trend exposure. So, to us we call that a sort of baked in mean reversion because we're taking chips off the table, whereas a hardcore trend following trader will ride that thing until it's going to go from 100 to 0 or from 100 to 1 million. So, for us in the core models, we have that baked in. And then we found across the financial markets, that pure mean reversion strategies, stand-alone mean reversion strategies, actually can work quite well. They are strategies that we'll look to buy into weakness or sell into strength, and as you sort of explained is that they have a high probability of success, and they take out sort of nice chunks when they're successful. When they're not successful, they take a disproportionately larger chunk, but they do it much less frequently. And it's just the nature of them because we trade the same markets on the directional volatility side. So if we are really sucking wind on a mean reversion strategy, we tend to mitigate it or even make it up for it on the directional side. And hence also a little bit the weighting because sort of our core is really looking for those directional Asymmetrical moves.
Niels: Performance wise, do the mean reversion strategies deliver 30% of the overall performance? Is it that simple over time? Or how much performance do they actually contribute compared with the size of allocation?
Kim: That's a great question, and I can't actually answer that directly because I haven't looked at it quite that way, and I think that perhaps we probably should, but I can tell you the way we do look at it. We look at it very much in the framework of the portfolio. So, again you could make a very compelling case for having models in certain markets and sectors that actually don't make money, but they have such a lower inverse correlation to all of the other activities that you engage in, that it helps mitigate your overall volatility of the portfolio, and, therefore, becomes a very valuable player.
So, that's really the way we look at it, particularly with regard to the mean reversion thing is because when… for instance, so let's take an example, so the 2008 financial debacle and subsequent the sort of decompression of the financial markets and the lowering of volatility over a multiyear period and sort of somewhat absence of directionality. The mean reversion models were very effective in carving out a steady return over that period of time. So, they really helped mitigate when you're always placing bets and looking for that directional move, and it's never materializing, so you keep… you’re bleeding, right? You keep bleeding your betting, your betting, and your bleeding, and you bleed, and the direction of mean reversion models were very helpful and sort of mitigating that. So, we look at it in that sense because we just find that they're very helpful in reducing the overall volatility of the portfolio and providing for a steadier, maybe higher confidence level of a constant sort of return stream.
Niels: Now you mentioned before if I’m not remembering incorrectly that mean reversion models don't have a stop, and of course we just experienced a so-called black swan event in the Swiss franc a couple of months ago. I know from your newsletters that you were not involved in that, but it does raise the eyebrows when it comes to risk management on mean reversion trades, and we do know that a lot of people got badly hurt during that time. Do you not worry about not having a stop for a mean reversion strategy?
Kim: Neils it's certainly worrying, and that's also why we dug in and we wanted to do some analysis post to the Swiss National Bank's surprise move and the subsequent enormous revaluation of the Swiss franc. Yes, does it make me nervous and uncomfortable? Yes. You know I think that what we try to do is as we talked about earlier, is we talked about processes and portfolio construction and risk allocation and what we spoke about is that we have strong beliefs about allocating a certain amount and broadly allocating across various sectors and markets and evenly allocating across them and a certain particular allocation across the various models and the models are running across various different markets. So, every individual single market gets a relatively small allocation, right? So, if you have a blowout in say the Swiss markets, yes you take a hit, but if you take a hit in a couple of percent or 3% or maybe even 5%, you still live to fight another day. You can't have the kind of concentration. I mean if we're talking about long-term capital management, where you're talking about 30x, 50x leverage, and you essentially are engaged in mean reversion strategies with that kind of leverage, and you go through a black swan… I mean that's exactly what happened, right? And then you're gone, but keep in mind what I told you is that our leverage is minimal. I mean our capital utilization rate is on average 15%, so we're trading on a leverage of maybe 2, 3 times. It's tiny compared to those kinds of classic mean reversion strategies.
