Trend-Following in a Crisis: A 2022 Review

Episode Transcript

Note: This transcription was generated using a combination of speech recognition software and human transcribers and may contain errors. As a part of this process, this transcript has also been edited for clarity.

Graham Robertson:

So Otto, before we talk about trend following, can you briefly talk us about what we've seen in traditional markets this year?

Otto van Hemert:

Absolutely, and thanks for having me Graham. So if we look at the following slide, we can see that it's been a very tough year for bonds. What this slide shows is the performance of bonds, US bonds in particular, during the first nine months of this year, which is the bar all the way to the right, and the same, the first nine months for each of the preceding 200 years. And the current year is as bad as it's ever been in those two centuries. So it's been truly a very bad environment for bonds. This of course happened against the backdrop of rising inflation that kept surprising up on the upside, inflation getting worse and worse and worse and central banks having to react to it, hiking rates, and that was to the detriment of bond prices. So as the picture illustrates, it's been as bad as it's been over a more than two century history.

Graham Robertson:

And then anything about stocks?

Otto van Hemert:

Stocks have been bad as well. In the following figure, you see essentially the same figure as on the previous slide, but then for stocks. So the right most bar is the first nine month return of US stocks, very bad indeed, but there have been worse years. So equities have been bad but are not at an all time bad situation in contrast to bonds. It should also be noted that oftentimes people think stocks and bonds are somewhat negatively correlated, when equities do bad bonds do better and vice versa, that obviously has not been the case this year, and that too has to do with inflation. Inflation is a phenomenon that negatively impacts both stocks and bonds, it's a common risk factor, and when that's driving the macro environment, both stocks and bonds can perform badly right at the same time.

Graham Robertson:

It sounds like volatile times, any comments on volatility yourself?

Otto van Hemert:

Absolutely, when performance is this bad, you can bat that volatility has been on the rise as well, and it has been. You can see that in the following illustration, you see a blue and a yellow line. The blue line is a measure for volatility of bonds, and that's been going up, and going up to unprecedented levels, again illustrating that for bonds in particular, the current environment is very, very extreme. The yellow line is for stocks, it's the fixed index that's used there, while it was the move index for bonds, but yellow is the VIX, and volatility has been increasing but it has been higher at previous locations, for example, during the global financial crisis. So bad performance for both bonds and stocks, but more extreme situation for bonds, and volatile times for both bonds and stocks, and again, the bigger outlier is for bonds compared to its own history.

Graham Robertson:

Interesting as well that volatility is kind of gradually rising, not that massive spike we saw over the Covid period here.

Otto van Hemert:

That's an important point indeed Graham, and that will have ramification for a topic we'll touch upon, trend following, which tends to do better if things go from bad to worse in a slightly more gradual fashion rather than a very big kind of almost overnight affair.

Graham Robertson:

Thank you. So really quite unusual times for traditional assets when measured on multi-century time scales. How is this compared to alternative investments and trend following strategies in particular?

Otto van Hemert:

That's a key question indeed, people often invest in alternatives for diversification, and preferably diversification when you need it most, when equities, and even worse, also bonds perform badly. If you look at the slide you pull up here, first in blue, we essentially repeat what we just discussed, left most bar is equities. In this case, not the US, but world's equity index down a lot near 20% over the first nine months of the year. Bonds down a lot as well, they're a bit over 10%, looks like down less than equities but bear in mind that bonds tend to have almost a third of the volatility, but the performance has been bad compared to what you would expect to be lower volatility. The yellow bars are then hedge fund indices, and they've not lost as much in case of the HFRI, the left most yellow bar, or the HFRX global hedge fund index, which is an investible version of the HFRI Global Hedge Fund index, have not lost as much but still negative except for a subset of hedge fund strategies, the right most yellow bar, and that is systematic macro and CTA type investments.

If you zoom in a little bit more to funds that are more focused on trend following type strategies, you get the right most bars, and you see the Barclay BTOP 50 and the SocGen Trend Index both mostly trend followers that are being represented there. Those type of strategies have really benefited from the current environment and truly given that diversification when you need it most when equities, and in this case, bonds, both are down a lot.

Graham Robertson:

Interesting, so both indices performing well but actually quite different, so constituents making a difference here as well.

Otto van Hemert:

Absolutely.

Graham Robertson:

So I guess that trend following really stands out, what we seen this year, can you maybe go into a little bit about where that performance has come from, tells a little bit more color on where that performance is?

