Bonds are riskier than stocks, and investors with heavily bond-oriented pension portfolios are at the risk of “under-funding their retirement”, according to Neptune Investment Management.
Meanwhile, those using absolute return funds are opening themselves up to “unpleasant surprises” due to a lack of accurate historical data on the products, the asset manager concludes.
Around 39% of funds in the targeted absolute return sector, which use different techniques to make consistent gains with less volatility, failed to beat the cumulative inflation rate of 4.5% over the past three years.
“That is a very nasty surprise that was unforecastable because there was no data on these types of products,” says James Dowey, chief investment officer at Neptune.
Stocks: Less Risky Over The Long Term
The accepted practice for measuring investment risk is to compare standard deviation, a drawdown, or the worst outcome over a short holding period, explains Dowey.
On these measure, using data from the Barclays Equity Gilt study between 1900 and 2015, standard deviation of UK stocks is 19.8% versus 13.5% for gilts. The worst one-year outcome for UK stocks is -58.1% compared to -32.8% for gilts.
“This paints stocks in a risky light,” admits Dowey. A reduction in equity exposure and increase in exposure to bonds and/or alternative assets as clients’ risk tolerance falls would be most financial advisers’ prescription.
Medicines would generally come in the form of absolute return or target-date funds, which dial down stock exposure as the investor nears retirement age.
However, if you take the same data from the Barclays study and compare the risk measure over 10 and 20-year periods, the results are “very interesting”, according to Dowey.
Over a 10-year period, the annualised standard deviation for UK stocks and gilts are both 5.2%. Over a 20-year period, the standard deviation for UK stocks is 6.5%, versus 7.9% for gilts.
Over a 10-year holding period, meanwhile, the worst outcome for UK stocks is -7.94%; for gilts it’s -10.75%. Over a 20-year holding period, the worst outcome for UK stocks is -4.28%; for gilts it’s -10.85%.
Digging further down into this data paints an even starker picture. Dowey shows histograms of real total returns over 10 and 20-year holding periods of both UK stocks and gilts during that 115-year timeframe.
In both cases, gilts deliver negative returns almost half of the time. For stocks, it’s about 15% of the time over 10 years and just 2% over 20 years.
“That is a radically different risk profile when we extend the time horizon and start dealing with the outcome risk of underfunding retirement.”
While he accepts the short-term ups and downs of the market are “incredibly stressful” and that it’s important to manage that, “it’s also stressful to under-fund retirement”. “It is absolutely essential that we elevate this risk above the risk of short-term ups and downs.”
Equity-Bond Dynamics Set To Change
The bond versus equity risk debate has not really mattered much in recent years to the 30-year bond bull market we’ve had. However, that may now be coming to an end. This means we may revert to longer-term trends, which look “much choppier”.
Dowey notes the fall in the UK Government bond yield from 15% in the early 1980s to less than 2% today contributed to really good total returns for bonds. Ignoring bonds’ inherent riskiness has, therefore gone unpunished.
But that punishment is lurking round the corner, he claims. With yields today at such a low ebb, “it will be extraordinarily difficult for bonds to deliver positive real returns”. The starting yield accounts for much of an investor’s 10-year returns. “We’re starting from a rock-bottom point so if we get any inflation at all over the next 10-20 years it’s going to erode the gains from that type of strategy.”
Finally, while bonds are seen as good diversifiers for equity portfolios, “this has been a relatively recent phenomenon”. In the past 10 years the correlation between the two asset classes has been around 0.7%; before 2005, there was only a short period in the early 1970s when the correlation was negative. Over most of the period between 1909 and 2005, the correlation was around 0.6.
