Portfolio risk when adding government and corporate bonds
As I suggested in Chapter 10, there could be merit in adding risky government and corporate bonds to the portfolio, but we did not incorporate them in the examples in that chapter. This is not out of neglect, but to keep the discussion simple and practical. If we attempt to add risky government and corporate bonds to the excel model outlined earlier (see Chapter 10) we would significantly add to its complexity and you would need to have a view on complex things like asset correlations. Taking an average correlation between risky government bonds, corporate bonds and equities would be too much of a simplification. During times of great distress, correlations between asset classes tend to go up (just like they do between various equity markets), and as we run the multiple scenarios we would need different correlations for different states of the equity markets, resulting in something like this:1
| Equity return | Sub-AA government bond correlation | Corporate bond correlation |
| 15%+ | 0.35 | 0.40 |
| 5% to 15% | 0.28 | 0.32 |
| −5% to 5% | 0.22 | 0.28 |
| −15% to −5% | 0.34 | 0.38 |
| Below −15% | 0.55 | 0.67 |
What this adjustment incorporates is that the addition of risky government and corporate bonds adds less diversification when you need it the most, namely when equity markets are poor. Since the bonds I suggest adding here are sub-AA rated return generators, in a crisis they may not be seen as a ‘flight to safety’ and go up in value. Your minimal risk asset will probably be seen as a safe haven and increase in price.
At the risk of oversimplification, other government and corporate bonds will add less risk to the portfolio than equities, but will not be risk free. Depending on the exact mix of sub-AA government bonds you add, you could perhaps see the other government and corporate bonds as adding half the risk of equities, although realise that this more a guesstimate than a precise calculation.
Adding great complexity to the simple model misses the point that all this is far from an exact science. The point of the excel model is to show that even the best-laid plans are subject to market risk, and to point out how we can incorporate this element of chance in thinking about our portfolio allocations and planning. By making the model excessively complex we run the risk of giving a false sense of precision.
____________________
1 This is still a significant simplification. The correlations in the various return scenarios would be an average expected correlation. Since each crisis is unique there is no guarantee that the actual correlations would be like those in the table.