Earlier in the week I was at a meeting whereby one of the attendees suggested that in a balanced portfolio (~70% equities) the equities proportion adds around 90% of the risk, and then in yesterday’s Australian Financial Review I see a headline quote from Stephen Nash (FIIG Securities) that said, “If investors replaced half their equity portfolio with long bonds, they could cut their risk by half and cut the return by about 1 per cent” .
Over the last 100 years or so, equities have outperformed bonds and bills by around 4%pa around the world so whilst I’m unsure a loss of 1% in return potential is a reasonable assumption these comments certainly got me thinking about what has actually happened with regards to equities versus bonds in terms of risk.
Source: Index data originally sourced from Morningstar via van Eyk
The above tables shows some analysis I did for both local and global equities and bonds over the last 21 years or so. I also combined the returns to produce a 50% allocation each to bonds and equities as per Stephen Nash’s suggestion.
These results who that, certainly over the last 20 years, declining interest rates around the world certainly resulted in much better returns versus equities on a risk adjusted basis (sharpe ratio) than equities and therefore resulted in a significant risk reduction of the 50/50 portfolio. Maximum drawdowns for the 50/50 portfolios more than halved over the last 10 and 20 years and this significant risk reduction was also evident when looking at theoretical Value at Risk (calculated using historic return and risk figures) or the actual 5th percentile of returns.
The time period where Stephen Nash’s comment didn’t quite ring true was in the 1990s when Global shares were the strong performer and the “Risk” reduction of the 50/50 portfolio didn’t quite make it to half. However the risk reduction was still significantly reduced (and more than halved) when looking at Sharpe ratio, theoretical value at risk, or the actual 5th percentile of returns.
This overly simplistic analysis does suggest that replacing half an equities portfolio with bonds may significantly reduce risk in terms of volatility or in terms of downside risk. However, what we face today are bonds that are yielding around no more than 4%pa across all maturities (and I am talking government bonds, not credit). So unlike the last 20 years where yields have decreased (5 years government bonds decrease from more than 10% in 1995 to around 3.6% today…RBA data I looked at only went back to 1995), the risk to bonds in the long term is increasing yields which means downside risk is far greater than we have seen for some time.
What is in favour of bonds is the enormous volatility equity markets so I tend to agree with Stephen Nash that bonds will half the short term risk of the equities portfolio but I do have fears for the long term risk of such a strategy.