CTA Fund Performance Report – 31 August 2020 – “Projections Versus Predictions”
Why is prediction such a dirty word for the trend follower? Well it really depends on how you view ‘complexity’ and in particular how you understand complex systems to behave.
For those of us that treat the financial markets as ‘complex systems’….we tend to describe the ‘now’ of a complex system as a particular ‘form or state’ of the market…..and a future market as simply being an unrealised different ‘market state’ that is somehow connected to this now through correlated relationships. We tend to view the outcomes of a market state being dependent on how the ’cause and effect’ architecture of relationships between participatory agents unfold over time. In other words the ‘form’ that a complex system takes is based on principles of emergence where a preceding ‘state’ directly effects the possible future outcomes of ‘form’ that a complex system can take.
We therefore see a story of emergence unfolding in a complex system where in hind-site…the narrative seems very explainable as a single historic path where State 1 affects State 2…then State 3 etc etc etc…..but when sitting on the right edge of the chart looking into the void…we have no understanding of how the ‘now’ that we currently sit in will unfold into the state that the market takes tomorrow. You see that future state is dependent (contingent) on all the entry and exit decisions made in the intervening interval which is yet to occur. We can always however project a future state based on an assumption of ‘continuance of a past state’ with no intervening events that alter the course of that state….but we never can predict a future state that is contingent on the outcome of future events that are yet to pass.
Fortunately our history provides us with a single path or single narrative of cause and effect which appears to be a very understandable linear progression of events in time….but as soon as we stand on that right hand edge of the now…we recognise that the potential variables that influence the very next moment are almost infinite in extent. That very next moment is contingent on those influential variables that ‘may be realised’ and the extent to which their confluence dictates that next ‘state’. Those variables are in ‘reality’ any variable that exists within the causal bubble surrounding that future event. That causal bubble extends at least 13.8 B light years around that instant in spacetime.
So when we ‘say we are predicting, we are actually not predicting a future event at all. What we are actually doing is projecting a past state into the future with the assumption that future conditions will not materially affect that projections trajectory. We are assuming a ‘stationery future’ or have an implicit assumption of equilibrium embedded in this assumption. For example we rely on the fact that there will be a tomorrow simply because the past single narrative has led us to that conclusion time and time again. It is not our ability to predict that tells us about whether there will be a tomorrow or not…but rather it is our method of projection based on past history that makes our prediction.
We ‘trend followers’ are simply dumbstruck by the intractable complexity of it all……so we avoid making any assumptions about the nature of a possible future state and we simply project a realised and known past state into the future. We opt to leave the future state 0f the market to the astrologers out there with the crystal balls ….and revert back to the safer turf of what ‘has happened’ and simply project a trajectory from this single past history into the future and trade that. In other words we simply assume that a past historic trending state will transmit it’s (cause and effect) momentum into the vast array of many possible future emergent states that arise from the moment of the ‘now’. Most of the time our projections are wrong and do not overtly influence a future reality……but occasionally our projections carry with them sufficient momentum of ’cause and effect’ to shape (or bias) the unfolding future possible states.
It is not saying that the markets are random…but rather that markets are actually very time-ordered but almost infinitely complex by virtue of the relational dependencies that exist between all agents that occupy that system….and these relationships are not one to one linear relationships….but rather nested non-linear relationships. To predict an unrealised future path is a hopeless endeavour. I actually think deep down we all understand this…but when we say we are ‘predicting’ what we are actually saying is that our prediction is simply an extension of an historic state. Our assumption is therefore that the past state will continue to correlate for a time with the future unfolding states. In other words we are hoping for the future state to remain correlated to where we are now….but there are never any guarantees as to the speed in which a system state changes….and that is where we find who has been swimming naked when the tide goes out.
Now it is not just us trend followers that have a supreme respect for the concept of ‘path dependence’ which arises from ’emergent form’ and appears to be a common fundamental feature of complex systems….but you will find that physicists have long recognised the importance of these principles. You see when you understand path dependence….you also develop a healthy respect for ‘tipping points’….and this is the gravy that us trend followers are really after. I will let ‘Sabine’ take it from here. You might learn a thing or two about chocolate. 🙂
…..anyway where were we….Oh yes…to the monthly CTA Fund Performance Report for August 2020 .
