Has trend following had its day as a trading strategy? The Profit & Loss editors go head-to-head on this debate, with managing editor, Colin Lambert arguing that trend following is dead and editor, Galen Stops, arguing the opposite. Which side do you find more persuasive?
Why Trend Following is Dead
The changing nature of markets tells Colin Lambert that trend following as a strategy has had its day…and then there’s the data.
When Galen and I decided to argue our cases over the relative merits of trend following, I immediately thought of amassing mountains of data around moving averages and breakout points. I then reminded myself this is not the way I do things and as such, decided to go the bluster route.
It is also, on a more serious note, important to remember that back-testing strategies works after a fashion, but to really understand strategy performance you need to know how many participants actually use it. In other words, the more people that follow a strategy, the more impactful it becomes on markets – at some stage it becomes a self-fulfilling prophecy.
At that stage, it also becomes vulnerable as a strategy, for as a trader, if you know and understand other people’s strategies, it represents an opportunity – if enough people simply wait for the trade to become crowded and then force the issue by going the other way, natural equilibrium is restored and the strategy is less valuable than it once was.
That is not – paradoxically – what I think prompted the downfall of trend following, however, technology and human nature did.
Trend following requires patience and a willingness to ride out some rough periods during which the P&L does not look that great. Back in the day (yes, my day!), data was not as readily available, investors did not look at the liquidity of their funds as much as they do now, they just stuck with the manager (within reason). They checked their portfolio once a month, quarter or on occasions, at the end of the year. Partly this is because they were not as “activist” as investors today, and partly because the data simply wasn’t available to compare performance.
The thing is though, humans like a steady stream of income – they are attracted to nailed on, guaranteed profits, and trend following doesn’t always deliver that. Shorter term trading models, especially those that arbitrage using superior technology, do deliver the type of income stream that people like. Thus, in the middle of the first decade of this century, investors started looking at the quant funds, especially those with shortterm horizons.
Of course, we now know that the latency arb model didn’t work for too long either, but in the brief period in which it thrived, it changed the nature of markets, possibly forever. Gone was the trade and hold strategy of the trend follower to be replaced by models predicated upon mean reversion and short-term gains.
The shorter-term models then became self-fulfilling at the expense of trend followers and money managers had to follow suit – if they saw a decent return they rarely resisted the temptation to take it and into the bargain they also traded counter-trend earlier than they thought, which in turn reinforced the market turn that ends the trend.
There is another societal change that I think has contributed to the demise of trend following – the much faster news cycle. Thanks to social media and 24-hour news channels desperate to fill airtime, news cycles through the system much quicker than it used to.
There used to be an information advantage in markets that the trend follower could exploit; that no longer exists. Previously, as information spread, more people got on the trade and the move was elongated and the fat tails that trend followers rely on so much were created. In today’s connected world, the news is with 99% of serious traders in seconds and while the move may be just as big as it used to be – say 3% for example – it now happens in 10 seconds rather than 10 minutes, or even hours.
That is not exactly a trend one can jump on, especially when one factors in how the news cycle will probably deliver an opposite influence at some stage in the coming hours. The only way to make money in these scenarios is to be one of the first into the trade, and one of the first out – and that means speed and shorterterm time horizons.
If you really want an indicator of how random the modern world is when it comes to news that influences markets, there is the current resident of 1600 Pennsylvania Avenue, who seems intent on confusing as many people as he can. There may actually be a trend trade there, but I am not sure the road is smooth enough for anyone but the very brave – and investors and money managers these days prefer the comfort of passive investment to the potential stress of brave calls.
So trend following is a strategy from a bygone era – one that struggles to meet the criteria of the modern investor and fails to handle the modern market structure. But if you are still drawn to my colleague’s case, let me leave you with some numbers.
The SG Trend Index tracks trend following strategies and so far this decade it has outperformed a simple, passive, buy the S&P500 Index on the first day of the year and sell it on the last, just twice. More pertinently for investors, the annualised return from 2010-2017 was 11.38% for the S&P and just 0.67% for the SG Trend Index – and if you don’t like car crashes look away now, because year-to-date (at time of writing) the S&P was up 6.6% and the SG Trend Index down 5.45%.