Niels: And I do accept that and my understanding is also of course that you're taking off chips from the trend following part so in theory at least, you could say that you might hurt on the mean reversion trades in a situation like that, but hopefully then some of the trend following or volatility directional trades would be kicking in and help you out. So, I do accept that as a concept. Although I would say, I think people forget… I mean there was a lot of damage done in the markets for that one event, and I think people forget that that was just one market. I mean imagine if it was the Euro breaking up and you had 17 different markets going wild at the same time because it wasn't just the currency, I mean equities in Switzerland got hurt significantly and so on and so forth. So, to me at least it does happen more often than people believe and anyway, I do like having these I guess, very quantifiable risks even though sometimes performance wise it's better not to use directly stops on your positions. Anyway…
Kim: Niels, in general, these kind of events are the kind of events that we live for.
Niels: I agree.
Kim: That's typically where we have our best performance.
Niels: Sure, and certainly January turned out to be just that, and I think this month's not looking too bad for most trend followers either. So I think that's good news on the way.
We've kind of already done that sort of digressing into the risk management side, which is very important I think generally speaking. I just wanted to ask you one thing from an overall point of view, in terms of risk how have you found best to define risk? What is risk for you? Is it the value at risk? Is it the standard deviation? How do you quantify risk?
Kim: So I think that we look at it partially from a top down and partly bottom up. We talked at length about portfolio construction. I think that we're big advocates of having a process and the process essentially embodies almost a belief in things such as, it's better to be trading eight sectors in 60 markets than trading one sector in two markets. It's better to have an even allocation across many sectors and markets than have very concentrated. So all of those types of things that, even though you might be able to construct a portfolio and then models that have highly concentrated and show exceptional historical performance, there's still a more humanistic decision making that comes in and says, "OK I just don't care. I don't care that this thing can return 1000% with !% drawdown over five years because it's basically one market. I'm just not comfortable building a business around trading one market."
So that's overall our viewpoint, and we embrace that. Underneath that we try to then justify it by applying the models, running the correlation analysis and seeing what the picture then might look like. If we think that picture is an acceptable picture in terms of the risk-adjusted returns and so forth. Then what we do on an individual level when we actually place an investment, as I said, probably about 80% of the positions have a predefined stop. We use the underlying volatility in that particular market to determine the position sizing which is a function of the overall assets under management. It's a function of the overall leverage of the portfolio and what constitutes the available risk capital to that particular market. So that's how we look at it.
Niels: I want to jump to another section, and this one will be a brief one for you because...
Kim: Niels, if I may just interrupt you on that former point. I think that perspective is fairly common but never the less unique to our industry. Because if you talk to a bond manager, or a stock manager, or a long/short stock guy, he'll tell you something very different. He will allocate outright a certain amount of assets or risk to a very particular trade or investment. Then he will do bar analysis and he'll do various shock analysis on this portfolio. He will come up with a profile - a risk profile to present to a potential investor. It's just very different the way that we look at it. We look at it sort of as a stream of serious bets and all of those bets we're willing to risk a certain amount on those bets which are typically constant but across very different assets. I think this industry has a unique perspective. I just met with a potential investor last week and we got into this discussion and I actually found it quite challenging to explain and to justify the way that we view risk management and how we look at it on every position sizing, etc. etc. He's was from a more traditional bond investor perspective, which has a whole different way of assessing risk.
Niels: Yeah...Did you convince him?
Kim: We'll see. I'll let you know. I hope so.
Niels: Yeah, yeah. I wanted to jump in and ask you about drawdowns because I think that having people talk about this is quite important and also I think actually for investors in this space it's very important to hear experiences. The problem with you is you haven't really had a drawdown yet, so I'm going to rephrase my question. I'm just going to drill it down to basically a couple of questions. I know that you expect at some point to lose between 15% and 20%, that's what your research shows you, but you haven't done anything even close to that in real trading. How do you prepare for this drawdown? Mentally that's going to be a different situation to be in.
Kim: You're right. I think, Niels, I guess it's almost impossible to answer before you're actually in it because I really do appreciate where you are coming from. I guess I would say that in my career a number of times I've been trading and trading for banks and financial institutions and trading for my own and so forth, so I have experienced drawdowns on an institutional level as well as on a personal level and you have to have a strong conviction. I'll tell you a story on that. You will appreciate it, I think. When I tried this business first time around in the early '90s, I went to one of these conferences. I saw John Henry was there, and I walked up to John and I said, "Listen, John, what advice can you give to an emerging, aspiring money manager of a CTA?"
