Kathryn Kaminski is deputy managing director at the Institute for Financial Research in Stockholm (SIFR) at the Stockholm School of Economics, Sweden as well as a contributor to CME Group. In 2014, she and Alex Greyserman wrote the book Trend Following with Managed Futures: The Search for Crisis Alpha.
In episode 282 of Trend Following Radio, host Michael Covel interviews Kathryn about her career and book, discussing survivorship bias, social networking as an example of risk, the importance of failure, about looking at markets like ecologies, and 800 years of trend following.
Read below for some extracts from the interview, or listen to the full episode here...
A good friend of mine was giving a talk about success. And what he said was, you must fail and fail and fail to succeed. This sort of idea of cut your losses…just fail smartly.Kathryn Kaminski
On her beginnings...
I’m originally from Nashville, Tennessee. And I’ve always been a huge fan of math, ever since I was a kid, I sort of had this pull towards quant. Then I ended up going to MIT and doing my undergraduate in electrical engineering. Then I went on to do my PhD at MIT in operations research. I worked with a professor by the name of Andrew Lo, who was also quite well known in the quant space.
On Convergent - Divergent...
I think this a fantastic way to understand what trend following is really about, in the following sense. In life you can really step back and think that pretty much everything we do is risk, and sometimes these risks are more quantifiable and sometimes they’re more about uncertainty. I think you need to use a combination of approaches to deal with any of the risks that we face on a daily basis.
So how would I explain the difference from convergent and divergent is that it depends on how you view the world. If you see the world as somewhat knowable and stable, quantifiable, then you’re already sitting in a framework where you already have some sort of beliefs in a system. And when you’re in that frame of reference, you’re more likely going to take your gains, but when you are wrong, you’re actually going to double up on your position because you’re going to think this goes against what I believe and think should happen.
On the other hand, divergent risk-taking approaches are more for the philosophical idea that you say I don’t really know what’s going to happen, I can’t predict the future, I just know that the structures of risk are too complicated for me to really make a discrete view about the world. In that case I’m going to use a set of approaches that are a little bit more agnostic to predicting the future.
So when you’re a divergent risk taker, what you do is basically; when you lose, you don’t necessarily believe that you should follow that particular strategy so you cut your losses. But when something goes you way, you follow that train. And this is exactly what trend following really is. It’s really about spreading lots of risks across many different positions where the intrinsic part of cutting the losses and following the winners give you this divergent risk taking strategy.
A good friend of mine was giving a talk about success. And what he said was, you must fail and fail and fail to succeed. This sort of idea of cut your losses; just fail smartly, in situations where things are really uncertain that’s when this divergent approach is often much better. Because you’re not too stuck in your ways, you don’t have a world view that is basically being challenged.
On her first aha! moment...
I think it was a long time coming, but I remember the first time that I was very interested, especially for the paper Crisis Alpha I wrote later on, this actually happened as a grad student. I was a student of Andrew Lo, so we were going and learning the adapted market hypothesis, and he was writing a lot of stuff about adaptive markets and seeing things in a different way than efficient markets.
I remember a good friend of mine was talking about how well technical analysis worked in currency markets, I thought that was so fascinating, because for me I went, wait a minute, you’re using the most liquid, most efficient contracts to trade using technical indicators, isn’t that sort of at odds with the theory? Because if you believe in efficient markets, you should never be able to use past prices or information to trade. And how come it happens to be the case that these strategies it works there? Something weird, and I remember that moment. Then several years later when I was studying trend following and thinking about Crisis Alpha, what happened was people kept asking me, why does trend following work in 2008 or in a crises?
And I thought, well you know if you go back to the efficient market, trend following should not work ever based on that assumption, but if you think about adaptive markets, it’s really about the composition of players in the markets and the environment that controls who is going to succeed and fail. So if you think about the world from that perspective, I thought, wait a minute, trend following is applied on the most liquid, most efficient markets out there, so it must be that this efficiency and ease of trade is at the core of their advantage in a difficult crises scenario. Basically I went back and did a lot of analysis and realised, actually this argument makes a lot of sense, when the market changes and there’s a crisis situation, this is exactly the moment where the environment is challenging for most players, whereas trend followers have the ability because they are diligent in the way that they apply risk to profit from this sort of odd difficult scenario.
On adaptive market hypothesis...