Otto van Hemert:

So the main macro driver has been the inflation, the sudden burst in inflation as mentioned, and different asset classes respond to it in different ways and at different time skills. Here in blue, we present the third quarter, and in yellow, the year to date through the third quarter performance of a trends program trading about 80 futures at a 50% volatility level. If you look at the performance of the different asset classes, maybe the most interesting one is all the way to the right, and that is equities are selling off, the trend following equities has actually not been the most profitable corner of the trend following space. That's because equities often do see some back and forth movements, but if you look at the other asset classes, rates has been very profitable, as you can imagine, inflation goes up, bonds sell off, and short bonds is a good position to have.

Different central banks react to inflation at different time scale and therefore some currencies start showing bigger moves supported by rate increases than other currencies, so trends pulling FX or currencies has been very profitable, and of course commodities are very intimately related to inflation. Some commodities are almost explicitly represented in that basket of goods that we measure with inflation and trend following, therefore commodity prices has been a good trade this year as well.

Graham Robertson:

So as you know, trend followings often consider a black box strategy and if there's a good way to wind us up, it's calling it a black box strategy. We think it's quite intuitive, can you perhaps take us through how intuition might get you to those asset class attribution that we've seen in this slide here?

Otto van Hemert:

Absolutely, so let's keep in mind those attributions and then indeed you can try to link it to what was the performance of those particular asset classes as you pull up here. What is very good for a trend following strategy is if you have a big move in a more gradual fashion, and that's what happened for example with the bottom line here, which is fixed income. Fixed income has been selling off but in a slightly more gradual fashion, and this links back to what I mentioned, people factoring in inflation but in a gradual way, all the time being surprised that inflation is getting higher and higher and higher. If you look at for example, consensus estimates of inflation, they initially only showed a small increase inflation and then a bit bigger and then a bit bigger and then a bit bigger. So markets kept being surprised in the same direction and therefore a more gradual decline in bonds was the result and that's an ideal situation for a trend follower.

So that's the bottom line in this figure, the yellow line for fixed income. The other line right above, the blue line, is equities, and to the trained eye, you see that equities has behaved a bit different in that it has had more sudden intermediate recoveries even though the overall move has been down just like for bonds. There are those distinct moments where there's a bit of a rally again, and that's tough for trend following strategy, and that's why on the previous slide we had equities not as a very good asset class this year in terms of trend following profits. Two more lines on this figure, the light blue line going up at the start of the year, that's commodities. So commodities were very well supported at the start of this more inflationary time period but peaked more or less around the middle of the year, meaning that trend following strategies did very well in the first half of the year on commodities, but since then, have given up some but not all of their gains, in particular, at the start of that reversal which took place around July of this year.

One more line to go, the green line, that's the US dollar versus other currencies, and US dollar has been very strong. It's not always that the US dollar appreciates a lot when there's an inflationary burst, it depends on whether the central bank, the US Treasury is aggressive or not so aggressive compared to other central banks in rising rates and countering inflation. In this case, the US Federal Reserve was one of the earliest to start aggressively hiking rates, thereby supporting the US dollar and thereby creating a pretty strong and profitable trend in the US dollar as you can see. Mostly starting from the second quarter of this year onwards when a lot of that monetary policy started to become more hawkish.

Graham Robertson:

Yeah, interesting. So it's clearly nice that trend following has done well in an environment where traditional assets have really struggled. Is that a surprise, is it surprising to see this kind of action given the environment we're in?

Otto van Hemert:

So we've done a lot of research on trends, as you may expect from us, but we've tried to go beyond just the last 10 or 20 years with our testing. We've tried to see how trend following strategies perform in different macro environments well beyond the last two decades. We've done, so for example, in two papers you pull up on the screen here, the left hand side paper looks at big equity selloff time periods and establishes how do different active and dynamic strategies perform. In many cases when equities go down, the performance of other investments isn't great either, but there are some exceptions, and trend following strategies they tend to benefit from those equity sell off time periods. Why? Equity selloff time periods often happen around a big macro economic change, this time around its inflation, it could be badly underwritten subprime loans like we had in the global financial crisis, or a tech bubble bursting, another example of a big equity sell off we've had.