He says to me, “Do you believe in your models?"
I said, "Yes."
He said, "If you believe in your models, never ever deviate from the models. Follow them."
What he's really saying is, when things look rough, and you have that drawdown, which you invariably will have, that's when you're going to be tested. If you really believe in your models and the work that you've done, you stick to your guns, and you'll do OK.
Niels: Yeah, which goes back to what I was trying to explain earlier on about the observations about the most robust systems actually are the ones that have been through the drawdowns and been there, done that. I truly believe that.
Kim: John Henry is a particularly interesting one because he has experienced so many drawdowns and severe ones, and invariably has managed to come back. It's an amazing thing.
Niels: Yes, but there are few people like that and everyone who has been around for 30, 40 years have been there and done that. I guess, one thing I would say is, in answering the question, which I think you did perfectly. It's about your experience. You're not the typical new CTA. Most young CTAs that we see start up, they start up when they come out of school almost and probably have very little experience to draw on once they go through their first drawdown. You have something which is equally as valuable, which is life experience from running and managing businesses. That's also stressful from time to time, and that's certainly no different from doing what we do.
Let's just jump to the next topic I want to touch upon. We've been chatting already for a couple of hours, and I want to just touch upon research. Clearly your son, Oliver, and you are doing this together. I can imagine that there are a few conversations over dinner about work as well. When you are together and in a research mode what are the... This I guess is another challenge when you have a small team, and that is how do you challenge each other to come up with new ideas and so on and so forth? I guess maybe the experience and the input that you got from the NYU group will have been very valuable in terms of new ideas and new ways of looking at it. What is the typical research conversation you have between the two of you when you sit down and talk about new things or things that you want to maybe explore and prove?
Kim: Niels, like yourself, you've been in the business for a long time and I've been in the business for quite a long time and I've had many clients and many clients in this particular space. I have observed so many different strategies being deployed over the years by these many successful money managers. I would say that the list of research projects is very, very long, actually. There's no shortage of research ideas and opportunities to explore. I think that the issue is one of prioritization and fit - does it fit inside of the overall portfolio that we want to run? So the idea, when we started introducing these mean reversion models were because we knew that the underlying behavior of the mean reversion model was essentially directly the opposite of directional trend models.
Before we started, we thought, hmm, it's pretty cool. If we can have an edge on both sides of that equation. If we can have an edge and then combine them, if things match up perfectly which they might, more or less, then you should get a better risk-adjusted return. Then you can leverage your portfolio proportionately, and you should be able to extract somewhat better returns relative to the risk that you take. I think it's along those lines - you sort of imagine what a particular strategy might look like and how it might fit into your overall portfolio, and if it adds, not just return, but what's the underlying correlation of that idea. If it's somewhat inversely correlated or low correlated, it becomes an attractive proposition. I would say we have a long list of ideas. The thing is to develop an idea and to thoroughly construct it and to thoroughly stress test it and to thoroughly get it to a point where you believe that it's robust, and it can be introduced. It actually takes a long time. I discussed this with my son and I've challenged him to say, look if you can come up with a new idea - a fully fleshed out idea and model once a year, that would be a great objective. I think it takes at least that.
Niels: Speaking on that topic, that is something that I often get asked by the listeners, and that's actually the converse situation, meaning how do you know that one of the models that you already have in the portfolio has stopped working?
