I sometimes wonder how much thought advisers give to the research and expertise that sits behind the asset allocation decisions that they and their clients rely upon. Without asset allocation, any consideration of risk would fall down, because there would be no basis of assumption for the likely growth of the asset classes that make up a portfolio, or their likely interaction in the various possible economic scenarios that we are able to foresee.
No expert judgement around suitability would be possible, nor would any sensible predictions of investment outcomes. All the major trends that drive the returns of our retail customers are rooted in the asset allocation models that reside at the heart of any firm's investment strategy.
One beauty of asset allocation is that the framework can accommodate any style of investment. Synaptic has had a long collaboration with RSMR, whose qualitative commentary provides insights into funds and managers that have become hugely influential. All our major adviser groups rely on this commentary to help direct their fund choices to a greater or lesser extent. Large cap/small cap, global/local, Emerging/Frontier, high yield, long/short duration, sovereign/corporate, convertible, hedged, smart beta, ETF/OEIC, multi asset or DFM… these are all the everyday concerns of those seeking investment returns, but all fit into the framework provided by asset allocation, which is used to map an efficient frontier balancing investment return forecast against risk.
The efficient frontier was a concept that helped revolutionise investment and we still adhere to the principles today. In his Modern Portfolio Theory, economist Harry Markowitz demonstrated how the correlation co-efficients between various asset classes could be used to optimise an asset class blend to maximise portfolio returns.
Markovitz' research has been used as a starting point by research specialists such as Moody's Analytics (formerly Barrie + Hibbert), who can now produce probability-based projections of all viable economic scenarios and asset class growth trajectories. The engine that produces these models is integrated into the Synaptic Modeller tool. By accessing these projections, advisers have access to all the metrics they need to prove suitability and hopefully delight their clients. This investment risk management approach offers:
- A consistent and proven approach to balancing risk and return;
- The extent to which losses in a bad year (1 in 20) will fall, allowing the adviser to accommodate the client's Capacity for Loss with the chosen investment;
- Reliable, research-based portfolio growth rate assumptions, that will allow a realistic plan to be formulated in line with objectives.
This mathematical modelling is done using simulation (Monte Carlo) where thousands of outcomes are generated and analysed, a technique also referred to as stochastic modelling. The accuracy of the model comes from the success that Moody's have had in building the rules that underpin the simulation. This implies of course the truth that not every model is a good one, so be careful which one you adopt. The proof is in the pudding, and Moody's have been proven for accuracy over many years including the 2008 downturn. Moody's (uniquely) attributed an accurate probability and loss to the event, meaning that clients of Synaptic and Moody's should not have been unduly concerned with losses as they occurred at that time and remain invested, in line with the FCA's guidance on risk and Capacity for Loss. The model is presented to advisers in terms of Value at Risk and these measurements form the boundaries of the Synaptic risk categories. Value at Risk is a more valuable metric to advisers than drawdown or volatility per se.
As well as having access to a framework constructed of Strategic Asset Allocations, advisers can review asset allocations of existing investments and project on recommendations that they wish to make. The ability to illustrate the risk and return characteristics of these investments, and frame them in the context of a company's investment strategy, makes the challenge of compliance in the MiFID age manageable.
How is a strategic asset allocation created? The graph (above right) shows output of simulation, hot off the press, used to define the Level 1, Cautious Portfolio in the new Synaptic 1-10 range.
|UK Corporate Bonds||30%|
|Global Equity ex UK||5%|
|Emerging Markets Equity||0%|
To identify the Synaptic Risk strategic asset allocation boundaries, Moody's uses 6 core asset classes (although the Synaptic version of the engine can process up to 22 asset classes including various fixed income asset classes which include high yield corporate bonds and index-linked bonds, global equities, commodities, infrastructure and hedge funds).
Processing these strategic asset allocations creates our efficient frontier based on risk metrics that look like this:
|Asset allocation name||Term||Type||Boundaries - Min Gain||Max Gain||Mean Gain||Risk Rating|
|Moderately Cautious||10||Strategic||-10.71||18.25||3.92||< 3.0|
|Moderately Adventurous||10||Strategic||-18.19||28.85||6.15||< 5.0|
The efficient frontier for the Moody's/Synaptic model has the advantage of being constructed around potential loss (so Value at Risk), rather than the more usual volatility bands.
