Best of TTU

But instead of relying on the black box of the brain to interpret everything correctly each day, it followed a rigid set of rules to be able to repeat the process every time it reoccurred.

Niels: And so, in that sense, just to be clear, were you convinced, even at that time early on, that if you were going to be successful from an investment management point of view, you had to be systematic? Was that something that just was obvious at that time? Or was it more of an operational issue where you say “Okay, it's actually easier if I can do it systematically.”

Because the systematic side of things is partly, I think, an operational issue, but it's also an emotional issue, in my opinion.

Mike: Absolutely. The one thing you get from systematic…the one main thing I think would benefit the majority of traders out there is the discipline that it brings to the trading. You can sit there and say as much as you want as a discretionary trader “I'm going to cut my losses at these levels, I've got a bad trade, I'm going to get out when it does this.”

But not everybody does that. And you may do that for a while, but at some point, there's that constant battle where you're sitting there saying “But it's a great position. I've got a great position.”

And I've been reading some books about the financial crisis. And one the things that you see there that's the same as what you had with long-term capital, is just that utter breakdown in risk management discipline. And it's easy for a person to sit and rationalize why “You know what? The risk management wasn't designed for this scenario.” Or “This was a great trade 10% loss ago. It's even better now.” So, I'm certainly not going to get out of it today and just ride it into the ground. So, the one thing, if there's nothing else that you get from systematic trading, is the disciplined risk management approach to trading.

Related to risk management comes draw downs. I'd love to know how you or whether you do predict in your testing and analysis a level of expected draw down for the way the portfolio operates. Is that how you come up with the overall balance to say, “We're happy to do this, because we expect it to have,” as you say, “a certain level of return, a certain level of standard deviation?” I don't know whether you relate that standard deviation to draw downs as well or how you look at that.

Mike: Yeah, so I consider drawdowns a true measure of portfolio risk. Volatility. You can look at 100 specific examples where volatility is meaningless. Certainly an option writing strategy that has no volatility to it does not mean it has no risk to it.

Conversely, one of the examples I give in the book was a strategy based on a time-price series that had pretty enormous volatility in it. It was over the course of a year…I forget what the actual volatility number was. I think it was in the 30-40% range of daily price activity. But it was an extra super safe strategy and highly predictable.

All I was showing was the average mean temperature in Philadelphia on each day of the year. It's something that cyclically pretty much, if you bet in the winter that it was going to go up between then and the summer, you were going to make money, but there was a high level of volatility during that period. There's a number of instances where volatility is meaningless as a measure of risk.

Where volatility and draw down break down in the relationship is where you don't have a balanced truly diversified portfolio. You're just doing one strategy or a sector of markets. Then the volatility that you're looking at is virtually meaningless. To me, Sharpe ratio is an absolutely meaningless measure on its own. It doesn't give you any idea of predictability of returns.

Niels: I couldn't agree more. In fact, over the years we looked at hundreds of CTA track records. What we found as a rule of thumb is that you should expect a maximum draw down for any manager of roughly five times their monthly deviation. If you have someone who is on a 5% monthly standard deviation, you should expect them to lose 25% at some point. As a rule of thumb, that's not too far off.

Mike: I think it's pretty close. If you have the less diversified their portfolio and strategies and markets, the bigger that draw down is going to be relevant to that standard deviation.

It's no different when we're making money or when we're losing money.

—Mike Dever

Download “Trend Following” Ebook
Listen to the Full Podcast Episode
More from Mike Dever
View All podcast episodes

Mike: It's no different when we're making money or when we're losing money. I don't have anything in my track record that shows I will not incur draw down. It's a natural process. It's just part of our program. For example, last summer, in mid-2013, we were in what I was referring to people as a perfect draw down. On an intraday basis, it was 8%. On a month end basis, about 7.5%. Peak the trough, which was at that point in our trading exactly where we should have had maximum draw down. It was a little delayed. It might have should have happened half a year or so sooner by then, so the probability was getting out there.

What we sent out in our monthly report, “Notice that hey, this is a perfect draw down. If you want to invest now, it's a great time to do that.” I'm not advocating timing, because if you happen to have the draw down, great. Take advantage of it. But you don't want to wait for a draw down, because you give up the positive return expectancy you achieve on a month to month basis. At that time it was just a perfect draw down. I guess react one way or another to it. It's just what it is. It's what it should be. I accept that as well as I accept the positive period.

Niels: Yeah. No, no. It's a perfect response, but I guess so few people out there on the investment allocator side will see it the same way as we on the manager side do.

Mike: It's funny, because they'll look at the back testing and the track record, the actual track record, and they'll see there's an 8% draw down. But then when you have it, they question it like there's something wrong. No, it's exactly what it's supposed to be. Nothing different.

07/11/2018 by Niels

If you would like to listen to the full conversation just go click here.

How Bill Dreiss Defines Risk

Niels: Let’s shift gears to a very important topic namely risk management. How do you define risk? I know that you are not a big fan of standard deviation, so how do you define risk in your own methodology?

Bill: I think, again, from my philosophical position I’m very suspicious of any kind of… things like VAR or any kind of standard risk metrics. I just think that risk is out there and your worst drawdowns always in the future.

I find if you look back over time, you find again some pretty standard risk profiles. You look at anybody who has been around for say 20 years or so, or maybe even 10, almost every one of those managers has had a major drawdown, which I would characterize as 50% or more, and usually just one, right?

If you believe that your guy’s going to stay the course, then you should probably stay the course too.

—Bill Dreiss

Download “Trend Following” Ebook
Listen to the Full Podcast Episode
More from Bill Dreiss
View All podcast episodes

The people who’ve survived in this business are people who have been able to weather drawdowns without losing their nerve. That’s exactly it. So that’s where the risk comes in. The risk comes in – is your manager going to lose it? That’s where the ultimate risk is. If you believe that your guy’s going to stay the course, then you should probably stay the course too.

Neils: I’m going to quote someone that you may be familiar with and who wrote a few years ago. It’s David Druz, and he wrote a few years ago that, “here’s an amazing thing about robust systems, the more robust a system, the more volatile it tends to be. This is because robust systems are not optimized to particular markets or market conditions. The converse is also true. You can design systems with excellent returns and low volatility on historical testing but which work only for a given period or given market. These systems tend to be curve fit and market fit and not robust.”

This is completely probably against what most people will feel that robust systems are the ones that are more volatile. I sense from our conversation today that, that is the conclusion that you’ve come to as well.

Bill: Yeah, I know David and he’s right, and I think this is a conclusion that most people have been in the business for a while would come to. It’s what I call… This is in contrast to perhaps the dominant quant paradigm which is picking up nickels in front of a steamroller. In other words, you can design systems like Long Term Capital or whatever that can give you smooth returns for a while but what those systems are doing, in my opinion, are what I call warehousing risk.

So essentially it’s like… And I think it’s the same sort of thing that Dave is talking about, you can design a system that essentially accumulates risk in some sense, and that when it finally breaks out it happens in a big way. This is, of course, what happened in the lead up in the financial crisis. You had all these people developing these securitizations and all these sorts of things and they were just setting the system up for a major break.

