Richard Brennan, Strategy Ambassador for East Coast Capital Management, speaks to Chris Gosselin, the CEO of Australian Fund Monitors, about the company’s diversified, systematic, trend-following approach.
Chris Gosselin: My name is Chris Gosselin from Australian Fund Monitors. Today, we’re talking to Richard Brennan from East Coast Capital Management. East Coast Capital Management run a systematic trend-following strategy across a whole range of futures contracts. Richard, welcome.
Richard Brennan: Hello Chris, how are you?
Chris Gosselin: I’m well. Thank you.
And it’s great being on here. Thanks very much for bringing me on.
Chris Gosselin: That’s our pleasure. Hopefully, you’re going to tell us all about what systematic trend-following futures contracts are all about. First of all, I’ll give a bit of background. You’ve got a track record of 4.5 years, performance since that time of 17 per cent plus annualised. So, that’s a great start to look at a strategy, but tell us more about it, because effectively it’s described as trend-following and systematic, so that tells us a little bit about it, but it leaves an awful lot under the bonnet or under the hood, as they say, that hopefully you’re going to be able to explain.
Richard Brennan: I suppose the key words in what we do is it’s a diversified, systematic trend-following approach. So, looking at diversification, we trade our models over, at the moment, 60-plus futures markets, which are all highly liquid markets. And we are using a trend-following model to basically take our signals in those futures markets.
So trend-following models are what we call rules-based models, which have been developed through extensive backtesting processes and quantitative analysis to basically ensure that, when those particular futures markets are offering trending opportunities, it activates signals which are then automatically generated by a system. And then there’s an automatic rules-based process applied to trading those signals with minimal discretionary override. Well, ideally, with no discretionary override, but everything we do with trading ultimately boils down to discretionary decision-making — for instance, our decision to do trend following.
Chris Gosselin: So let’s have a look at what this means in practice. You’ve got 60 potential markets to trade. They’re all futures. You can go long and short. Is that correct?
Richard Brennan: That’s correct.
Chris Gosselin: So, when you look for a trend, presumably you don’t get in at the start of a trend, you have to wait for the trend to form?
Richard Brennan: That’s correct, Chris. We’re waiting for the signs of a material trend evolving. And once that trend evolves, either long or short, signals are then activated to enter those particular trending price series and hold on until the trend is over. So, we don’t set what we call profit targets. Once we enter a trade, we just let that trading model run with the price while it’s still trending. And, ultimately, there’ll be a trailing stop mechanism which basically gets us out of that trend when it has ended.
Chris Gosselin: So, the old adage of “let the trend be your friend” is really pretty applicable to what you do.
Richard Brennan: Exactly, Chris. There’s another adage you might have heard called “cut losses short and let profits run”. Our rules-based models enforce that principle in all of the models we apply to these futures markets.
Chris Gosselin: And if you’ve got potentially 60 markets which you can trade, we’re talking across a variety of commodities and currencies, bonds, equities, all those futures markets, are you particularly concentrated or is it diversified?
Richard Brennan: So, when we cast our net to identify our signals, we are very diversified. So, we have what we call very small position allocations to each individual market on entry. But, for instance, when one of those markets starts materially trending, we can find that we are riding considerable profit on that particular market. Now, what we do is we let it run. We don’t sort of interfere with the process of our rules-based process, because in our extensive backtesting and research, we identify that that is the most sound way to approach those trending markets. The trends that we capture are often referred to as outliers or anomalies because they’re so significant in extent. And with these outliers, they inherently are very unpredictable in terms of their endurance, how long they’re going to last. So, therefore, we don’t have any predictive process applied to say when to cut off that trend. We let our models run until the conclusion of that price series trending mechanics.
Chris Gosselin: And how large are the positions when you start out? If you’re across, potentially, 60 different markets, they must be quite small.
Richard Brennan: So, in relation to our total overall capital, we are applying about a 1.5 per cent trade risk per trade. In other words, when we enter a material trend, we allocate about 1.5 per cent of our capital towards that trade and then we just let it run. And so, some of those trades develop where you then find that they might represent 10 per cent of your total portfolio, but that’s a significant outlier, which is all profitable to us, so we let that ride.
Chris Gosselin: And then, at what point do you cut the trade? Is there a stop-loss mechanism or is it a rolling stop-loss?
Richard Brennan: So, when we enter a trade, immediately we enter a trade, we always have what we call an initial stop in place, but then as soon as the trade activates and we are riding that trend, we have what we call a trailing stop that trails behind that price movement or lower. So if we’re going long, our trailing stop sits behind going long, that trending price series, and if we’re going short, it’s vice versa. So, there’s always a mechanism to get us out, but it’s what we call a trailing stop that ratchets behind price, giving enough breathing room for that price to be fairly volatile, but still overall trending in its signature.
Chris Gosselin: So, Richard, that’s the system and that’s the process. Let’s have a look at, in practice, what happens. Give us an example. Let’s have a look at two or three examples, maybe a couple that have worked and maybe one, if there has been, that hasn’t worked, and in what markets doesn’t it work? Let’s start with the winning trades first. Give us an example.
