Creating a Powerful Portfolio Blend from Scratch – Part 1
In my previous article which introduced you to the world of practical portfolio management, we summarized the key criteria you need to consider when creating a portfolio from scratch. This article was theoretical in nature…..but now it is time to put the theory into practice and take you on an exciting practical journey where we apply these sound principles to create a robust portfolio performer offering sustainable risk-weighted returns.
Selecting the Primary System
Before we begin the compilation exercise we first need to find the engine of the portfolio (or system) which provides the foundations upon which we then manipulate various asset return streams to create a powerful blend.
For the purposes of this series of articles I will be using a simple momentum trader as our core foundation that regularly harvests short term uni-directional momentum bursts in either the up (long) or down (short) direction.
Now it is important to note that while I will be demonstrating the shady art of portfolio magic using a single system only applied across a number of different instruments, this is for demonstration only. A portfolio manager usually utilises a suite of different systems designed to tackle a variety of different market conditions. However unless you are prepared to sit back and read a Tolstoy epic, I thought it may be simpler to demonstrate the process in an abbreviated way, using a single simple system.
Sample Strategy – Short Term Momentum Trader
In the portfolio management space, the rules are different to that experienced by a retail trader devoted to a single system and/or instrument. Attention is placed on how you blend your return streams and what position sizing to adopt in determining how to address your resulting portfolio. This is unlike the methodologies adopted by the retail trader who demands a desired cashflow and therefore tends to overfit their individual strategy/instrument to current market conditions and hence pay the ultimate price when market conditions change.
Portfolio management is all about survival in the long term. Provided you have survived the long haul then there will be times when you reap the benefits of what the market delivers. It is the market that dictates terms relating to your future success…..and not your system rules.
The core strategy that I will use for this demonstration will capitalise on divergent market conditions as opposed to convergent market conditions. Divergent markets are forward looking where price takes an excursion away from an historic (previous) mean. In other words, where price continues to move in a short or long direction for an extended period of time. Examples of trading strategies that fit into this system suite include trend trading models and also momentum trading models.
On the other hand, convergent trading strategies are backwards looking and assume that current price will naturally over time want to revert back to a historic mean.
Now remember that a portfolio manager worth his salt will combine strategies that address both divergent and convergent market conditions so that when one class of strategy is in drawdown, the other class of strategy is in draw-up. Under this complementary arrangement, the volatility of returns delivered by the composite is significantly reduced….however for demonstration purposes only, we are simply going to create a portfolio that harvests alpha during periods of market divergence. The result will therefore be a portfolio that is naturally more volatile in nature than a more conservative blend of systems.
Given that our primary strategy is a divergent trader, we naturally want to be able to capitalise on those market conditions where momentum bursts of activity are active. We are also deploying a trading strategy that is short term in nature and seek to capitalise on those conditions when momentum bursts are particularly significant. Furthermore we will be deploying an open ended profit condition in our design considerations which therefore ensures that if recent momentum is maintained, then the profit potential of the strategy can be significant.
The system that we will be demonstrating is a very simple breakout trader on the 30 minute timeframe which uses a standard deviation channel (SDC) of 1.0 width to analyse price action while under a short term momentum surge. We deploy an SDC channel over recent price action from the prior major swing high or swing low to define the overall bias in the price action series to its current price.
Having defined where price action is heading using this short term historical look-back technique, we are now in a position to project the standard deviation channel into the future on the assumption that current momentum will continue. This projection provides the rules upon which we trade this momentum surge with systematic application. Let’s look at how we apply these rules with a simple example.
Step by Step Walk-through
Refer to the chart below on the 30 minute timeframe for the Dow Jones Index (DJI30). Price action from the 25 September 2017 to the 2 October 2017 has clearly been on an uptrend.
If we apply a standard deviation channel (SDC) (of 1.0 standard deviations) to the price series from the swing low of 25 September 2017 to the most recent price on 2nd October 2017 we can then define the general attributes of momentum in this journey from its swing low.
We can also define shorter term momentum within this longer term context using a 1.0 x SDC channel commencing from the swing low of 28 September 2017 to 2 October 2017. By drawing this shorter term momentum drive within the longer term overall direction, we can tell whether momentum is on the increase or decrease. If momentum is on the increase then we can project the trajectories of both these SDC’s to be representative of where we think future momentum is heading. From this projection we can define:
- Our pending order entry levels for both momentum projections;
- Our initial stops which are placed at a location where our definition of the momentum projection break down; and
- Our trailing stop which follows the projection of each SDC channel.