Those big macroeconomic shocks lead to all kind of ripple effects in markets, and markets are not immediate in pricing that in but rather need time to digest all the implications and trend following benefits from the more gradual adjustment to bad equity sell off time periods. The right hand side paper is focused more on inflation, the macro shock we're living through at this very moment. In that particular paper, we go back almost a century because we want to have previous inflationary bursts. We obviously had some inflation in the sixties, and the seventies, and early eighties in the US, but also in the forties and some other decades in the previous century, and we've looked at different strategies again, how well do they do when inflation rises, which is that painful time for equity and bond markets. It turns out that commodities tend to do well when inflation rises, which happened again this time around, and indeed rising inflation is also an example of a big macroeconomic shock that tends to lead to price patterns favorable to trend followers. So trend following has traditionally done well in this, as illustrated in this empirical study during previous inflationary bursts over the last century or so.

Graham Robertson:

So really important to make sure that when you're doing any studies on inflation to go back beyond kind of the last 10, 20, 30 years, but that much further when we've actually seen inflation in the past.

Otto van Hemert:

Absolutely, and sometimes people don't go back that far because the data gets a bit more messy and harder to come by, but we firmly believe it's worth the effort, it's worth making some assumptions and proxy returns. It's not a perfect analysis you can do if you go back almost a century, but it's worthwhile because you start seeing the patterns of what works in what type of macro environment, and in particular, in what type of inflationary environment pieces.

Graham Robertson:

So history suggests that trend following should do well in crisis periods, and sometimes these crisis periods result from inflation. I noticed a lot of debate at the moment about whether inflation's peaked, with that kind of inference, I guess people are trying to figure out whether it's time to get back into traditional assets or not, how much of that is true and what does it mean for somebody looking at either traditional assets or trend following investments now?

Otto van Hemert:

So I agree, that is a question on the minds of people, and sometimes with the suggestion that when inflation peaks and starts coming down again, that then previous return patterns would immediately reverse, but the situation is not as simple as that and maybe we can illustrate it based on the slide you've pulled up here. So let's first share what's here. The top left panel is asset, and class, and strategy performance during the full rising inflationary time period, where you see that many asset classes, the ones mentioned on the right of that top left panel, have a tough time. Equities, bonds, the 60/40 portfolio does bad when inflation rises. Going all the way to the left, you see commodities do well when inflation rises, trend following as well. So that's the top left figure, and the top right is the same idea but then the final part of the inflation rising period, the last six months, but broadly speaking the same picture, equities and commodities and trends are good.

Now to your point, what happens when inflation peaks? What happens the months following that peak? And who knows, maybe inflation is about to peak or has peaked in some countries, and therefore It's interesting to analyze what happens. In the bottom left panel, you see what happens when inflation, the 12 months after that, and the bottom right, the six months after that. And you do see that some asset classes indeed show the mirror image, start reversing the moves they've had prior to the peak of inflation. For example, bonds and equities may start to recover and show some positive returns following the peak of inflation, and commodities may reverse some of their good performance pre-peak. Post peak, they may give up some of those earlier gains.

Trend following however is an adaptive strategy, it doesn't tend to stick to its same positions and therefore has a chance to reposition itself, and if inflation peaks and then comes down again, if asset price patterns reverse but not overnight but in a slightly more gradual way, a trend following strategy may have a chance to pick up on that. And I think there are reasons to believe that the patterns are somewhat gradual, that's because inflation picks because central banks have sufficiently raised rates because the economy is on some slower growth, therefore taking off the pressure of inflation, but the real challenge for central banks is that when they take those actions, those actions have very delayed impacts, so they can only know long after they take the action of raising rates how much impact that really has on slowing down the economy.

So they're very unlikely to get it exactly right, they may very well not raise rates fast enough, in which case, inflation may continue to go up or they may overshoot, in which case, inflation may come crashing down and you may even get in a more disinflationary or even deflationary scenario. Because the impact of their actions is so delayed, they're likely to go through a cycle of under and overshooting to their ultimate objective, having a calm inflation of perhaps 2% again, so that under an overshooting naturally may lead to some more trending type situations, again in bonds, in commodities, and equities to the benefit of trend followers. So long story short, trend followers are dynamic and as long as the market environment changes, not overnight, but in slightly more gradual fashions, they may be able to hop on to the new trends, in some cases exactly in the opposite direction, and benefit from that yet again.

Graham Robertson:

So is it maybe fair to say that if you are able to call the top of an inflationary cycle then you can perhaps time your investments into traditional assets like bonds and equities, and maybe out of commodities, but perhaps if you're not quite so sure when inflation's peaking, if you can't time the peak, then maybe trend following is a nice way around that problem?