Richard Brennan: So, a very good example is a recent cocoa trade, which we’ve been significant beneficiaries of. So, just to put that into context, about 1.5 years ago, 2 years ago, cocoa was trading at around about US$1,500 per metric tonne. Subsequent to that, we started getting a bullish trend emerging in cocoa to a point this year where it’s now up to around about $10,000 or $11,000 per metric tonne. We’ve been riding that price as it’s been going up with our trend-following model. And it’s basically been operating. We’ve had a holding period for that of around about a year, a year and a half, just sitting on that trend and taking the opportunity as it unfolds. And that’s a significant profitable outlier for us and it’s significantly improved our portfolio equity.
That’s an example of a successful trade. And these outliers that we are catching, they are few and far between, so the stories of our unsuccessful trades, it’s a lot of them, but very small. So, if you could imagine, with our models that cut losses short and let profits run, we do get numerous losses with our trades. So, if we are looking at the percentage of winning trades and the percentage of losing trades over a large data sample, you’ll see that our winning trades are about 40 per cent and our losing trades are about 60 per cent. But where the opportunity arises or the edge that we are getting from our models is on the extent of that trend, the magnitude of those wins. So, all of our losses, we cut them short so that our initial stop and that trailing stop will always get us out after we get an adverse price movement in relation to that price movement, but it’s always a small loss. We get lots of small losses. We never let a small loss turn into a large loss with our rules-based process.
But our wins can be significant. And the example of cocoa might be it’s 15 times the average loss. And so, those very large winners pay for all of those small losses we get with the 60 per cent losses. But 40 per cent of our overall trades are winners. And some of them, 5 to 10 per cent of them, are significant winners, such as cocoa, such as orange juice last year, such as soybeans. All of the liquid markets we trade have the propensity to offer these outliers.
And when we look at our quantitative models, because we undertake quantitative analysis on price series, so our technique is a quantitative technique, we find that all of the liquid markets we trade have displayed, over the course of 30 years or so, what we call fat-tail properties. And it’s these fat tails that are giving us these extensions in our trends. So, we trade both long and short, so when we are looking at the trade distribution of returns, we could be either capturing the left tail or the right tail with our models, but we stunt our losses with short losses so we never have adverse losses as tail events, if that makes sense. So we’re capitalising on positive tail properties of the distribution of returns of those markets.
Chris Gosselin: So, it’s dependent on having some large trends, even if you have some short-term trends and you get stopped out of those, you need to have a large trending market, which hopefully, across 60 different markets, you can achieve, you can find those large trends.
Richard Brennan: That’s right, yep. That’s why we need to diversify so extensively, to cast this net wide because most of the time we’re inactive in the market, we’re not participating. So, there might only be one or two of those 60 markets which are displaying these properties. We’ll be jumping on board then with our models, but we won’t necessarily be active in all of the other markets that are in our watch list.
Chris Gosselin: You’ve mentioned quite a few agricultural markets, so soybeans or orange juice or cocoa. Let’s have a look at some of the other markets you trade, say currencies or equities. Equities are, by their very nature, pretty volatile. Can you find those long trends without being stopped out by your stop losses?
Richard Brennan: Yes, so when we are referring to these outliers, Chris, we might, for instance, examine the performance of the S&P 500, which is one of the indexes we trade. And we’ll notice that post-GFC we’ve had this bullish trend for about 10 years or so. That is an outlier to us. So, if we have a model that is riding that overall bullish trend, which lets profits run and is quite loose as far as its model parameters, we can ride that for 10 years or so and have a profitable opportunity over the course of time for the index, the S&P 500. We trade numerous indexes, Hong Kong, Nikkei, etc. And that gives us geographical diversification across our equities.
But also, in the currencies, for instance, we have very significant outliers, for instance in the yen pairs. The Japanese yen has exhibited these significant trending opportunities as it’s progressively devalued over the last decade or so. There’s been significant opportunities there in our particular approach, applying trend models to that currency. And there are other currencies as well. For instance, with the Brexit decision, we might get a significant trend emerging over a very short period of time, which we can capitalise on with our models.
Chris Gosselin: Do the models differ depending on the commodity or the market that you’re trading? So, we’ve talked about equity indices, we talked about currencies. I presume there are metals as well, soft commodities, agricultural commodities. Is the underlying model different? Are your algorithms different for different models?
Richard Brennan: The way we deal with that is a normalisation process. So, with our models, when we are trading different asset classes– like, for instance, we do trade Bitcoin as an asset class. It’s just one single asset class. It’s an opportunity for some trending opportunities there. That’s a very volatile currency. But we also trade bonds, which is a much less volatile currency. But through our normalisation process, we’re applying the same models to those different asset classes. But our normalisation process means that we’re only ever risking the same equalised risk allocation for entry for every one of the markets that we trade. So, we are applying an equal-weighted risk allocation to every one of those 60 markets. So, for instance, cocoa has as much gravitas as the S&P 500, even though one is an entire index and the other is a single commodity. We are applying this equal-weighted positioning to all of these different assets.
Chris Gosselin: Richard, really great to speak to you this afternoon. You can get more information on the Fund Monitors website. And the fund is available for applications via Olivia123.
Richard Brennan: Thanks very much, Chris. Lovely to be here.
Chris Gosselin: My pleasure.
Ends