We use NilssonHedge for reporting purposes which allows us to expand our performance coverage to include a broader array of long term established FM’s who occupy the CTA space and have been in operation since 1 January 2000 to the current day. This performance report focuses only on those funds with a long term track record (approx 20 years). The reason we adopt this long term horizon for reporting purposes is that to survive in these financial markets over such a long timeframe and still be alive today offering absolute returns to the client takes a special breed of Fund Manager who has expertise in surviving the turmoil of a variety of different market regimes. We like these guys and that is why we focus on them. As the years roll on we will progressively expand our coverage to include those FM’s who narrowly miss out in their inclusion when they reach the 20 year performance track record horizon.
So far for the month of August 2020 we have 48 CTA’s reporting and within that grand total we have 33 Systematic Global Trend Following funds. We have to draw the line somewhere and the slow coaches unfortunately miss out.
For those that like the detail, below are the index constituent performance results for the CTA Composite Index (48) and the TF Global Index (33).
- CTA Composite Index (Program Composition)
- CTA Systematic Trend Following Global Index (Program Composition)
The CTA Composite Index 48 was up 0.31% for the month with the calendar year offering modest growth in a volatile year of 5.47%….and the TF Global Index 33 was up 0.14% with a YTD contribution of 4.84%.
Now as ardent trend followers ourselves, we like to narrow our focus to the Systematic Diversified Global Trend Following community of CTA’s.
Top 10 by CAGR since 1 January 2000
Below is a performance table and an equal weighted performance chart of the top 10 performers of the Long Term Trend Following Index Composite in terms of annualised returns to investors (net of all fees and expenses) since 1st January 2000.
Here is a scatter plot that highlights where the top 10 sit in terms of their Compound Annual Growth rate (CAGR) and Maximum Drawdown over the performance monitoring period.
Below are the performance metrics of the Top 3 from this Top 10 list by CAGR. Just look at those returns. It might be a bumpy journey along the way….but when these guys nail it…they hit it out of the ball-park.
Top 10 by Risk Adjusted Return (using the MAR ratio) since 1 January 2000
Now onto the risk adjusted return category. This category is for those that get ulcers when riding the drawdowns of leveraged volatile equity curves. Here are the results of the Top 10 in this category.
….and the top 3 from this Top 10 category.
Top 3 Equal Weighted Combination Portfolio – The Blend of Blends
Now as great as the individual risk adjusted returns of the Top 10 in the prior category are….we can do better when we look at the performance of possible combinations from the TF list. We iterate across the TF Index to find the Top 3 in terms of Risk Adjusted returns as a combination portfolio each month. So for those avid readers looking for the best risk-adjusted result of a possible equal weighted threesome….here is the result for the month.
Just look at those risk adjusted metrics. With a bit of leverage we can reach for the stars with this combination.
The 3 contributing funds for this month based on an equal weighted blend are as follows:
Top 10 for the last 12 months
So how are the guys going in the short term? There is enough style drift in this camp to observe significant variation in performance returns over the short term. Some of the mob have performed strongly over the last 12 months.
….and the top 3 from this Top 10 category.
Well that’s a wrap for the month…
Remember that in complex systems built on principles of ’cause and effect’ the current ‘market state’ is path dependent. This state can only be understood from a knowledge of the path of preceding states and cannot be determined through simple ensemble averaging. The relationships that exist within complex systems make this system unfoldment intractable in nature. Path dependence really affects decision making as path dependence means that the choices we make today are actually limited by the choices we have made in the past.
So what do we mean when we say that we are trend followers. What we are really saying is that momentum is not a simple price pattern but rather a biasing ’cause and effect’ influence that shapes a future path based on the inertia carried forward from preceding states.
As Mark Twain says so elegantly….. “History Does Not Repeat Itself, But it Rhymes”.
Trade well and prosper
The ATS mob