In the short term, while that was going on, everybody was happy because everybody was making money and the apparent risk was very low. That’s what I think a trend follower tends to do. I’m essentially taking risk – on an ongoing basis I’m realizing risk. I’m not warehousing it, I’m recognizing it and realizing it and that’s just part of the game. I think any sort of viable, long-term investment has to do that because otherwise it’s a disaster waiting to happen.

Niels: Sure, A lot of people say, you hear entrepreneurs talk about it all the time that they learned a lot from the adverse time in their career and so on and so forth, what about drawdowns? What have you learned from your drawdowns? Have you learned anything from the big drawdown that you had more recently that was different from what you’ve seen before?

Bill: No. There’s two kinds of drawdowns. Certainly there’s the garden-variety drawdown, which for me might be 20% to 30%. That’s just part of doing business. They’re not pleasant, but on the other hand the problem is that even the ordinary drawdowns are likely to cause anguish to your clients. In terms of your system, these are pretty much just part of the game.

When you get what I call catastrophic drawdown, say something 50% or greater, one, I think those are pretty rare. As I say, a trader might experience something like that once in 20 years, or whatever. So once again, it’s just a matter of living through it when it happens. In terms of my experience with that, no I suffered but I didn’t change anything. As I say, the following year I was up 85% or something.

So to me the existence of drawdowns, large or small, in my experience have not had any particular bearing on… In other words, I don’t use those as evidence that my system is not working, or that something needs to be fixed. The reason for that, as I say, I have this underlying faith, and I think faith is the only word for it, that markets will continue to exhibit persistence. That is that trend following is a universal or at least a persistent behavior of the markets and, as I say, that’s supported by the fact that the markets are subject to the fundamental economics that don’t change.

31/10/2018 by Niels

If you want to catch the full conversation then just click here

Turtle Takeaways

Niels: Brian and Jerry, what are the key takeaways from the whole Turtle experience when you think back on those years?

Jerry: Hmmm, the takeaways? I think I learned a lot and it was just a magical time. Those were four great years. I had no expectation of future business. I think we all would have been incredibly happy to continue managing money for Rich for the rest of our lives and that would have been a good idea.

…is proper trading a real business?

—Jerry Parker

I think that some of the other guys that we competed against at other CTAs, in my opinion, probably did not have nearly the training and support that we had and yet they have large businesses. A lot of these guys went out together and worked together so I think that was maybe an opportunity that we missed – most of us going out on our own. So I really wish, in some kind of crazy way, that we had the best training, the best experience, the best four years that anyone could ever hope for, but I don't know that I personally…

I went to a good business school, even though it doesn't really look like it. Maybe part of my problem was that… I suggested a question for Rich because I've asked this question many times to myself (maybe it's a bad question) but, is proper trading a real business? To some degree, I've tried to trade properly and not pay attention to clients as much as maybe I should have and not vol targeted and not taken profits and not made it a better experience. I think to some degree that was one of the Turtle characteristics that clients and others are going to lead you down the bad path so stick with your system. Stick with what you believe to be true, but maybe a little more compromise would have made me have a larger business.

Niels: Hmmm, yeah.

What about you, Brian?

Brian: I would say that the most important concepts and things that always resonated with me were that you just have to have really strict risk management, don't overtrade, take lots of losing trades, don't get out of your winning trades until the trend is confirmed that it's over.

I think risk management was rule/key concept number one. Then, concept number two was: keep looking at new systems, blending different values together, different timeframes, to see if you can come up with something better than you already have. So, we've invested a lot of money in our research infrastructure, and we're always searching for the next best, great systems. I think those were the two concepts that I took away from the Turtle program were system development and risk management.

Niels: Now, just going back to you, Rich. I can imagine you telling Jerry and Brian and all the other Turtles, “Trade small, follow the system, do the hard thing, do the right thing, and innovate and so on.” What if you were to teach a Turtle Program today, would you teach them the same thing or is there anything different to what you would teach them in 2017?

Richard: Well, I wouldn't teach the same thing and, to tell you the truth, I don't know exactly what I would teach because it's just harder to do objectively now. Maybe it will get easy again, but people have asked me that question, and I don't have a good answer for them. I don't even know if it would work again, to tell you the truth. These exogenous variables that have kind of truncated trends – they're a problem. There's no doubt about it.

The Outcome of Developing Counter Trend Trading Models

Niels: As we come to the end of our conversation I wanted to ask all of you if you would like to bring up anything with each other? Perhaps you, Jerry and Brian, would have a question for your mentor and vice versa, is there anything that you, Rich, would like to ask Brian and Jerry (so I'm not the only one asking all the questions today)?

Brian: I would ask Rich: I had heard that you had developed several counter trend trading models, maybe 10, 15 years ago. I just was curious if you found anything in that realm that has worked and that you continue to use?

You probably still believe that the trend is your friend when really the rules are your guardian angel.

—Richard Dennis

Richard: I would say almost nothing, you might find something that works, but it would trade so infrequently that I'm not sure it's worth the effort to track it. I had a thought, and it was a decent system, but it almost never happened. It basically was simple, if the price of anything got down to 10N, then you should buy it because that was so out of wack that you could take a counter trend position, but that might happen two or three times a year – not enough.Page Break

Brian: On the other end, say something is so overvalued that it would be a sell?

Richard: Ah, you're never going to get away with that. (laughter).

I think, was it John Maynard Keynes who said, “The market can stay irrational longer than you can stay solvent.”

Brian: That's right.

Niels: What about you, Rich? What about a question for two of your Turtles?

Richard: Well, let me make a comment. I really don't have a question, and it will be, like most of my thoughts, simplicity itself. You probably still believe that the trend is your friend when really the rules are your guardian angel. I'm not sure which one I would rank as the first, most important principle. I think it's pretty much a dead heat.

View All podcast episodes

Identifying the right Parameter Set for Risk Management

Bill: The big trend followers, and certainly trend following dominates the CTA space, you’d think they’d all just be cannibalizing themselves. I think this gets back to what I was saying, that I’m a technical trader that believes in fundamentals. So there’s a world out there that can’t be controlled or that’s beyond the reach of market psychology or market methodology and I think that world is subject to forces – generally longer term forces – that are pretty much universal and timeless.

Niels: Human behavior, yeah.

Bill: That go back in history as far as you want to go and that will go into the future as far.

Niels: You know, I agree with all that, Bill, and it’s interesting because there’s obviously still so much resistance, it’s fair to say, by a lot of people, certainly on the investor side, to embrace this and you always have to justify why trend following works. Even if it has a year or two of under average performance then it’s their case for why it has stopped working and it’s never going to work again. So certain things don’t change.

I want to go back also to another point which I think differentiates you a lot compared to the managers that we see out there. Maybe you can explain more that where I’m going with this is that I know that you, or at least part of your system is not looking at parameters, meaning you’re not trying to optimize a certain parameter set while, if you use moving averages, or price breakout, whatever it might be, clearly a big part of the research is really identifying the right parameter sets to use. Explain to me about that and why you’ve chosen this way of looking at it.

