Dynamic risk models
Dynamic risk models can be used to measure and detect warning signals that may trigger a reduction in exposure. These include:
- Momentum Overlay: exposure is reduced on assets with a moderate or strong downward trend;
- Volatility Overlay: exposure is reduced at the level of asset classes and portfolios in response to the rapid increase in volatility;
- Correlation overlay: exposure is reduced across the entire portfolio when there is a short-term recovery in the risk of a bond sale spreading to other asset classes.
Execution is often overlooked, yet it plays an important role. What good are the most accurate signals if they can’t be executed efficiently? The focus on execution naturally dictates the instruments a strategy should use to construct a portfolio. In our view, the most suitable instruments for the application of dynamic risk models are primarily futures contracts. They are generally very liquid securities, but if the overlays activate, they can significantly reduce risk over a short period of time. The resulting strategy can therefore remain profitable if execution costs are kept low.
Dynamic risk management naturally requires more transactions than the average portfolio. This requires an execution infrastructure capable of reducing explicit costs (commissions, settlement fees, custody fees, etc.) and minimizing implicit costs (bid/ask spread, timing, market impact, etc.). To this end, investors looking to implement such a strategy need a sophisticated execution platform capable of effectively executing highly dynamic risk overlays. Building and maintaining such a platform is a big undertaking, requiring a lot of investment, experience, and quantitative talent to deliver results.
In summary, dynamic risk management – facilitated by an efficient execution platform – enables the use of leverage to enhance risk-adjusted returns, helping a strategy perform in good times and reduce declines in periods of market stress.
Case studies: Covid-19 and the GFC
Below, we highlight two examples of a sell-off in risk assets: Covid-19 and the 2008 Global Financial Crisis (GFC). meaningful, providing a useful illustration of how exposure and risk levels can practically be tailored to market conditions, and how such a strategy can help avoid excessive risk taking.
The Covid-19 crisis cannot be said to have hit global markets “unexpectedly”. By the end of December 2019, the infection had already started spreading out of control in China and Southeast Asia. Yet global markets ignored its impact and global stock and bond markets continued to rally. It was only at the end of February that the markets became aware of the pandemic. The response of a theoretical volatility targeting strategy in terms of asset class exposure is highlighted in Chart 1a.
In January and most of February 2020, when markets were quiet, the strategy required a gross exposure of 350% in order to achieve its balanced risk profile and 10% volatility target. In this benign period, the strategy was fully invested and had nearly double the volatility of a 60/40 portfolio (Figure 1b) allowing it to better participate in the market rally and outperform 60/40 in January and early February ( Figure 1c).
In late February, when the pandemic began to infect global markets, dynamic risk overlays reacted quickly; volatility first, then momentum, and finally the correlation overlay. This reduced the exposure from about 350% to about 30% (Figure 1a). In terms of volatility, however, the volatility of a 60/40 portfolio increased significantly from 5% to 30% (Figure 1b). This is because the exposures in a 60/40 portfolio are static and the risk of the portfolio therefore reflects higher volatility in the markets. The continued decline in risk assets made the strategy’s outperformance clearly visible in March 2020, although a rebound in April led to underperformance (Chart 1c).
In summary, although the volatility targeting strategy entered the Covid-19 market event with 3.5x total leverage, it outperformed an unleveraged 60/40 portfolio. This was achieved thanks to the two facets that we presented previously:
- Diversification and balanced risk-based allocation that mitigated initial equity losses through mid-February;
- Risk management overlays, which significantly reduce risk in the event of increasing market volatility.
Figure 1. Covid-19 case study
Simulated past performance is not indicative of future results. Illustrative example. For information only.
Source: Man Group, Bloomberg. A management fee example of 0.95% management fee has been applied. A 60/40 composite index is composed of 60% MSCI World Net Total Return Index and 40% Barclays Capital Global Aggregate Bond Index (USD hedged).
In the eyes of many commentators, the Covid-19 crisis had its closest parallel during the GFC in 2008.
The theoretical volatility targeting strategy entered the period around 2.5x with leverage (Figure 2a) but quickly reduced exposure to around 0.5. As a result, the strategy did not experience the same increase in risk as a static 60/40 portfolio (Figure 2b), and it outperformed as markets continued to decline (Figure 3c).
However, we see in this example how the strategy shifted as markets stabilized towards the end of 2008 and into 2009. Total exposure does not reach pre-crisis levels because volatility remained high, as can be seen in the case of a static 60/40 portfolio in Figure 2b.
Figure 2. GFC Case Study
We have illustrated that, for two of the most significant market events in recent memory, the leverage of a volatility targeting strategy before a crisis does not translate into significant risk or losses during the crisis itself. -same. In contrast, a traditional 60/40 portfolio with less leverage at the start of each crisis ended up being more risky during crises and suffered greater losses.
We believe this demonstrates that leverage is not always a good indicator of portfolio risk, especially in the presence of diversification and active risk management.
This post was funded by Man Group