In the Beginning there was Trend Following – A Primer – Part 1
Trend Following Primer Series – An Introduction – Part 1
Primer Series Contents
- An Introduction- Part 1
- Care Less about Trend Form and More about the Bias within it- Part 2
- Divergence, Convergence and Noise – Part 3
- Revealing Non-Randomness through the Market Distribution of Returns – Part 4
- Characteristics of Complex Adaptive Markets – Part 5
- The Search for Sustainable Trading Models – Part 6
- The Need for an Enduring Edge – Part 7
- Compounding, Path Dependence and Positive Skew – Part 8
- A Risk Adjusted Approach to Maximise Geometric Returns – Part 9
- Diversification is Never Enough…for Trend Followers – Part 10
- Correlation Between Return Streams – Where all the Wiggling Matters – Part 11
- The Pain Arbitrage of Trend Following – Part 12
- Building a Diversified, Systematic, Trend Following Model – Part 13
- A Systematic Workflow Process Applied to Data Mining – Part 14
- Put Your Helmets On, It’s Time to Go Mining – Part 15
- The Robustness Phase – T’is But a Scratch – Part 16
- There is no Permanence, Only Change – Part 17
- Compiling a Sub Portfolio: A First Glimpse of our Creation – Part 18
- The Court Verdict: A Lesson In Hubris – Part 19
- Conclusion: All Things Come to an End, Even Trends – Part 20
Welcome to the shadowy world of Diversified Systematic Trend Following. This primer is an account of a curious special breed of trader who beats to the tune of their own drum and disregards the news, hot tips, and economic commentary in favour of applying a simple trading technique. Namely to simply follow price either long or short with the faith that liquid financial markets can at times exhibit trending behaviour.
For any financial market observer, they will sometimes experience moments in time when markets exhibit directional price moves that extend well beyond what can be produced by a random distribution of price action. There is some causal impetus that drives this trending price behaviour. These are the price moves that trend followers want to capitalise on.
While the focus of economic literature over the decades has been towards the notion that markets are normally distributed (or random) and are efficient in nature (e.g. The Efficient Market Hypothesis), while being applicable to the markets most of the time, such notions assigned to how markets behave are more academic in nature or gross simplified statements as opposed to being true descriptors of how the actual market behaves. We find when plotting the market distribution of daily returns for any liquid market, opportunities do exist outside the Normal distribution curve where arbitrage exists for traders seeking a slight edge. However we need to take care as by far the majority of price action defined by the distribution of market returns are indeed random in nature.
Fortunately, we only need a slight edge in the market to then capitalise on this small bias in market data over a large trade data sample where that small edge compounds in a non-linear manner and can lead to exquisite long term net wealth over time. Granted the market spends most of its time fooling the speculator with ‘false promises’ which makes us particularly cautious in the way we manage risk, however provided we leave ourselves open for the exotic ‘non-random’ market anomaly, we can reach for the stars in our long-term wealth ambitions.
Trend followers are regarded as ‘price followers’ as opposed to ‘price predictors’. We like to think that we react to current price as opposed to attempting to predict its next move. Our spoils are generated when trends extend into significant directional moves. Like a surfer we wait until a swell is observed and then we focus on simply entering an existing trending price series, either long or short, and simply follow the trending behaviour till its conclusion. There is no prediction involved. We simply hitch a ride on any existing directional trending price series to see where it takes us.
Given our preference for only participating in a trade event when price is exhibiting trending behaviour, we thereby avoid everyday market machinations and focus our attention on the more exotic market conditions where markets only exhibit this trending condition. It is in these more unpredictable market regimes where predictive models tuned to exploit repetitive market behaviour can ‘break-down’.
Trend followers unlike other forms of trader, are more like spiders who simply place their trading traps in likely areas where predictive models break down. They capitalise on any stray victim who simply got their prediction wrong and whose failed trades fall into our webs.
As a trend follower, we are less concerned with ‘why’ markets can at times exhibit trending behaviour, and more concerned with the simple fact that markets ‘do’ occasionally exhibit this trending behaviour. The world in which we live constantly changes and that is the source of all trends. That is as specific as we need to be in our conclusion. In a world of change, then trends will be ubiquitous.
So by restricting our trading opportunities to those more exotic times, when markets trend, a trend follower hopes that once on board for the ride, the trending move continues. We simply wait for a directional price move of significance and hitch a ride on board the possible gravy train while always managing risk.
While trending price series can be constructed from random price data, a trending price series can at times possess directional momentum. In quantitative circles, we refer to the presence of a directional bias in a price series as serial correlation. Serial correlation means that price at say time t=1 may influence price at time t=5. Price at one point in time may be correlated with price at another moment in time. Under a serially correlated relationship, one moment in time may be dependent on another.
Now the presence of serial correlation in a price series is what makes that price series non-random and path dependent in nature. We will be regularly visiting this notion of path dependence in this primer series. Path dependence is a very important concept to understand due to its influence over long term wealth building. So, any form of quantitative trading method with a definitive edge is seeking to harvest this principle of serial correlation. In other words, looking for serial correlation is not solely a feature restricted to trend following.