Bill: Well, of course the idea of data fitting has been the nemesis of anyone who’s tried to design systems. So one of the attractions to the fractal approach was that you’re dealing again with very fundamental patterns, but you’re dealing with patterns as opposed to numbers. You’re dealing with pictures instead of the numerical approach. So in the first place, if you adjust the algorithm the way I’ve described it, is not a matter of optimizing on any kind of numerical parameters. It’s a matter of setting up a certain structure and then, in a sense, graphically utilizing that structure to translate that into patterns.

…went through and tried all the different possibilities and picked out the best one.

—Bill Dreiss

Now there’s certainly data fitting in the sense that you’re fitting what patterns that you think are significant versus those that you don’t. You’ve obviously got to have some choice there. For instance, trading weekly charts versus daily charts, that’s obviously a parameter that you’ve selected. But these might be numerical parameters but they’ve been selected on a qualitative criterion. They haven’t been selected because I went through and tried all the different possibilities and picked out the best one. It was a much broader type of judgement that was made.

So the advantage, again is that, in terms of designing a system, you’re not really focusing. You’re coming into it with an analysis that’s based upon, shall we say, qualitative judgements about how the markets work and so on and so forth. Then that’s implemented more or less directly without having to go through a lot of optimization, testing, and so on and so forth. Obviously you’re going to backtest your methodology and if it doesn’t look like it works you’re not going to use it. So right there you are data fitting.

Nobody uses a system that doesn’t test out. You certainly can’t avoid that kind of parameterization completely, but you can certainly keep it to a minimum and also avoid being deluded by it. I think the worst thing that comes from parameterization is you tend to think you’ve got something that’s a lot more, say magical, than is actually the case.

Niels: When it comes to risk management, though, parameters are difficult not to apply: how much should I risk, etc. etc. Obviously people usually come to these conclusions based on research and so on and so forth. How do you, just sort of broadly, frame that in terms of just the parameter side of things, not exactly the risk management, we’ll talk about that a little bit latter?

Bill: Yeah, well but risk management is pretty much common sense. Everybody’s risk management is the same. Risk management is driven by the realities of the business. To some extent it’s a matter of choice. In other worlds, I’ve decided to operate in a certain, shall we say level of leverage. You find that the level I operate in is probably towards the high end among CTAs. If you get much higher than that then you are out of business.

Most CTAs, as they get big, they tend to cut their leverage. So the typical CTA is probably about half of what I use. But again, that to me is pretty standard. In other words, I can just glance at somebody’s performance record or whatever – just basic stats, and I know what region of risk or of exposure that they’re operating in.

If you look at those people, the people who are established and have been around for a while, they’re all in the same ballpark.

You look at me versus DUNN or various other people with similar shall we say standard deviations – monthly standard deviations or similar drawdowns, or whatever, we’re all doing the same thing. We’ve got about the same margin to equity.

So I think that, to some extent, money management’s another issue. Once your system generates a series of trades, which may or may not be correlated. Typically, you’re going to have about 40% winners and 60% losers and on and on. That’s why I say that different people have different systems and, certainly from a marketing point of view, it’s nice to say, “Talk about how special your system is.” But you’re really pretty much driven by what the markets are doing and what they offer you in terms of possibilities.

Someone who… A well designed system is going to do a reasonably good job of capturing what the market offers, but it’s more of either you’ve got one or you don’t, it’s not a matter that there’s a lot of distinction in terms of what the outcome is. Once again, the outcome is pretty much driven by the level of the amount of leverage that you take. If you normalize to that leverage, then anybody who’s been around for a long time is going to have pretty much the same performance.

We have a risk management concept that overlays the portfolio that's based on marginal utility. So, we're harvesting profits along the way which is very different than what we did learn in the original Turtle trading programs. We're still doing the same things, just a little bit differently than we used to.

Also, if you enjoyed this short insightful takeaway from a past episode of this show then you will love the free book I’m giving away right now….it’s called “The Many Flavors of Trend Following” and it includes some of my best insights on this perhaps the Most Dependable and Consistent, yet Often Overlooked Investment Strategy… Get your free copy here

The problems with Smart Beta

“Nigol: So, the short-term index is flat since inception, or negative or actually down since inception. So yes, “challenging” is, as I said, you’re being extremely gentle.

So, why are short-term managers still getting allocation? This is because of diversification, but also because of convexity. If you’re long a put on the market and it’s going to cost you ten percent a year, fifteen percent a year, twenty percent a year, you can get the same protection for five percent a year negative. Investors who are pricing these things accurately will see a great investment.

…what’s critical is that the short-term space is much more easily crowded

—Nigol Koulajian

So, lucky for us, we’ve been able to provide substantial alpha relative to the CTA indices, whether short-term or long-term. So, in the short-term space what’s critical is that the short-term space is much more easily crowded. When you’re trading short-term, you’re typically trading more stops and trading on stops intraday rather than VWAPing (Volume Weighted Average Price) or trading market on open, market on close.

So, you’re very sensitive to spikes in the market, up or down, and you’re getting whipsawed much more if you have short-term noise. So, what’s critical to do then is to have the right filtering techniques.

Of course, you want to buy cheap convexity, and you want to sell expensive convexity, realized in the markets. One way to do this is, if everybody is trading a ten-day channel breakout, you want to short ten-day channel breakout.

If you look at the short-term CTA index, you can have seventy percent correlation to it by trading ten-day channel breakouts. I know short-term CTAs are much more short-term than that, but ten-day channel breakout I think has seventy or eighty percent correlation.

So, this is kind of like the smart beta version of the short-term CTA index. You want to be shorting that and going long momentum around it. So, if everybody wants to buy the S&P, nobody wants to buy stock number five hundred and one, short the S&P and go long small cap. It’s typical arbitrage of equity long / short.

The same thing applies in the CTA space. You want to short smart beta and go long everything around it. In the short-term space, in particular, where you’re highly affected by the liquidity of the markets this becomes very, very critical. So, there are ways to trade mean reversion where it’s kind of like the “lazy man’s trading,” where you want to find the positive convexity around it, the same as I explained with the S&P.

Niels: Just curious, maybe on a slightly different tack. You bring up the words “smart beta.”Of course in our industry, and in particularly the trend following space, over the years it has been… Certain firms promote trend following as being a very easy risk premia to replicate, so they sell their products very cheaply. Yet, I have not really seen that these products have outperformed the true veterans in that particular strategy.

I’m just curious, but the smart beta products, some of them, have raised billions of dollars because people look at the fees and say, “Oh yeah, it’s easy so we shouldn’t pay so much for it.” So when you say you should short beta and do everything around it, what should you then do? I’m curious whether, in fact, is smart beta (and I’m referring to trend following because it’s easy for me to understand), is smart beta overrated in some ways? That it’s maybe not as effective in capturing that risk premia in some ways?

Nigol: It’s very effective when nobody has money in it.

Niels: Right, OK. (Laughter)

Nigol: The issue with smart beta is that with transparency comes a certain amount of decay. So, although you have multiple firms, each firm is managing three, four, five billion smart beta type of products; or AQR is managing, I don’t know, fifteen or twenty billion, the returns were great just prior to them raising the money.

So, take an AQR replicator: let’s say you use twelve-month momentum, six-month momentum, three-month momentum in combination, you end up with like an eighty-five percent correlation. But then you see that, relative to that basic replicator, once they actually raised the assets then they had substantial decay in performance relative to their own replicator.