Otto van Hemert:

Yeah, I think you could say that, it's very hard to call the peak on anything, including inflation, and the market has been consistently wrong in thinking that inflation would be subdued and the rise wouldn't be too bad. The nice thing about trend following strategies is that they have their own built in mechanism in order to reposition themselves and therefore it's kind of programmed in based on, as illustrated, almost a century of evidence, it's programmed in how to respond around that change moment, you don't need to make a separate call on whether you think inflation has peaked or not.

Graham Robertson:

Great. So I guess we've received quite a few questions from the audience over the last 20 minutes or so, perhaps we can pass some of these on, start with the first one. So the first one is actually linked to that point, is it right to consider trend following as a mechanism for detecting regime shifts?

Otto van Hemert:

I wouldn't necessarily put it like that, so regime shifts sounds very binary to me. You're first in this regime and then you're in a different regime, I would say the reality is much more gradual and much more fluid. So what a trans following strategy will do is it may bet on rising inflation by its positions as it did during the earlier part of this year, but then it may transition into a pattern that is unrelated to inflation, if that's what's dominating markets, or very specific, if there are very specific markets where, for example, commodity, where there are supply issues and subsequently the commodity price may go onto a big trend, it may hop on that. So it's really not a binary call that's being made on it's this regime and no it's that regime, it's much more fluid, it's much more gradual, and it's really much more multidimensional in it can go in all kind of different directions depending on which market is showing the better trends.

Graham Robertson:

Great, thank you. Yeah, lots of questions coming in, here's another one. Is a short bonds position tenable in the long term, isn't that betting against carry?

Otto van Hemert:

I've had this question before. There is this notion indeed that short bonds is a tough position to hold because the carry goes against you. The way we as trend followers would think about securities is simply in terms of their total return in case of a bond that would be both the price increase you have and the coupon payments you get on the bond. It's really the total return that matters, and if the total return is trending upwards, then you'd go long, and if the total return is negative, you'd go short. And in some ways you don't explicitly account for the fact whether the return comes from coupon or price appreciation, and therefore the bonds are really quite similar to other assets, just like equity indices, equity indices will go long or short depending on what the total return to equities is, including dividends, and as soon as the total return starts being positive, you'd go long equities, you'd go long bonds.

Specifically on the bond comment, for the longest time we've gotten questions on can trend following even perform well if the falling rates environment that was a bull market for bonds, if that would stop? We don't get that question as much anymore because this year clearly there's an example that being short bonds can be profitable, but also from our historical studies, back in the sixties and seventies, if you look at the data and if you apply trend following type modeling on it, then there too, short bonds was a very profitable strategy, including the negative carry you may have on that bond.

Graham Robertson:

Yeah, it's interesting this year, so we've been profitable in a crisis being short bonds. Lots of questions coming in, perhaps we've got time for one more, especially in regard to the BTOP 50 and the SocGen trend being really quite different, both up, but really quite different this year. The question is are old trend followers the same, do they provide similar crisis alpha?

Otto van Hemert:

Trend followers can be quite different in a couple of ways, they can be different in what exact markets they trade, what markets they trend follow, they can be different in terms of what models are employed. Are they employing more crossing of moving average, gradual models, or more explosive breakout models, as they're called? And they can also be quite different in terms of how fast they reposition themselves when the trend seems to be changing. In regards to the crisis alpha aspect you mentioned, we tested that explicitly in the papers mentioned earlier, as well as other studies, it turns out that being a bit faster in your trend measurement helps for that crisis alpha feature, which perhaps is somewhat intuitive. Sometimes a crisis comes about quite quickly, not really overnight, but still quite quickly like the Covid induced selloff in February and March of 2020, and following trends a bit faster really helps for that defensive characteristic. And that's what we've consistently found not just during this first quarter of 2020, but throughout history, faster trends following formulations have this tendency to be a bit more defensive in the sense that they tend to perform relatively well and particular during those equity selloff time periods.

Graham Robertson:

Great. Time has got the better of us, Otto thank you very much. We've received lots more questions than we can answer obviously here. If you have asked a question, speak to a relationship manager and we'll get back to you soon as we can. And with that, we'll conclude the call and say thank you and we look forward to speaking to you soon, thank you.

Otto van Hemert:

Thanks.