Back in July, as part of our Ideal Portfolio series, we looked at the concept of risk management in detail (Ideal Portfolio: Managing Risk, 24 July 2015).
We analysed US versions of the three Investors Chronicle Ideal Portfolios (cautious, balanced and adventurous strategies - see diagram for allocations). There was also a complex adventurous portfolio that specified S&P 500 holdings, spread the equities allocation internationally, and added a property exposure to US real estate. The portfolios were assessed according to a standard measure of value at risk (VaR) and the proprietary RealRisk™ model of PrairieSmarts, the US risk software company. A reminder of the principles underpinning these tools is provided in the box 'Risk metrics' below.
Around the world, 24 August 2015 was a bad day for equity markets, the S&P 500 closed 3.94 per cent below where it started trading and the four US Ideal Portfolios saw significant losses. Of course equity markets are volatile, which is the trade-off for the additional returns premium that they offer. Successful investing is about managing exposure so that the risk/reward ratio is as favourable as possible. While no risk model can predict market falls, the more accurate an approximation investors can make of the potential losses they could suffer, the greater the opportunity to spot and take steps to mitigate worrying signs of imbalance in their portfolios.
What did risk models allow for a sell-off?
Estimates in the table below are of extreme one-day losses (with a 0.5 per cent probability of occurrence) for each of the four US Ideal Portfolios. The two middle columns show the average percentage losses risked by investors according to PrairieSmarts and VaR respectively. Looked at another way, the models are saying that 99.5 per cent of reversals should not be as bad as the stated figures. In the final column, we can see what actually happened on 24 August, which serves as a good reality check as to whether VaR and/or PrairieSmarts accurately reflect the true 'riskiness' of the portfolios.
| Portfolio | 1-Day PrairieSmartsRealRisk | 1-Day Value at Risk | Actual fall on 24 August |
|---|---|---|---|
| Cautious | -1.853 | -0.950 | -1.302 |
| Balanced | -3.183 | -1.588 | -2.378 |
| Adventurous | -4.641 | -2.214 | -3.053 |
| Complex Adventurous | -2.892 | -1.497 | -2.279 |
Taking the balanced strategy as an example, PrairieSmarts predicts that on the very worst days for the stock market, losses are likely to be on average -3.183 per cent. The implication is that the portfolio's -2.378 per cent fall on 24 August was not at the extreme boundaries of probability. This is in stark contrast to the VaR model's assessment of the balanced strategy, which suggests that losses on the worst 0.5 per cent of days, should on average be of a magnitude of -1.588 per cent. From both past empirical evidence and recent experience of market volatility, it seems that there is a danger that models such as VaR seriously underestimate risk.
Planning adequate diversification of risk
Understanding downside exposure helps enormously in deciding an appropriate mix of investments. It is also useful to know how the level of diversification affects the overall risk profile of a portfolio. Dividing the level of extreme risk (for 0.5 per cent probability events) for a portfolio of investments, by the sum of risks estimated for the individual holdings, gives a diversification index. A higher diversification index demonstrates that the pattern of co-movement between securities mitigates risk overall; in theory making the portfolio more robust in periods of market stress. For our US Ideal Portfolios, using figures arrived at from PrairieSmarts' risk estimates, reality has supported theory. As the table below shows, the higher the diversification index, the smaller the falls in value on 24 August.
| Portfolio | Diversification Index | Actual portfolio movement 24 August |
|---|---|---|
| Cautious | 0.417 | -1.302 |
| Balanced | 0.338 | -2.378 |
| Adventurous | 0.183 | -3.053 |
| Complex Adventurous | 0.407 | -2.279 |
There are encouraging signs for the complex adventurous portfolio, which while not weathering the recent market volatility as well as a cautious strategy, has proved more robust than the balanced or adventurous allocations. When the US portfolios were run in February, the past data since 2006 showed the annualised returns had been 5.22 per cent for cautious, 6.79 per cent for balanced, 7.25 per cent for adventurous and 8.17 per cent for complex adventurous. It is pleasing to note comparing the performance of the two adventurous strategies that allocations can be tweaked to achieve superior returns, not just in absolute terms, but also relative to risk.