…The issue with smart beta is that with transparency comes a certain amount of decay

—Nigol Koulajian

So, what I’m trying to say is that although it seems that individual CTAs are trading one, two, three, four, five percent of volume, or ten percent of volume VWAP, and very gentle, and making markets. Still, markets are heavily influenced by CTA trading. At the points where liquidity is required, there’s not that much liquidity [where they are trading].

CTAs are a very big percentage of volume in the market. It’s easier to see that, trading short-term, saying to yourself, “How are my assets influencing my slippage,” (or that type of thing) and seeing what type of influence you have.

We’re trading one or two percent of volume or three percent of volume, but I think we still have impact. Analyzing the impact on short-term and then going and applying the same techniques of analysis to longer-term, you’re saying that the same exact thing is happening to long-term trading where you’re running hundreds of billions.

So, going back to smart beta, smart beta is great when nobody has money in it. The more assets that come in, the more those specific entry points are going to be affected… those entry points are going to become counterproductive.

So, trend following works. I can give you that as a general statement, as a technique, because it captures something that individuals or investors don’t want to do which is trade…

Niels: Buy the high and sell the low which intuitively feels weird.

Nigol: Correct. The natural thing for human beings, whether in kindergarten, whether in high school, whether you go to business school, everybody teaches you to buy low and sell high. Nobody wants to do the opposite.

So those psychological biases are why trend following (which is a technique which is cyclical in return, it’s definitely not a straight line) sometimes works and sometimes doesn’t. But, when people believe in it, it stops working because assets influence it.

This is a very, very big factor that you cannot look at smart beta outside of this factor. It’s not that investors are coming in because there was a run in return… I don’t believe it’s purely a timing issue, like coming in 2009. I believe that the assets coming into a strategy are having a major impact. That’s my opinion based on what we’ve seen in the short-term space and, again, applying the same things to long-term.

Niels: Just staying with that a little but because now you’re talking about industry, capacity, and so on and so forth, what about the short-term space in itself? Where do you think that some of the techniques that you use… You say one or two percent of volume, but even with that, you can see the footprint. Where is enough, enough, in your space? (Not to end up in the same place as all the ones who have sold smart beta, but are not delivering for their clients.)

Nigol: So the death of channel breakout… Let’s say the Turtle system (buy at fifty day high, sell at fifty day low sort of thing, with a shorter-term stop loss) came when the model went public in the late ‘90s early 2000s. Those models had straight-line equity curves and flat lined in the 2000s.

Then longer-term techniques such as those typically used with moving averages and exponential moving averages, flat-lined later in 2009. So, in the short-term space—let’s say things more similar to vol. breakout—flat-lined, definitely, since 2009 when the vol. has been low. But they were, with filtering… let’s say the smart beta in short-term, that was dead that long, long time ago.

Using certain types of filtering you could still survive and make alpha. Then it became extremely… the more the vol. compressed, the more difficult in became. I would say, short-term smart beta is dead, long dead, and as a matter of fact you can trade against it which is what all those mean reversion models that you see today are…”

Is Filtering the “Holy Grail?”

“Katy: Nigol, you talked a little bit about filtering. Could you maybe clarify more about how you think about that and what you mean by filtering?

Nigol: So, filtering is like the Holy Grail. Basically, the alpha comes from filtering. So, filtering, fifteen years ago trade momentum or trend following occurred only when the vol. was compressed. That's one way, that's the way it was.

Then maybe five or ten years ago it was trade momentum in the direction of the negative skew. The skew in option pricing tells you that, effectively, if a market starts to go down there's going to be a trend. There is going to be a very strong drift.

…The market is telling you that if we go down, we're going to go down a lot

—Nigol Koulajian

If the S&P starts to go down today, the market is expecting that it is going to lose twenty percent a year – that type of thing. So that's another way of filtering. The market is telling you that if we go down, we're going to go down a lot. So you say, OK, we'll go along with you, no questions.

Katy: So do you think that there is more that you need to spend more time thinking about that in today's markets as opposed to a few years ago? Is that the secret sauce, I guess we might say?

Nigol: There is more clarity gained not around the way the markets are but by the way that people trade. I don't believe that the markets are a certain way; that if you're long the S&P you make seven percent a year, and if you're long CTAs you make ten percent a year. It depends on what people are thinking and the way that people are trading.

The way to (on the run) evaluate the level of conviction and risk people are taking means asking, ͞Where is the surprise potential? So vol. compression was fifteen, twenty years ago; convexity was five, ten years ago. Today a lot of what we do is around crowding.

Effectively trade against… Whatever feels good, do the opposite. (Laughter) You can do this looking at the relationship of the between the returns of different sectors within momentum. You can look at certain markets within a sector and assume that things are going to mean revert.

Katy: So if it's uncomfortable that's a good thing.

Nigol: Correct, if it's uncomfortable or… The way that we're programmed to think… How does the way we are comfortable thinking lose money? That is why we bet on that.

Today, you can hire people who are undergrads and who can run optimizations that quadruple PhD's couldn't do twenty years ago. The techniques available for optimization are lacking – they are not free. As a result the ways that people are chasing returns, and therefore reinforcing short-term bubbles and then breaking them down (how the markets de-pattern themselves), are really critical and has changed.

Our job is to find ways to provide this convexity with positive alpha at the same time by relying on what people are comfortable doing and trading against it. In all time-frames, it͛s available. You can hire ten quants, and they all tell you the same thing, and you do exactly the opposite.

Niels: Is there a risk, in some ways, that…

Someone shared this with me, and I thought it was interesting. If you look at the European part of our industry, and you look at the US part of our industry, there is no doubt that in the late 90s (I think) the Europeans, the managers were seen as being more scientific. Therefore, a lot of institutional investors preferred that over the US managers who were seen more as ex-floor traders who systematized their rules – whether it be Turtles or someone like that. But, it also meant that a lot of the European managers were recruiting from universities to emphasize this scientific approach to investing.

Of course, as we know, in the universities they teach from the same books. So, their approach to (for example) risk management is, perhaps, more similar than what we think in that they start looking very similar. I don't know, it's not a study that I've done, and it's not a right or wrong, but you're saying is, do a little bit the opposite, be the rebel. In a sense, if you recruit from the same university and they all end up doing the same thing (maybe that's not down the route yet, obviously with a few exceptions), that's a lot of money still flowing to those types of strategies.

Nigol: Sure, there are two aspects to that trade. First, universities don't teach you what you need to know; they teach you what recruiters want you to know in order to go sell their products. So, it͛s not necessarily the best knowledge available. You know how it works.

The next phase is the scientific phase where are we progressing scientifically in everyday life? Again, investing is not something that can be managed through a scientific process – absolutely not. It's a complex system where high conviction results in the pattern breaking.

So if you hire a scientist who says, ͞People who wear white shirts typically go to the beach, and people who wear black shirts are going out (whatever), you can do that in everyday life. It doesn't affect the process, but in investing, the more advanced… With science, you need to do it just to understand what people are thinking and feeling, not because it gives you an idea of what the markets are going to do. It's very different.