Synaptic Risk Rating – graph showing efficient frontier using Moody's Analytics 'Capacity for Loss' quotient / Min Gain (1 in 20 Year Loss %pa)
The benefit of the Moody's approach is that the model represents every viable investment outcome, including all sequence of returns. The simulation aspect of the exercise harnesses the power of probability through the generation of thousands of scenarios.
MiFID watchers will have been horrified at the guidance put forward in those directives that allows for investment growth assumptions to be made by averaging previous years' performance. Just looking at a table of asset class returns shows how impossible it is to predict returns in the short term by looking at past performance. It is only by stretching the term and capturing the trends and behaviours correctly that our projections become meaningful.
The benefit of the Moody's approach is that the model represents every viable investment outcome, including all sequence of returns. The simulation aspect of the exercise harnesses the power of probability through the generation of thousands of scenarios. Stochastic modelling was a bit of a joke a few years ago in the mainstream (though certainly not at institutional level – for example Moody's have been used by Royal London for years). Not anymore, as practitioners start to appreciate that the probabilistic approach leads to far more reliable forecasting, and the ability to embed a transparent process means that advice becomes consistent and reporting is made possible for compliance purposes.
|Asset allocation name||Type||Term||Min gain||Mean gain||Max gain||Attitude to risk||Risk rating|
|Cautious MPS||Strategic||10||-11.36||3.49||17.25||Moderately Cautious||2.80|
|Balanced MPS||Strategic||10||-15.82||4.87||23.84||Moderately Adventurous||4.00|
Asset allocation becomes a prism to evaluate any and every portfolio. In the rear pages of this magazine, you will see a wide range of fund groups who have participated in our risk rating. The Synaptic Risk Ratings are calculated by verifying the asset allocation of the investment with the asset manager, and projecting using the Moody's stochastic engine. This is mapped to the Moody's efficient frontier, giving us a risk profile that can be correctly mapped to the risk appetite of the client. The Strategic Asset Allocations have been developed to align perfectly with the ATRQ provided.
Notice that the methodology is different to the others in the market, who are less confident of their ability to project asset classes and therefore provide a committee-based 'rating' rather than a research-based rating. The problem of this, in our view, is that it removes the perspective of the adviser or investment committee, who should be providing the 'qualitative overlay' (with the assistance of RSMR's insights). The Moody's approach also means that the adviser has an objective means of monitoring an investment, rather than relying on a rating agency whose culture is to obscure the underlying dynamics and play down the fluidity of the risk universe.
See above for an example of how the Risk metrics make up the Synaptic risk ratings for Brewin Dolphin MPS portfolios. The 'min gain' is used to map to the boundaries of the risk categories. It is also the Value at Risk which must be matched to the client's Capacity for Loss.
Another great benefit of working with Moody's stochastic model is the ability to model inflation. This is a thorny challenge for advisers, because of the difficulty of gaining a useful model for inflation, and illustrations that accurately capture the ravages of inflation are not useful sales aids(!), unless the adviser is shown the likely impact of inflation on a cash holding as a point of comparison. This focusses the mind on the value of investment as a means of protecting wealth over the long term. Moody's calculate the impact of inflation historically to average 4.5% loss per year, in any 10-year period. The illustration (top right) indicates an 85% likelihood of losing money on 10-year cash investment (without costs!). The average return in the model at £8.6k is not as bad as in real life, as we are using the investment definition of cash as short dated bonds (1-year government bonds which have a greater growth potential than cash in the bank).
Putting it all together, for research purposes you can combine the projections of the asset allocations that underpin your advice. We are looking at the Strategic Asset Allocation, the allocation of the recommended investment, and we are overlaying costs and inflation for the full, beautifully compliant picture.
The next results (right) show the value of advice – an 85% chance of losing money has become an 60% chance of making money. Now if some charges could be shaved off, the picture starts to look more like our nominal projection (over 90% chance of gains)… but that is another story. The recommendation here was using a 10-year projection with 1.1% costs (adviser charges, platform and fund). Remember the 'min gain value' is our Value at Risk amount, the extent of the losses expected in a 'bad year', defined as 1 in 20 (5%) – enabling us to make our Capacity for Loss assessment with credible values. The blue line shows the projected returns of our strategic Balanced allocation. The green line shows the asset allocation of our recommended investment. The red line is the latter includes costs and inflation.