We could imagine this effect as that created by the impact of repeated favorable news events over time. Each news event progressively confirms that conditions are improving leading to a building consensus amongst participants of this improving state. This leads to buying pressure which dominates selling pressure over the news cycle leading to a bullish trend.
The presence of serial correlation (also called auto-correlation) leads to a bias in the price series over time. Much like the physics principle of a moving objective with mass, once price is seen to trend either long or short, then we infer that there is overall buying or selling pressure imparted in the price (referred to as momentum) creating inertia in the time series which hopefully continues for an extended period.
Serial correlation is rarely omnipresent in a price series and typically clusters. It waxes and wanes over time and is a feature of complex adaptive systems that are never stationery. Most of the time price is independent to its time series and its behaviour approaches a random walk. At other times a trending series with serial correlation may be just part of a larger mean reversion cycle as opposed to an enduring directional event under momentum.
As a result, most of the time when we enter a trend thinking that momentum will persist, we are wrong in our assumption, and shortly after entry, price either retraces in the other direction to our trend following direction, or simply stalls in its momentum and stays where it is. So, to manage the downside risk associated with unfavourable price movement, trend followers are known to apply a simple rule of ‘always cutting losses short’. We do this by the rigorous application of stops and trailing stops for each market in our portfolio. This way, we ensure that while we incur many small losses along the way, no loss is material in nature to our overall level of trading capital.
However, for those less frequent times that we may be right in our trade entry, a trend follower simply ‘lets profits run’ until the market finally decides that the trend ends and the serial correlation in the price series, that leads to this trending bias in the price series, fades. Hopefully by the time the momentum dissipates, a trend follower has been able to exploit this directional bias with a significant gain. Trailing stops are a prerequisite for the principle of ‘letting profits run’. Having profit targets may offer psychological benefits, but their application compromises our ambition of possibly riding a trend of ‘infinite potential’. A trend follower places their faith in the notion that the less frequent outsized rewards over time will exceed all the small losses that are incurred with this technique along the way.
To capture the infrequent ‘outsized’ gains, trend followers need to be extensively diversified in geography, timeframe, system, and market type across liquid markets. Such diversification allows trend followers to only participate in the strongest trends where serial correlation is more likely to reside which have the potential to possibly turn into outlier moves. The typical portfolio of a diversified trend follower extends across different asset classes such as foreign exchange, soft commodities, metals, treasury instruments, equity indexes and even cryptocurrencies. The intent of such wide diversification across asset classes, that frequently display different market behaviour, is that somewhere, someplace, a market instrument is displaying trending behaviour.
Under a diversified portfolio where our philosophy is non-predictive in nature we adopt a perspective of making many small equally sized bets for all the composite return streams. Such a philosophy ensures that no single return stream dominates the portfolio and risk dollars are equally spread throughout. Given the need to allocate risk evenly across markets with differing volatility we typically use the Average True Range or other normalization methods to define our stops levels and trailing stop levels. While many believe that the use of recent volatility is a sign that trend followers actually do use predictive measures, we prefer to view this normalization as non-predictive in nature where we give an equal weight to any liquid market to trend within our diversified portfolios.
While most markets at any moment in time in a diversified portfolio offer no trending opportunities, the intent of diversification across markets, timeframes and systems is to improve your chances of participating in the occasional significant trending opportunity that pays for all the small losses incurred in this waiting game.
An ideal trend following portfolio equity curve is one which possesses a continuously rising equity curve across a time series. This continuous ascent means that at all points across the time series, trending markets were present and sufficient to pay for all the small losses along the way.
It is however unlikely for a diversified trend follower to achieve this steadily rising equity curve. The infrequent nature of trending environments makes our equity curves more volatile than many alternative trading techniques. The reality of adaptive markets is that they can enter protracted periods of noisy or mean reverting condition where enduring trends are simply absent. This means that our trend following portfolios enter periods of stagnation of building drawdowns while these conditions persist.
This does not mean that the portfolio systems are broken. This is a very natural symptom of trend following where our success is contingent on trending environments.
Any trends that may arise within these noisy or mean reverting regimes typically have no substantive momentum embedded within them resulting in a trend follower being victim to many ‘false trends’. This can lead to slow building drawdowns that can last for considerable stretches of time.
Such enduring periods of drawdown are a feature of price following techniques that patiently wait for trending conditions. While there is little skill in simply following price, your skill as a trend follower lies in how you mitigate the impact of these drawdowns through risk management approaches.
This skill relates to how a trend follower can survive over extended periods of unfavourable regime and preserve their capital by mitigating the impacts of drawdown so that when trending conditions resume, a trend follower can fully participate in these more favorable regimes.