When everybody believes that buying the S&P after two down days is a one hundred percent trade or a ninety percent trade, the market de-patterns. Science is not very good at figuring this out unless you're looking at the science of the mind (we're getting there).

Another way for us to measure this is the level of conviction that people have in certain trading strategies. So, you evaluate the behavior of the returns of different trading strategies, and that tells you how the people trading these are feeling based on the Sharpe Ratios, and you can do certain things with that.

…First, education is not ideal. It's intentional

—Nigol Koulajian

Download “Trend Following” Ebook
Listen to the Full Podcast Episode
More from Nigol Koulajian
View All podcast episodes

First, education is not ideal. It's intentional; it's not pure knowledge. Second, they are cyclical, and science will not understand markets. Depur Neuton, I think, was a good example of that.

So, you cannot have a purely rational mind when you're trading. You need to realize the cycles of emotions that investors are going through. At a certain stage that might change, but the frustration today…

So the people are saying, want to invest with quants, that is because they are frustrated with the macro managers who haven't done anything because central banks have made it impossible for them. That's, again, a cycle where people will run away from quants and go back to the thinkers one day.

I hope you enjoyed these insights about Smart Beta from one of the most successful short-term trading firms in the world. If you want to check out the full conversation… just click here! Also, if you enjoyed this short takeaway from a past episode of this show then you will love the free book I'm giving away right now…. it's called ͞”The Many Flavors of Trend Following” and it includes some of my best insights on this perhaps the Most Dependable and Consistent, yet Often Overlooked Investment Strategy… Get your free copy here

10/10/2018 by Niels

If you would like to hear the full episode then you can just click here

Katy: I have a question too, one of the biggest challenges we’ve had in the industry has been return dispersion for investors too. It’s that you have one fund up, another fund down, it seems that there’s a lot of difference in terms of how performance has not been very consistent across managers in the space and that creates a lot of confusion for the investors.

How has your experience been with this? Have you encountered… Do you have some thoughts about how DUNN sees that and how you have fared in this sort of change in the industry, and what do you think drives it?

There’s always a sweet spot for your period of time you’re looking back over.

—Fname Lname

Marty: Gee whiz, Katy, you’re the expert in this area. You’re the one that should know these answers (laughter). This is my opinion, which you can tell me whether you agree with my opinion or not.

Katy: I want your opinion.

Marty: I think it really has to do with time periods that people are looking at. There’s always a sweet spot for your period of time you’re looking back over. I think one thing that has worked well for DUNN is that we don’t put any restrictions on what’s available to the program from a time constraint. It can go as short as a week and as long as four or five years if it chooses that.

Now, the parameter selection process is automated, it happens weekly, but it doesn’t matter that it happens. It’s not imperative that it happens each week. We can go eighteen months without any real difference in return, so it’s not sensitive to the parameters.

I think it’s important that you allow the system to evolve between shorter time periods and longer time periods given whatever the investment environment is. That’s the only reason I can see why… Well, that’s one of the many reasons why, I think, we’ve been able to manage this environment well. Between the ARP, between the uncorrelated revenue streams that we have, and the adaptive nature of our system, we have an exit strategy that has been implemented over the last five to seven years. All these things have made us better.

Katy: Yeah, I mean it’s been a very… If you think about it, the reason that I’m interested is that you’ve done very well for your investors over the last few years and it has been very challenging. Quantitative easing, very low trends…

Marty: It’s been one of the most difficult environments that I have ever seen in what we do. I thought this was going to be easy, and you look at the post-financial crisis and it has just been a hard road for trend following systematic managers. What’s happened is the central banks have created this environment where everything is correlated across all the markets. So, in reality, I ask myself, “Do we have diversity?”

What is diversity? So, people are investing in their stock portfolios. They go to a mutual fund, “Oh, we’re a diverse mutual fund. We have a hundred and fifty holdings. Nobody has more than two percent of the AUM, all these stock names, pretty diverse, right? I would proposition no. If all those stocks are highly correlated aren’t you truly only invested in one thing?

That’s the same thing that happens in the managed futures space when all the markets become correlated. It’s a lot harder for our markets to become correlated because it’s the softs, it’s the energies, it’s the metals, it’s the currencies, it’s the interest rates, it’s bonds, it’s equities, across all geographical areas.

But, after the financial crisis and the central banks were all working in tandem, yeah, those markets became very highly correlated. That is the most difficult time for us to make money is high correlation, low volatility. There are no trends. Everything is trading sideways. Because of low volatility, people tend to want to put more positions on because they’ve got to get to their risk targets.

This is what started our research on the Adaptive Risk Portfolio because that doesn’t make sense to do that. So, we’ve now incorporated that into our model so that we don’t do things that don’t make sense.

The thing that’s the most encouraging that we’ve seen in the last quarter of the year and is exciting to me is the correlation matrix is broken down. Everything is starting to break apart; you’re starting to see European bonds move different than US bonds, Asia moves differently than the rest of the world. These give us opportunities to make money.

All we need are opportunities, and the winners are always going to outweigh… The winners are big, the losers are small, that’s the whole point of trend following.

All we need are opportunities, and the winners are always going to outweigh…

—Marty Bergin

Niels: I think there are two really interesting observations here. One is, and I don’t know that people really know that, but since 1990 when the BTOP50 Index was started to be computed, there are only seven years where it’s down, but six of those seven years has been since 2009 when quantitative easing started. So, there’s a very clear correlation to exactly your point. Of course, since Katy is here, the queen of divergent and convergent strategy, I’d love to hear your opinion as well to what Marty says about what we, as trend followers, really need.

Katy: Well, I mean, Marty is exactly right is that high correlation low vol is a challenging environment, but we can also add the fact that interest rates have been very low too. So, we haven’t had the carry moving directionally a lot of assets.

So, he’s exactly right in that if you’re focusing on being adaptive to those environments and adjusting your risk and implementing that into your process that is really trying to think proactively in how you can handle those different types of environments without blindly saying, “This is a model that worked,let’s execute it.” So, it sounds like that’s the direction you guys went. In that you were looking at adaptive risk and better thinking about correlation…

Marty: You hit on it. I hate to admit it, and I think anybody in our space hates to admit it, but a lot of the returns, back in the ‘70s and ‘80s were driven off of interest rates. Let’s face it.

Katy: It was up to eighteen percent.

Marty: When sixty percent of your assets are held in cash and are earning interest, that could be quite a nice little hurdle. Now, I will tell you, from DUNN’s perspective; we’ve always excluded that in any of our model development. So, we’re OK working in a zero interest rate environment.

All the backtesting that we do, and the out of sample testing that we do and the development of our systems are assuming, basically, a zero rate environment. We don’t plan for any money to be made on our excess cash.

That doesn’t mean we don’t look at… Clearly, futures are more dynamic in the interest rate environment of the older days. The other adage is that managed futures can’t make money in a rising interest rate environment. That just doesn’t make any sense to me. I’ve heard people that will look at me and say, “Well, you’re crazy. Doesn’t it make sense to you? It’s obvious.”