Kim: OK, so we monitor the theoretical performance and the historical profile on a real time basis with the live models. We look for any deviation from expected behavior and profiles. That's more prevalent and obvious on the shorter term models just because of the velocity of trades are much greater. It's somewhat less in the longer ones - much less in the longer ones. I think actually the longer ones, when you're talking core directional trend, more traditional type models it's almost impossible. You either believe that this thing is going to work over time or not. It's very difficult to monitor. On the shorter term models, you get a lot more data points. You can observe it sooner. Niels: One of the final topics I also want to touch on is just a little bit on the business side, and it's kind of random the questions that I sometimes pick depending on how our conversation has been so far. I think we've touched on so many things already, so I just want to ask a question that I actually think is relevant for managers to consider right now, and for investors to be aware of as well. Since you mentioned that you're moving towards having a fund - you're launching your first fund. You were mentioning that 15% of the money you get in your fund will be put to work through margin, on average. That leaves you with 85% of the cash. In a zero interest rate environment, where potentially there could be some significant risk, maybe not obvious to the eye in the fixed income markets at these historical low-interest rates. What are your thoughts about what you're going to do with the 85% of the money that you don't need for managing the new fund?
Kim: Niels, that's a real conundrum. Historically it has been one of the areas that have actually added a very nice cushion and a very consistent return profile to our business because typically the money management business is the inverse. Typically it requires leverage. Therefore you have to have a cost of capital. Most investment strategies have that characteristic. Ours is the opposite and it's a beautiful thing because historically you could put the money into T-Bills and you could pick up a couple of percent returns back in the '80s, maybe 5% - a lot of really good returns with hardly any risk.
So you're right. What do you do today because there's no return to be had? If you were in Euro, you have an inverse problem, so it's costing you to keep the money in cash, so what do you do? What we have done is we think there is decent value and decent stability in short term corporate bonds. So we've actually put monies in short term corporate bonds, one to two, three-year duration. We've essentially bought some ETFs, and they pick up a couple of percent or so and we think that's a reasonable area. Corporate America is doing quite well. Maybe you could argue that there's less risk in corporate America than there is in government America. At least you get paid a little bit for owning those papers. So that's where we put the monies. We haven't invested up to the full 85%. I would say we have been very patient to deploy that. We've had a mean reversion perspective on that. If they dip a little, we'll buy some.
Niels: Yeah, absolutely. Now, you're at an interesting point in your business because you already have probably had lots of conversations with potential investors, due diligence meetings, due diligence phone calls, and no doubt there will be many more to come. I wanted to ask you what your observation is so far, in terms of those conversations. What are the questions that investors should be asking you but they're not? Do you know what I mean? What should they really drill down into understanding what you do, yet they may not actually go in that direction in the way that they phrase the questions they have for you?
Kim: Right, so...
Niels: What should they be asking I guess is the short way of saying it?
Kim: Well I would say, look, I haven't had all that many conversations because I would like to think or hope that we are sort of at a flexion point where more investors are going to be interested to learn more about what we are doing. Whereas, up until this stage I would say the majority of communication has been one-sided in terms of providing information, as I told you through monthly newsletters, and illustrating how we are trading and so forth and so on.
We have had conversation, but most of the conversations we've had I would say are with these early adopters. Besides family, friends, business associates, which have a certain emotional component in terms of why they give you the money, now we're into the category of say the early adopter. The early adopter is very savvy. They really, really understand the space and they really grill you left, right, center, and all around. I'm not sure that they really leave any stone unturned. They're pretty thorough.
I think that when you get into the broader investor base, let's say family office, small institutional clients who are not entirely exclusively focused on investing in this space...The thing that I stress is this idea of this is an attractive potential return stream. It's an attractive risk-adjusted return stream, and that's a another very attractive component is that it has a very low correlation to your traditional investments in stock, bonds, and say real estate. So that's a very attractive high-level situation.
I actually think that they less sophisticated that you are, the more basic the conversation is, and I actually don't think that there is anything wrong with that, meaning the investors should understand what is my risk. We talked about at some point do we pull the plug? I say to the investor who is not a professional in this space, I say, "listen, you're initial risk is max... if you are going at the worst possible time historically over the last 34 years, maybe you could be down 20% of your initial investment. You have to be willing to stomach that." We think that we can mitigate that risk over the first 12 to 18 months. We think because our targeting return profile is somewhere between 15% and 25% annualized. I will say to an investor, "listen, if you can hang with us for 12 to 18 months, we think we can mitigate most of that initial risk."