Investing is all about performance and trend following isn’t cutting it and thus is losing friends quickly. It used to be said the trend is your friend – much like in the social media world that has helped to destroy it, friendship can be fleeting.
Why Trend Following Isn’t Dead
Trend followers must have more lives than cats, given the frequency with which they’re pronounced dead. Galen Stops argues that the critics were wrong before and that they’re wrong this time around too.
So, the first thing to say is that in arguing that trend following is going the way of the dodo, Colin completely misunderstands the role of these strategies within an investment portfolio. Fortunately for him, this is an issue that we address elsewhere in this very edition of the magazine.
Yes, investing is all about performance, but what Colin is either ignoring or ignorant of is that “performance” can mean different things to different investors, and for many of them, avoiding massive drawdowns is as important as generating returns. The investment universe is not made up solely of high net-worth individuals chasing outsized returns that need to be delivered tomorrow, just think of all the massive pension or endowment funds for whom consistency of returns is vital.
This is why trend following strategies still have value within the portfolio: they’re there to diversify the investor’s exposures.
The obvious example here is 2008. While the S&P500 went into freefall (don’t even get me started on the fact that Colin is using this as a benchmark, another common mistake that is discussed elsewhere in this issue), following the financial crisis and ended the year 38.5% in the red, the SG Trend Index finished the year 20.88% up. I think it’s safe to say that there’s a vast swathe of investors out there who would take a few years of mediocre returns, or even losses, to avoid the kind of drawdown that occurred in equities following the crisis. To say that “Trend Following is Dead” one has to posit that it has no use in an investment portfolio, when clearly these strategies do, even if that purpose isn’t massive returns every year.
But okay, 2008 feels like a long time ago now, so let’s look at some more recent data. In 2014, the SG Trend Index finished the year up a massive 19.7% – I repeat in 2014, when Twitter, 24-hour news, electronic trading, HFTs, newsreading algorithms and whatever else Colin wants to blame the so-called demise of trend following on, all existed. With the possible exceptions of Donald Trump’s tweeting habits and the current hysteria around “Fake News”, does the way that news is generated and disseminated in 2018 fundamentally look different to 2014?
I would argue not, and as a case in point, I would ask readers to cast their minds back to 2013 when the Associated Press’s Twitter account was hacked and a tweet was put out stating that the White House had been hit by two explosions and that Barack Obama was injured. As soon as the tweet went live, the algos did their job and promptly caused a 143- point fall in the Dow Jones Industrial Average, and then equally did their job and caused the market to recover minutes later when it was revealed that the tweet was inaccurate. So, the argument that we’ve entered some strange new world where social media and a faster news cycle has rendered trend following strategies as incapable of delivering returns and therefore obsolete is clearly wide of the mark.
And so is the argument that is sometimes put forward that financial markets have generally become so efficient that trend following doesn’t work anymore. Increased automation does not fundamentally change how markets operate, they go through the same stages as always: accumulation, run up, distribution and run down. While these stages exist, trend followers can exploit them.
Talking about inefficiencies, let me tell you one area of the market where trend following is only just getting started: crypto trading. Since the launch of bitcoin futures, which enable firms to more effectively short crypto exposure, some CTAs have been deploying trend following strategies in the crypto space to great effect. The price is still largely driven in the OTC market by retail punters and, let’s be honest, is the herd mentality more prevalent anywhere else than amongst retail crypto traders? It’s a trend follower’s dream, and the CTAs that I’ve spoken to are having great success in this market as a result.
Now, while crypto trading is still a pretty niche segment for trend followers and might be an edge case example, the broader point here is that most trend followers are not just sitting on their hands waiting for the next financial crisis to come around so that they can shout “I told you so” at the rest of the market. They’re launching new funds that utilise more sophisticated trend following algos that have different signals for when to enter and exit a market, not just CTAs but hedge funds in general are looking at how AI tools can be used to analyse data and identify more subtle patterns in the market, and then trend followers are always looking for new products and assets that can be incorporated into their trading strategies.
And one final point. Type in “Why Trend Following is Dead” on a Google search and have a look at what comes up. You’ll find that, following every period in which trend following has produced sub-par returns for a year or two, there is a plethora of articles declaring this strategy to be dead. And yet, somehow, it never is.
Forget the dramatic eulogies, this is just part of a cyclical pattern. Or, dare I say it, a trend.