Well, no, because futures prices build interest rate into it. It’s a very simple concept. It’s saying, “In the future, the bond price is going to go up or down by this much based on the projected interest rates.” Either the interest rates move the way you expect it or they don’t. If they don’t, then the price of the futures moves differently than what was projected.

People try to make it too complicated sometimes, and start doing this evaluation of interest rates and how you should be, “Oh interest rates are going up, you should be short bonds.” Well, no, if interest rates don’t go up as fast as you think they’re going to go up then you should be long bonds because that’s the right trade.

Our system is still holding. All the European bonds are still long. We’ve gotten short the U.S. bonds. The other thing that we’ve seen in looking at our simulations and our out of sample data, in our live real trading is that we never take positions on the short side as strong as the long side positions. It just doesn’t happen. I think it’s because people’s perceptions about interest rate are off kilter. So, I think people always think interest rates are going to go up faster than they do. They never go up as fast as people think.

Katy: They’re afraid they’re going to go up faster.

Niels: I think, actually also, when you do research, and if you look at certainly the simpler trend following systems, what you do realize, and this could be tied into the thirty year interest rate cycle that we’ve had, but when you do that, actually, eighty-five or ninety percent of the returns come from long sided trades, whether that’s because it’s easier to capture because short sided trades can turn very quickly, this and that.

But, I agree with Marty that there shouldn’t be any concern that there’s a particular environment where trend following doesn’t work unless there are no trends.

Marty: Right. So, to me it makes perfect sense why most of the trends are on the long side. It’s inflation. The economy is always growing, prices are always growing, and everything is going up. It just doesn’t go up in a straight line

How The Turtle Got Its Name

Niels: Now there has, over the last three decades, been so much talk about how this trading experiment was named: what the inspiration for the Turtle name really was. Some people say that it was related to you seeing a turtle farm in Singapore I think I heard, and another story I heard was it was related to a rock band called The Turtles that performed back then. Why don't you put us all out of suspense and share with us the true story about how the name came about.

…you're going to get a “Heavens NO!” on that one”

—Richard Dennis

Richard: I'm going to stick with the first story about the turtles in Singapore. That actually, that's how they got the name, it was kind of a misnomer but it sort of stuck. If I had a dollar for every plastic turtle that people have given me, I'd be indeed rich.

Niels: But also, talking about the name itself, and I wonder whether seeing the turtles in Singapore or something that happened years before you did the program, but I also wanted to talk about the inspiration for the idea behind creating the Turtle Program. Again, we hear so many stories relating to, one that seems to be very popular is that you and your partner back then, Bill Eckhardt, having seen the movie Trading Places with Dan Aykroyd and Eddie Murphy, where there was a bet made about how you could train anyone to be successful in trading.

Richard: That you're going to get a “Heavens NO!” on that one.

Niels: Oh well good! Excellent, excellent! Well, can you share with us how the whole idea behind the Turtle Program came about?

Richard: Sure, so one lazy Sunday afternoon I was hanging out with Johnny Walker Black, and I started to think about my own trading and realized that a lot of it was just sort of rules that were informal and that I noticed that other traders operated according to rules. Also, some of those rules were very bad, like a lot of traders at that time, their one rule was always buy soy beans.

Having thought about my trading and the rules, it seemed to me that, at that time, just to put a number on it, I thought that two-thirds of trading was following rules and maybe one-third was intuition – the dreaded flare that we talked about during the course and that. So, as the ice cubes melted, I started to make some notes about what I thought was true; what you could do to prove it. It could turn out to be one of those endless debates that never comes to any conclusion. It seemed to me that we could resolve the question by trying to train people and giving them rules and talking to them about intuition and things like that, and that was the genesis. And nobody told me it was a great idea, but nobody wanted to tell me it was stupid either, so we did it

Niels: Sure, sure.

How long before the actual program started? Was this something that you reacted on very quickly and said, “Yeah, this is a great idea, let me do it.” Or did it have to sink in for a while before you created the program?

Richard: It was only a couple of months before we put the things in motion, like advertisements in newspapers, that lead to starting in January of '84. So, not such a long period of time after I thought of it.

Niels: Fantastic! I think that's a great story, and thanks for finally putting the record straight on this point

Jerry's Training Process

Niels: Now speaking about that, and maybe we come back to you Jerry, you've got (as far as I understand) very little training so to speak. Tell us a little bit about that, but also tell me, when you were taught these rules initially, whether this made sense to you the first time you heard it or did it take awhile before it made perfect sense?

Jerry: Well, we got great training, sufficient, more than enough, so our training could not have been better. And on top of that we got a lot of time with Rich and Bill, and support – crazy support, which you don't get in the real world. So, what was the second part of your question?

In fact, these rules seem a little stifling.

—Jerry Parker

Niels: No, I was just going to say, so you had this training which is it fair to say that it lasted two, three weeks?

Jerry: Right, it lasted two to three weeks. Oh, I know! We did take some time where we could go up and have lunch with Rich and trade with Rich. Some of those opportunities were very beneficial because I could ask certain questions and start understanding more of the philosophy and what was going on with the trades and how to think about the trades and the market's systems.

So, that was our Trader for a Day with Rich, which was very important. I've said on other podcasts that it's so important, this sort of support we had in an environment that maybe we'll never have again, where if you're doing the right thing, that's going to be the bottom line on your performance or your evaluation of your performance versus, “Please make me some money, and I don't really care how you do it.”

In fact, these rules seem a little stifling. You should probably use more flair for instance, which is a typical client. I've said before (it's a bad analogy), but it's kind of like saying that if you print the rules then people can follow them or they'll know some good rules but, “You won't be a marine by reading a manual.” I think that we became much better traders because of time spent with Rich and Bill, and the way they ran the program was simply far beyond just a set of rules.Page Break

Richard: I'd like to say this about the rules, in my old age I've come to believe that mediocre rules are better than good traders judgment. That's because one thing in this business that you need is persistence and it's almost impossible to be as persistent using judgment. Obviously, if you follow a system and do everything, you're being one hundred percent persistent. So, it's a part of rule orientation that's under appreciated.

Are the Original Turtle Rules relevant nowadays?

Niels: Now if we just go back to the subject about what was taught back in 1984, I just want to ask all of you if you feel that now, some thirty years later, that this is grossly out of touch with the reality of how trading should be done because things do change; or at the same time, in life and in markets, a lot of things stay the same. Feel free to go first, any of you on this one.

…what's more of a problem is the lack of volatility

—Richard Dennis

Richard: Well good luck with that one. The markets have changed a lot. What works is changing and is a bit of a problem, but what's more of a problem is the lack of volatility. Volatility seems to me to have trailed off over the years intermittently. You know, I'd rather have the volatility back. I mean that's a variable you can't control, but I think that it's more important than adjusting the system, although adjusting the system is important too.

Niels: I mean maybe I should ask that to Brian and you Jerry, so applying sort of the raw Turtle rules in the year 2017, that would still work?

Brian: Well I would say the basic philosophy hasn't changed. You're continuing to do research, finding robust systems, and that means systems with the least amount of parameters that tell you how to initiate, liquidate, or stay out of a trade. We're always looking to build systems that are based on momentum or based on range dependent discrete time frames where you're confirming that a trend is in place. So, you're always looking to capture directional price movement. Obviously, managing risk is paramount, so you manage risk from the trade size, you limit it in the markets and sectors that you trade.

