Diversification did work in 2008, if you had some. This conclusion may be a surprise but looking at the returns you will see that equities and credit did very poorly, government bonds did very well and real estate did okay. The problem was that many plans thought they were diversified but werenât, in fact they had nearly an entire portfolio exposed to the same cyclical economic/corporate profit/credit risk factors. A typical institutional portfolio of 60% equity (½ Canadian, ½ Foreign) and 40% bonds (½ in credit, which was typical of fixed income credit overweight yield carry strategies) had an 80% exposure to the same common factor because equities and credit were highly correlated. The 60% in equities was all correlated because geographic diversification turned out to be a fantasy.
2008 returns are shown in the table below. Canadian government bonds returns were +11.5% and it was the only asset class that gave a meaningful offset to equity returns.
| Asset Class* | 2008 Annual Return |
| Cash Equivalents | 3.5% |
| Cdn Govât Bonds | 12.1% |
| Cdn Corp Bonds | 0.2% |
| Cdn Universe Bonds | 6.4% |
| Canadian Equity | -33.0% |
| Global Equity â Hedged | -41.0% |
| Global Equity â Unhedged | -25.8% |
| Emerg Mkts Equity | -41.4% |
| Real Estate | 3.1% |
| Real Return Bonds | 0.4% |
* Asset class data from Dex, S&P/TSX, MSCI and IPD.
The lesson learned is that diversification still works, but there must be assets in portfolios that are diversifiers in times of financial stress. Government bonds were that asset class in 2008 and will likely repeat as the top asset class in future periods of financial market stress. Going forward the return for assuming credit risk may look attractive, but abandoning government bonds (real or nominal) may not be a good idea. All assets should have a purpose and not all should be for generating excess returns. Government bonds might play an important defensive role in most funds.