In many ways, Niels, I think the simpler you keep it, the better it is because you can very easily get lost in a lot of intricate details that are very interesting to you and I but to an investor it just sort of glazes over the eyes. Then the risk goes into a category of I just have no idea what they're talking about. I don't understand it. It makes me uncomfortable, and I'm not really going to do this.
Niels: I do agree that many of the early adopters they are pretty savvy, and they ask good questions, so I think that's right.
We're going to jump to the last section Kim. We're almost there; there's light at the end of the tunnel here. It's one of my favorite categories or sections, which is the one I call general and fun, so not a lot to do specifically with any models or systems but just something that gives people a little bit of color and hopefully one or two good insights to take away as well. I wanted to ask you, just to start off with; you tried a lot already in your career. Along the way, you may have read a few books that have guided you to take some of the decisions that you've taken. If you were going to recommend a book or two for people to, could be either improve their trading or improve their business, which books would you say have had the most impact on your career?
Kim: Niels, I've been around for a long time and I've read quite a few books. The early books with Welles Wilder, some of the Larry Williams books, Jack Schwager, the Top Traders is always very inspiring and interesting. I think a more recent book called the Quants (Scott Patterson), and then there's more specific books that have more detail around - Ralph Wintz that goes into the optimal F risk in a portfolio and so forth. They get much more technical in nature.
So I would say it depends on where you are. If you're looking for an introduction and you want to understand this business from a more general than a higher level and you want to get inspiration from other people who have succeeded in this space, Jack Schwager's books are probably some of the best. Then I think if you're really, really, seriously on a professional level I think it's good to read a lot of these books, but I think you have to toss them out. You just have to toss them out and you have to sit down and you really have to think critically about the markets and the microstructure and how they work and the psychology around the markets and why markets behave the way they behave.
Then I think you have to think about how to capture that behavior mathematically with some sort of logic, and then you can maybe go back into the tool kit, into the books and stuff and you can go in and you can maybe extrapolate out maybe data or a concept and you can say, OK, how can I use that tool to try to uncover this opportunity? Can I use that, or do I need to come up with something completely new to try and capture this idea. I think that's got to be if you want to be in this space - the cutting edge, and you want to compete, you have to be somewhat original. Not completely from scratch, but you have to be somewhat original in the way that you construct your models and the way you try to solve the problem of extracting earnings.
If you could buy something off the shelf and just plug it in, and then off you go, that's just not quite realistic. But there are also a lot of components. There are so many sides to this business. You could develop an incredible alpha model, but if you don't understand the risk components, like how to manage the risk. If you don't understand how to construct a portfolio properly then you're probably not going to be very successful. So there are so many different components and I think a lot of the books will help you think about all the various different areas that you need to consider and that you need to incorporate into the management of investing, trading in the markets, and building portfolios and managing monies. At the end of the day, you have to think about it critically yourself, and you have to construct it from that vantage point.
Niels: It sounds like, on my side, unfortunately, Kim, we're losing a little bit of our crisp connection here, so I'll try and wrap up with the last few questions before we lose the Skype connection entirely. I wanted to ask you, as you build your business, are there any other CTAs, bigger ones that you look at to aspire to? Are you looking outside and saying, "Prolific Capital, I would love for them to be like so and so in a few years time?"
Kim: I've had many of the successful CTAs as clients over the years, and I know that one of the clients you interviewed whom I think is a really brilliant and an original thinker...
Niels: I can guess who it might be.
Kim: Do you want to guess before I say?
Niels: My guess would be Roy Niederhoffer.
Kim: Yes, you're right, exactly. So Victor was a client of mine early on, his brother, and subsequently Roy. Roy and Victor worked together back then, and they were clients of mine and I had a lot of interaction with them. They're just really original independent thinkers. I think that's very inspirational, and I think that it's pretty cool what they do. Somebody who came out of that group who was also a client of mine who was extremely successful was Trout - Monroe Trout. So very interesting. I had encounters with Paul Tudor Jones and he has also subsequently been a client and it's just a very inspiring background where he comes from a cotton pit trader and developing the business as successfully as he had. Bruce Covner, there's a lot of these guys that are really quite admirable in terms of how successful they have been. I think there's a... to aspire to any and all of those guys is pretty cool.