View All podcast episodes

More From Brian Proctor

26/09/2018 by Niels

If you would like to hear the full episode then you can listen by clicking here.

Rob: In statistics, in machine learning and decision theory, many of these concepts that are talked about actually got their start in the ‘70s, in the ‘60s, even in the ‘50s. Where a lot of the nuance has come today, or a lot of the excitement has come today, is in where and how they are applied, in what data sets they’re applied to, and to what degree you can automate and systematize the application of these tools.

Machine learning is a toolbox. To say that we use machine learning to build our algorithms is kind of like saying, “I use tools to build a house.” It’s not really additive in terms of your understanding. So, let me break that down.

Machine learning is a toolbox.

—Robert Sinnott

Let’s talk about what actually we do with machine learning. Before we go into the tools that we use, which I think are really interesting, let me also break down to the problems that we try to solve because I think there’s a lot of hype about machine learning and I think that there are some kinds of problems where there is a lot of potential for growth.

I think that there are some kinds of problems that, regardless of the amount of machinery that you throw at them, are still going to be a challenge and are still going to be a source of where people who are well versed in machines, and specifically their limitations, will be able to still add value as humans rather than just automatons.

So when we think about machine learning, I would say that there are two kinds of problems. You have what you would think of as classification problems, or (as what I would like to call) stationary problems, meaning that the problem that you’re working on doesn’t change over time.

Here’s a great example: Google has come out with a lot of really interesting results and a lot of really impressive, fast algorithms for identifying various things in videos and images. You go back ten years, and it was a really hard thing to identify a cat in a photo. Now it’s a really trivial thing. In fact, you can do that for arbitrary objects. You can just go online, and there are online classifiers that allow you to make these decisions.

One of the very, very, early uses of these things was (in financial markets) things like counting cars in parking lots, or identifying cars in parking lots, or identifying the levels of oil in silos, or trying to predict crop yields, things like these. Those kinds of questions, where it doesn’t matter how many people are looking at the field to identify if this is going to be a good forecast or a bad forecast, or is it going to be a high yield crop or a low yield crop? That doesn’t change the success of detection of that yield.

It doesn’t matter how many people are looking at whether or not that’s a cat in the video or a cat in the image. That doesn’t change your success rate. That’s a static problem and the more data you can throw at it, the more training samples that you can throw at the problem, the better your algorithm will be up to some asymptotic.

What are the challenges there? Well, the challenges there are an abundance of features. The more things that you potentially know about a data set the harder is to glean what’s true. Also noise, the more noisy a data set, the more pixilated an image, for example, the harder it’s going to be to get your answer. But again, the more data that you have, the more training samples that you have, the better your algorithm will be in the end.

There’s a second kind of problem, and that’s, unfortunately, the problem that we typically have in trying to forecast financial markets. So again, in the first case where you can do machine learning for yield discovery, you can do it for forecasting earnings, you can do it for trying to predict the number of skews that will be sold by a retailer. Those are all great things. Those are all classification problems. Those are all stationary.

But when you start trying to answer the question, “Well, this earnings level and this book value and this momentum indicator and this sentiment out of the CEO on his earnings call, what that means to the return that’s going to happen between today and tomorrow? That’s a much harder question because there’s a feedback loop.

The more people try to answer that question, and the better their answer is, the less relevant it is in figuring out where it’s going to go tomorrow. That is to say, the solution to the problem, the more people looking at the problem, make that problem harder and may make the features that you used to think were useful in making your prediction no longer useful because they’ve been fully priced in.

So, as a result, you have this competition, this fight to be first, this fight to be right among lots of intelligent people in the market is going to make something that means that whatever you design for the second kind of problem, this forecasting problem, it’s going to disappear over time. So, it may have been in the past that simple value signals or value pricing in equities may have been very successful at determining the direction of markets. Or, for a more concrete example, if you were to go back to the ‘80s and ’90, simply knowing which direction the price moved over the last five days or ten days was a really good indication of where prices were going to move over the next five days. It was a sharp two or sharp three strategy depending on how good your transaction costs were. That simply isn’t the case anymore.

So, the balance that we have to fight is because of the second category of problem. It’s whatever we think we know may work for some time, but then that will decay as more people figure out the things that we figured out, even if we did have forecasting power before, we may not have forecasting power tomorrow.

…things that used to not work may also start to work again.

—Robert Sinnott

So, it’s not just model overfit. There are lots of problems with model overfit, especially noisy data like financial data, but it’s also model decay, and it’s this evolution of market participants that will make things that used to work cease to work. At the same time, if people stop paying attention to these particular features, things that used to not work may also start to work again. But that’s the challenge that we see today.

Niels: It’s interesting. It’s very, very interesting as an actual field and not a field that I know a lot about; I have to admit. I kind of posed the same question a few months ago when I was interviewing the founders of AHL and to my surprise both David Harding and Martin Lueck and I think Mike Adam as well, actually, they didn’t sound too overly enthusiastic about artificial intelligence. Actually, to some extent, from memory they were talking about, well, in a sense we already have that because each of our brains, each participant in the market kind of makes up that structure.

To break it down in a very simple way, for me, is of course that I think of it as the more examples the machine sees it learns more and more. But, it learns more and more from what’s happening in the recent history, and we know that things constantly change and might even go back. You know, we don’t have this kind of data from fifty or a hundred years ago, but actually, to some extent, you could argue that markets go back and replicate themselves into how things worked fifty or a hundred years ago. It doesn’t always have to be new change. It could be changed back to the way it was. So how do you overcome all of that? How do you make it work?

Rob: I don’t actually agree with them, to a large extent. I think that, as I gave with that example, I think that you can very easily overdo it with machine learning. I think you can very easily overfit your process and that overfitting can come from just finding noise and thinking it’s signal, or it can come from decaying of these attributes.

The things that we’re detecting, the things that we’re allowing our models to adjust to are generally fairly slow moving features. These are not things that are going to change from one month to the next. These are features that are going to develop and disappear over the course of years. Anything shorter than that and we have no hope of detecting and having any sense of confidence in it.

They can never give too much emphasis to a trend.

—Robert Sinnott

Download “Trend Following” Ebook
Listen to the Full Podcast Episode
View All podcast episodes

The difference in the way we design our models is that we design them so that they can (those algorithms can) shape those particular nuances between markets, but that they have a common structure, and have common constraints, and common features across markets that we know (in some sense) hold economic value or has some economic truth to them. So again, putting my statistician’s hat on, putting my model builder’s hat on, when you have an infinitely wide dimensional feature space, anything can go into forecasting any market, potentially. You had only so much data.

Suppose we have forty years of data, well congratulations, you have ten thousand data points on a daily basis. Yes, you could chop it up second to second, or minute to minute, but for most of the things that we’re talking about (especially for momentum and trend following), it’s not going to be additive. The autocorrelation of the signal is too high.

So, how do you deal with that problem? You deal with that problem by building structure. You deal with that problem by saying, “You know what, my learning models, my mechanisms, they can only learn a certain class of problem. They can only learn (within trend following) how I relatively weight my different horizons?”