Note that for the purposes of this portfolio blending exercise we use a 1% trade risk per trade. On a $200K account we therefore risk $2,000 per trade (being the distance to the initial stops of A with a 0.22 lot size and B with a 0.15 lot size. Now let’s continue our walk-through which describes the systematic manner in which we manage these trades from order activation.
Does the Sample Strategy Meet the Criteria as a Robust Portfolio Solution?
Having defined the primary system that will be the workhorse of the portfolio the next question we have to face in our Portfolio Creation story is whether the logic of the strategy meets our criteria to act as a primary strategy for our Portfolio Blend.
If we refer back to our previous article titled “Smoother than the Sum of the Parts – An Approach to Portfolio Optimisation” we identified the need for the following criteria to be present in any primary system we engage for our portfolio design:
- Only consider systems in your selection that offer Positive Expectancy – Well to save you the long-winded tale of how we arrived at this short term momentum trader as a primary system, it was necessary to conduct an extensive back test across a wide universe of instruments to determine whether this technique had what it takes in offering weak alpha in its own right. In a nutshell this strategy has been tested across 24 liquid instruments between the period 1 January 2000 to current day….and to state it mildly, this strategy clearly does exhibit an edge. It thrives during favorable volatile market regimes where other trend following strategies also thrive, namely periods of crisis alpha….but during unfavourable market regimes, this technique also manages to strictly preserve capital. The edge found in this strategy can be attributed to its trade efficiency. When momentum rears its head in a liquid market, then this strategy is onto it immediately and with its open ended profit nature, allows for unlimited upside provided that momentum persists. If momentum wanes, then the tight trailing stop that is located alongside the momentum projection ensures that we are out of the trade immediately. This technique pays tribute to the trend following mantra of ‘letting your profits run and cut your losses short.
- A small edge is all that is required. We are not after curve fit solutions that cannot survive extended periods of market uncertainty, but rather simple robust solutions that have a history of being able to navigate difficult extended market conditions while protecting your capital and occasionally offering excellent returns when market conditions are favorable. The requirement for a small edge rests upon the assumption that under a portfolio solution, the composite returns of the portfolio can magnify returns without further adding to your risk exposure. In other words, when we find a robust strategy with a small edge, we have a quality ingredient which we can work with under a portfolio solution to go far beyond what a single system can deliver in isolation. Fortunately, the strategy we will be using in our example is a very simple one…..however use this in isolation on a single instrument….and you won’t get far with it.…but use this as a strategy for a portfolio and watch that equity curve reach for the stars.
- Strategies possessing positive skew are viewed favorably as the primary system contributor. This particular momentum strategy is highly positively skewed. What this means is that the majority or the trades undertaken are either small losses, break-even opportunities or small wins. This is sufficient to allow the portfolio to weather the storms of market uncertainty.….however occasionally….when market conditions are favorable, this strategy can deliver knock out profits that make the game worth it.
For a diversified portfolio, we are after systems with positive skew, namely lots of small losses with the occasional big win. For portfolio management, it is essential that your strategy is able to release ‘risk steam’. Yes you will have a more volatile return distribution for each single portfolio addition, but as discussed, you can iron out this volatility through blending methods.
Having a volatile profile of individual system components is a sure sign that you have release valves in place to ‘blow off’ intrinsic risk from your portfolio. A single system with an almost linear equity curve is a massive warning bell for the Portfolio Manager indicating either a system component that is curve-fitted, or a Martingale Variant or simply the result of a particular market condition.
Convergent strategies such as mean reversion tend to be negative skewed and offer great, almost perfect equity curves for a period of time and high win rates attracting traders who do not like admitting they can be wrong…..until the pressure cooker blows and it is too late to defend.
So now that we are comfortable that we have a robust strategy that meets the criterion for portfolio inclusion and offers a small edge over a range of market conditions, we are now ready to enter the next phase of portfolio development where we start building our portfolio by selecting our liquid instruments that span across asset classes to deliver knock out risk-weighted returns.
Stay tuned for Part 2 of this installment article where we start to apply our blending processes to create a powerful perfect recipe.
Trade well and prosper