Niels: With those connections, Kim, feel free to invite them to come and talk to me on the podcast.
Kim: I'll give it a try, no problem Niels.
Niels: Clearly you son Oliver works with you. You have three other children, as you mentioned. If you could pass on just one of your own skills to your children, what would that be?
Kim: Interesting, OK, I think that in life it's a very good quality to be open minded - open minded with a positive perspective. I think in life you want to try to position yourself for success, whether that is in your professional vocation, personal relationships, if you are competitive in sports, or if you are in this business which is a competitive business. You want to be open minded and positive, and you want to be able to take in whatever comes your way in life. You want to be able to take it in with an open mind and with a positive perspective because when you do that, your brain has a much greater capacity to process those inputs, and to organize them so that you can position yourself around it for the most successful outcome. So if I were to pass on something to my children, I would say keep an open mind, a positive mind, and use those circumstances to position yourself for future success.
Niels: Great advice Kim. Can you tell me before we round up, can you tell me a fun fact about yourself? Something that even people who know you may not know about you?
Kim: I don't know if anyone would classify me as a super fun guy.
Niels: I didn't mean necessarily that you do standup comedy in your spare time, but something... it could be a talent, it could be something that even people who are around you may not know that you enjoy doing or not enjoy doing for that matter.
Kim: OK so, actually I didn't mention this but since we're having this conversation from your home in Switzerland, my Dad is in Geneva and my brother is in Geneva, Switzerland, and my brother is an avid Polo player. So when I go to Geneva I invariably, he convinced me to get on top of a horse and play a few chukkas with him. So a secret is that I grew up on a farm in Denmark with horses. I actually in turn raise horses, so I do know how to ride. It's not something I do every day, so you can imagine the challenge of going over there for a periodic visit and getting on top o an Argentine polo horse and playing a few chukkas was kind of an interesting thing, and it is a comical thing when I land on the ground.
Niels: That's where the fun part comes in or the fun factor comes in. I hope people caught that because I can hear on my side that our connection is breaking up. Before we finish, and this will then be my last question, I said earlier that I think it is important that investors ask the right questions, so I'm also going to turn it on myself and ask you if I missed something today in trying to cover all the ground about you and what you do with Prolific and if there's anything that we need to cover towards the end here in order to do justice to you and your firm?
Kim: Niels, I thank you very much. I think you've been super thorough. Normally with most people it's an elevator pitch - 30 seconds and you're out the door, so having the opportunity to talk about what we do for an extended period of time - I think we've covered a lot of ground and a lot of detail and I appreciate that. Look, there are other things we can talk about, but I think you covered the vast majority and hopefully gave our listeners a good insight in to what Prolific is all about.
Niels: Sure, absolutely. Of course, we have to thank Marc Goodman for making the introduction. Most people will know Marc as one of the founders of Kenmar back in the day. So I want to put that on record that, that was very kind. This has certainly been a great conversation. Not just because we have our Danish history in common. It has been enlightening; it has been... I think a lot of people can take a lot away from this conversation on many different aspects, so I really appreciate that. I appreciate your willingness to be transparent about a lot of things which is something that not everyone is comfortable with, so with that in mind Kim, I want to say thanks very much, and I want to ask you where people can best find you and learn more about Prolific?
Kim: Well, we're located in New Jersey, and I guess the best way to contact us would be through email, and you can reach us on firstname.lastname@example.org.
Niels: Great, and of course I will put all of these details in the show notes for this episode on TOPTRADERSUNPLUGGED.COM. I hope we will be able to connect later in the year and hear about the great work that you are doing. Of course, at some point perhaps even introduce Oliver to the podcast as well, so all I have to say now is thanks so much and hope to speak to you later.
Kim: Thank you very much Niels. Thank you for your time.
Niels: Your welcome, take care, bye bye now.
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: 31 Mar 20154 comments