Efficient market hypothesis basically says that all information is incorporated into prices. So what this means is that you can’t go out there and do some analysis and make money off it because the market is so efficient and so many people competing that there is no opportunities. Now adaptive market hypothesis comes more from a biological view of the world. It's more about thinking about the world as an ecology and the market as an ecology.
We can think about it in that perspective, everyone who is participating in the market, so that can be hedge funds, HFT traders, pension funds, individual investors, everyone are sort of members of this market ecology. So it’s a combination of the distribution of players in this market ecology and the opportunity sets and resources that are available which define sort of how the ecology evolves. What Andrew has built up in his work is the framework where we can see markets more like ecologies, so therefore what we have to do is step back and say ok, what is the distribution of players in this ecology and what is the certain sense of environmental factors we are dealing with and how does that going to play out if we want to see who is going to succeed and who is going to fail?
That’s the sort of way you want to think about it, it’s more about biology and ecologies and understanding the forces of natural selection, competition and resources/fighting for resources will help you understand better how financial markets work.
On trend following during crisis alpha periods...
What I found is that on an average day, it’s not that obvious that trend following is going to make money. If you look at the return profile of trend following they have lot of losses. But the difference is that this strategy is poised and ready to move when markets diverge. So in the book we talk about divergent risk taking and then we say that trend following is not long volatility, its long divergence.
And what divergence comes from is actually the reallocation of resources and forces within this this ecology that is the market. So from time to time there are moments where things happen in markets and it takes longer to adapt than efficient market hypothesis would predict. And we can take 2008 as a perfect example of that, in the adaptive markets world, this is something which Andrew and evolutionary biologists would call a punctured equilibrium. Well what does that mean? It means when an equilibrium gets destabilised it takes some period of time to re-stabilise itself into a new equilibrium. And divergence is sort of the result of that. What does divergence mean? It’s when prices are somewhat more predictable than a random walk for example.
On defining Crisis Alpha...
Crisis alpha is the alpha opportunities that are available when there is a financial crisis or severe dislocation. What I would say is trend following as a strategy makes money outside of crisis well, but crisis alpha is called alpha because nobody else seems to deliver it. It’s not something that’s easy to capture without implementing the strategy as a whole. Because several different investors have asked me, "I know I can just change my positions right before the crisis". And I just laugh because no, when I was asked to explain crisis alpha, what I said is, we have to step back and think about the environment of a financial crisis.
There are three things that happen when there is a financial crisis. Firstly, when equities go down, people lose, in aggregate you’re sitting in a world where people are losing money. And if we believe that behavioural finance has an impact on markets, this is going to be the time where it’s the most severe. So people are going to be driven by their behavioural biases more during crisis than other times.
Second of all, most of the rules and risk management techniques we have today are contingent on loss rules, draw down, and increases on volatility. So we have an embedded natural reaction in our portfolios regardless of whether or not we consider ourselves rational or irrational during this type of moment.
And finally, we all know that we have a home bias towards equity markets and whenever equity markets move, it tends to have impacts on other markets as a result. So other asset classes like commodities or interest rates. So what we are talking about during a financial crisis is a period where the most amount of people are somewhat driven or forced into action in markets. If you think about that in an ecology perspective, things are really behaving in an erratic way during a crisis period, which means there is more divergence in markets. And when there is divergence, a trend following strategy is basically set up to profit from divergence in prices.
On 800 year-0ld data...
I think that was an exciting part of this book (The Search for Crisis Alpha) because we wanted to start with history, because we think history frames peoples frame of reference and it builds and basically makes things relevant. So granted the mathematical techniques aren’t as precise with such a long historical analysis, but it’s still helps us to think, you know over centuries and centuries of time, was this a good risk taking strategy? I think we quote David Riccardo and his quote, he basically said "cut short your losses and let your profits run on", in 1838.
What that kind of sets is this idea that is sort of ingrained in human behaviour and trends are just part of financial markets and that has been true since the beginning of time. The different techniques for doing it have become more sophisticated, but the idea is the same. I think that was why beginning the book with history, 800 year analysis, looking at inflation across different centuries and trying to think about that was not meant to be a rigorous academic exercise, it was meant to create a historical and ideological understanding of how sort of innate trends are and how they are just a part of financial markets.
The goal of the book as well was to start with history and move into the science and the art. I think with trend following, that’s a really good way of explaining it. History sort of shows that trends are part of markets. The science is really learning about how you cut your losses and follow winners and building a system and framework for how to apply it. And the art is how you personally apply it yourself.