Frequently we find that following extended periods of noisy or mean reverting market environment, markets can, on the turn of a pin, suddenly start to exhibit trending behaviour. We also find that, due to the interconnected nature of financial markets, a transition in one market can, like a set of dominoes falling, lead to these market transitions spreading across markets and asset classes.
Trend followers with their diversified portfolios can then capitalise on these events and find that in a very short period, previous drawdowns quickly disappear leading to far higher equity levels and new high-water marks being achieved.
Seasoned trend followers who manage risk well typically find that drawdowns take time to build yet can quickly evaporate as soon as market conditions become favorable.
What gives us faith in the enduring nature of trend following as a robust trading method is the decades of academic research that supports our cause combined with the very long-term track record of professional fund managers that apply this technique.
You see, for those that practice this trading method, we acknowledge that for a financial market to act as an efficient means of facilitating trade through perceptions of value, which is the primary role of a financial market, then a market must at times exhibit trending behaviour. If it did not, then there would be no reason for transacting price in the form of investing, hedging or speculating in a financial market.
This transactional trade event which is facilitated by the market leads to an emergent outcome over time where net wealth is transferred, from the many to the few. The winners and the losers change over time and this internal transference leads to trends (or flows) inside and across financial markets.
While speculation in a market is a zero sum game (a negative sum game when including the costs of trading), it is tempting to conclude that the market is random. However the flows that arise inside the market between participants is real and leads to non-random trending behaviour. You won’t necessarily see it on the outside as a gross statistical statement when looking from outside-in, but you will feel it as a participant privy to only a part of this gigantic game of transference from inside-out.
Given that trends are a natural consequence of a changing society and planet, for a liquid market to not display trending behaviour over a long-term data set, then something very strange is going on. It is being deliberately suppressed. Post 2010 we have seen coordinated action taken by the Central banks to suppress the very natural market tendency to want to re-equilibrate and offer trending environments via transition. Like any complex system, the changing nature of participation over time of its constituent parts makes trends a necessity to allow for a complex system to adapt over time and re-equilibrate in accordance with changing participant behaviour.
Our job as trend followers is a deceptively simple one. To simply participate in those trending conditions where and when they occur and at all other times to always manage risk and preserve our finite trading capital.
Now as simple as this statement sounds, to be a participant for trending opportunities requires enduring patience.
We can never be sure of ‘the where’ and ‘the when’ that a financial market will display trending behaviour and sometimes there is a very ‘long wait between drinks’. We are at the mercy of the market in this regard.
So as opposed to other trading techniques that are predictive in nature and exploit an opportunity based on a repeatable pattern of behaviour, trend following is regarded as a non-predictive method that simply jumps on board existing trending price behaviour and hitches a ride while the trending condition lasts.
Over-trading arising from the frustration of patiently waiting for trending conditions is the enemy of a trend follower. We must wait for the market decide to trend and not try to force the situation by attempting to predict the top or bottom of trends or trading any price series that inconsequentially moves. Over-trading inevitably leads to many further small losses that can compromise our ability to ‘pay for them all and then some’ when we enter favorable trending market regimes.
While this technique is deceptively simple, it flies in the face of human intuition as our brains have been sculpted to predict, so any non-predictive method is an exceedingly difficult method to psychologically tolerate. Most retail traders hate the uncertainty with trend following. Unlike predictive methods that attempt to take a controlling stance in dictating our fortunes and pounce on immediate opportunities, we must wait until a trending opportunity arises and then jump into the prevailing trending current and float along with existing price action.
Most of the time we are wrong in our trade entry decision and consequently continuously feel the sting of small losses (referred to as whipsaws) which slowly contribute to building drawdowns, however we participate in this game because during those rare instances we are right, the market can deliver us a windfall that makes the patient wait so worthwhile.
We make our fortunes from the outlier as opposed to the everyday churn of the market and therefore we need to literally invert our thinking to alternative predictive trading techniques that attempt to exploit a repetitive feature of price action.
Now hunting for the outlier sounds like looking for a need in a haystack. To be able to capitalise on these infrequent unpredictable events necessitates that a trend follower needs to be in the right place at the right time. Now while this sounds like a statement of luck, a trend follower adopts a particular technique that turns this apparent statement of luck into a rules-based process that attracts those traders seeking long term sustainable returns.
You see being in the ‘right place’ necessitates that trend followers deploy extensive diversification as a key weapon to participate in trends wherever they may occur…… and being at the ‘right time’ requires that trend followers adopt systematic methods that are available to participate in trending conditions 24/5.
This introductory primer has hopefully opened your eyes to this peculiar world of diversified systematic trend following. As simple as the technique sounds of simply following price, the practical application of a diversified systematic trend following technique is quite an endeavor, but over a series of future primers we are going to give our readers a step-by-step guide to how to adopt this very robust trading technique that is applied by many of the most successful traders in the world.
So, sit back and relax as we take you on this journey of discovery that hopefully will change the way you decide to interact with these markets and set you on the road to long term financial wealth. Stay tuned for our next chapter into this sordid tale of trend following.
Trade well and prosper
The ATS mob