They can never go shorter trend. They can never give too much emphasis to a trend. The kinds of lessons that they can extract are in an, actually, very, very narrow region of that unlimited space because, otherwise, I can guarantee you’ll pick up noise.

An epochal story (probably epochal), it very well may be true, told to me by Andy Gelman of Columbia was that when, either he was a grad student or when he was an early professor, he had a student try to find the single best predictor of the S&P 500. What the student came back with, after studying thousands and tens of thousands of databases was the quarterly price of butter in Bangladesh. Indeed, over the in-sample period it would have phenomenally high r², and on an out of sample basis, it was worthless.

The only way you can avoid those kinds of instances, besides model discipline, is by adding structure to your model: by knowing what kinds of features you can input, by knowing how those inputs can be used to develop that function, to develop that model, and what form the output will be. That’s where the art, so of speak, comes from.

The only difference between what we do and what Harding did and AHL did back in the day was that we’ve tried to make it so that the algorithms add much of that asset by asset, or asset class, by asset class nuance. We parameterize how they learn, rather than trying to code and tune every asset on an item by item basis.

Why do we do that? Again, because if you have two or three parameters that allow us to figure out or to constrain how seventy or eighty markets are learning or are being detected or are following trends, we’re going to be much less likely to overfit our process. We’re going to be much less likely to overfit those parameters than if we have to fit eighty different markets or two hundred seventy, or two hundred forty different parameters per eighty markets. So, by narrowing that we’re actually taking advantage of these tools and of the systems and of the computation to make our jobs…

18/09/2018 by Niels

If you would like to listen to the full episode then you can do so by clicking here.

Nigol… “Certain investments that don’t have great risk-adjusted returns can temporarily turn into amazing bubbles because the return stream has certain characteristics which give time for other investors to come in. So it becomes very self-reinforcing like a positive feedback loop. There are ways, in certain industries, where you can reinforce that intentionally. If you’re an activist investor, you put on a position and then you say, “Here, I got my position,” and other people come in, you can do great. Effectively, it’s like a pump and dump of penny stocks in the eighties. Crypto currencies are a little bit similar. Traders receive text messages saying, “Buy this!” and “only this much for each account”.

…it’s like a pump and dump of penny stocks in the eighties

—Nigol Koulajian

So, effectively these things are being a little bit manipulated. In the same way, regulators, if they don’t control these things, it can be very, very harmful for investors in the long-term. In the case of the VIX, starting from the Fed put, starting in 2009, which was when the market was not going to go down fifty percent anymore. It’s only going to go down twenty. Then you had equity hedge funds which said, “Since it’s going to go down only twenty then I’m going to buy at minus fifteen.” Then, because the minus fifteen is there, you have short-term swing traders who bought at minus ten. Then you had short-term CTAs who bought it at minus five, and then you had artificial intelligence or machine learning techniques who said buy it at minus three.

So, it’s become very self-reinforcing where effectively people are now justifying that the economy is doing great and that’s why the vol. is low. There’s an aspect of the reality that is that markets lead the economy and not the other way around.

I would say, today, what’s going on in the equity world, and the VIX, in particular, has become a very self-reinforcing bubble which is suppressing the vol. in the market. So when an implied vol. in the market comes down; when people sell options in the market until those options expire, effectively you have mean reversion trading. If the market goes down, the options sellers are buyers of that market that is going down, and the option sellers are sellers of the market when it goes up. So they suppress the vol. of the market.

Typically, that lasts until the options expire. So, look at the open interest in a market. What drove the implied vol. compression? With options of this maturity, that’s typically when you’re going to see the mean reversion of the market. Then it typically breaks out. So the VIX is quite a serious thing because a lot of, I’m going to say, very sophisticated investors are putting large amounts of money at play in it. Based on volatility and normalized risk techniques they think they’re taking very little risk.

Niels: “ On that note, Nigol, you mentioned the word manipulation, there’s the VIX, you talked about big trades, and so on and so forth. But, I think there was an article in Bloomberg recently that mentions that on December 20th of last year, 2017, the “VIX elephant” put on a very large position – something like two hundred and sixty thousand lots. The article went on to talk about how relatively easy it is to manipulate the settlement, which has a huge impact if you’re trading options and you’re waiting for expiring, etc, etc. Do you have any thoughts about this? One of the things that our industry has always been known for is liquidity. We’re not manipulating markets. We’re not creating trends. We’re not doing anything like this. The VIX seems to be a little out of the ordinary when it comes to this.

Nigol: (Laughter) The VIX, by the way, in 2017 there was a six month period where the Sharpe Ratio was fifteen. On the Euro VIX, it was thirty. So, yes it’s a little bit out of the norm. I would say about one hundred times over.

So, with volatility, the amount of risk or the amount of gamma that you have in the market is highly variable over time as you approach expiration. So, you come in and you say, “My vol. is this, and my delta to the market is that,” and then as you approach expiration, your gamma, or the level of mean reversion that you are willing to do, or the amount of liquidity that you are effectively providing to the market, increases exponentially.

…I would say, in particular, it’s easy to manipulate the VIX.

—Nigol Koulajian

I would say, in particular, it’s easy to manipulate the VIX. Today in an industry that, I would say, there’s potentially some misalignment of interests between the interest of hedge fund managers where they participate in the up side but not the downside towards year end. There could be quite strong incentives to make sure that the market stays where it needs to be for a good payout year-end. So effectively, if consciously we’re talking about risk being volatility based, and I say that my position is this (and I have a ten percent vol. and I’m going to keep a ten percent vol. and short the VIX), you’re going to create this bubble very, very easily in the VIX. You don’t need to do this on the VIX.

If you look at equities, if you look at fixed income over time, or the all-weather portfolio, or risk parity (where positions are inversely sized to, let’s say one year, three year, five year vol.), the fact that people are allocating to equities, and the vol. is coming down, they are constantly increasing their leverage and they are reinforcing that cycle. The leverage today, in a ten vol. risk parity product, is six to one, where, at a peak in 2007 it was four to one. In 2003 it was about four to one as well. Today, just the fact that people are chasing or saying, “We agree I’m going to put on ten percent vol.” “You don’t complain,” it’s stated.

Now I say do this: In the market, if I do it in a way that I’m timing my entries correctly, I can suck in a lot of quantitative traders in the process. I’m saying that, in the VIX, it’s very easy. Even in equities and fixed income you have the same thing going on. If you buy the S&P when the realized vol. is low, you typically lose money. If you buy when the realized vol. is high, you typically make money. Effectively you can use only volatility as a way to time the S&P or equity indices just because of the way that people are trading it and the way that people are sizing their positions.

Katy: Do you trade the VIX?

Nigol: We put on tiny positions, as I said, mainly for entertainment. So, we can extract convexity out of the market by trading delta one futures. We don’t need the VIX to get that. If we want the negative skew, we can generate negative skew by trading delta one as well. So, models that bottom pick the S&P have Sharpe Ratios of four and five in the last three years. Again, we provide those for free. You can get that convexity without trading the VIX. We trade it almost with a smile.

Interesting insights from one of the most successful short-term trading firms which I hope you found useful. If you want to check out the full conversation…just click here!