News and articles

Our Risk Framework: A decade on

Efficient frontier

Asset allocation is responsible for 90% of the variability of portfolio returns, so it’s vital to apply a rigorous approach, one which Advisers trust and clients can depend upon to deliver a level of risk adjusted return consistent with their expectations. Parmenion’s Strategic Asset Allocations have been in place for a decade and underpin the risk framework governing the majority of investment solutions on our platform. How are these asset allocations derived and do they remain fit for purpose?

What is the academic basis of our approach?

Parmenion’s approach is based on Modern Portfolio Theory, the brain child of Harry Markowitz. First published in 1952 and expanded as his PhD in 1954, later winning him the Nobel Prize for Economics in 1990, it introduced the concept of the Efficient Frontier. The theory postulates that Risk and Return can be related and that the combination of assets into portfolios can improve your risk adjusted return. This outcome, which is not intuitive, is an effect of less than 100% performance correlation between individual assets. In simple terms, a portfolio is more likely to have some assets going up in value more often if it has a good spread of different asset classes, when its well diversified.

Mapping individual assets in their almost infinite combinations on a risk and return chart allows you to identify which portfolio combinations provide the highest level of return for your client’s specified level of risk. Join all these optimal portfolios together with a line and that will reveal the Efficient Frontier, as illustrated below.

Some practical questions

Implementing the theory in the real world means tackling a number of difficult challenges which revolve around the simple question: ‘how do you measure risk and return’? Here are some of the related problems which arise when putting theory into practice and what approach we have used over the last 10 years:

1. Which asset class indices should be used and is the data reliable?

Our data is sourced from FE Analytics and Parmenion pay for all the necessary index licences. We have a preference for mainstream indices which often form the benchmarks of our passive and active funds.

2. What frequency and over what period should the analysis be based?

We analyse monthly data, with income re-invested, over a rolling 20 year period. This ensures we have a statistically valid sample size.

3. Should all time periods of analysis be equally weighted? Or some periods discounted?

We don’t include data which is less relevant to 2018. For example if it reflects very different political, social, regulatory or economic environments from the distant past.

4. Should analysis be historic or based on forward-looking models?

We see predicting the future as subjective, and as such we see historic data covering multiple business cycles as the better reflection of enduring asset class relationships.

5. How many portfolio combinations should be analysed?

We analyse all combinations including sub-asset classes.

6. If a program is used to assist with the larger number of calculations, how does it work?

We do not use a third party program because we want to understand and control the modelling in exacting detail. Furthermore we want to fully understand the risks and controls which are utilised in the process.

7. Should fees or tax be included?

Fees and tax are excluded as they vary by investor.

8. Should any qualitative limits or constraints to the model be applied?

Sensible constraints are applied. No more than 20% of a portfolio can be in one sub-asset class. This is to ensure the output is appropriate and portfolios are well diversified, guarding against potential ‘black swans’.

9. How often should the analysis be refreshed?

We re-perform our analysis on an annual basis.

What is the result of this work?

Our Strategic Asset Allocations can be found published in our Quarterly Investment Reports, available publicly here. They are also included in the Investment Management Reports generated when placing business with Parmenion. Our client services team can show Advisers new to Parmenion how these can be quickly and efficiently produced.

As our asset class data is refreshed on an annual basis, the shape of the efficient frontier we create will change. But after careful review, our target asset allocations have not needed to be changed since 2008. Their merits are evidenced by the alignment of our risk graded portfolios over every reasonable time frame and by the strong risk adjusted returns achieved against relevant benchmarks. The theory is working in practice.

Any modelling approach will have flaws, but as long as you understand them you can control issues as they arrive. It is clear that the Efficient Frontier offers a rigorous, dependable, academic approach and we have no reason to doubt its value in the future. Markets may surprise us, but the relative movements of our portfolios should not leave clients with unwelcome surprises.

If you have any queries as a result of the above article, then please speak to your relationship manager, contact client services on 0117 204 7600 or email and we will be happy to assist.