The paper, co-authored by Innova Asset Management managing director Daniel Miles and Milliman principal and consultant Joshua Corrigan, argues that the underlying assumption behind so-called ‘balanced’ portfolios – that equities and bonds are negatively correlated – has proven to be wrong.
“From the 1960s to 2000, the correlation was largely positive – in both falling markets (such as the 1970s) and rising markets (1982 onwards),” said the paper.
“Essentially, if ‘growth assets’ go down, the whole portfolio goes down. In other words, the concept of risk reduction has been completely misunderstood,” it said.
Part of the problem is that volatility is simply a proxy for risk, and traditional portfolios tend to have far more of their ‘risk budget’ tied up in equities than they think, according to the paper.
A breakdown of the short-term risk allocation of a typical balanced portfolio shows that 52 per cent of the portfolio has valuation risk, and 23 per cent has inflation risk.
When it comes to the long-term risk allocation of a 60/40 fund, around 51 per cent of the portfolio is tied to economic growth risk and 27 per cent is in inflation risk, according to the paper.
Instead, investment managers would be better served spreading their risk evenly among 10 factors, including: economic growth; valuation; inflation; liquidity; credit; political risk; momentum; manager skill; option premium; and demographic shifts.
“The only real insurance against black swan events, big regime shifts and large drawdowns is to seek out multiple sources of risk premia across a host of asset classes,” according to the paper.
“Allocating capital across asset classes and investment styles represents superficial diversification if payoffs are exposed to the same set of risk factors,” it said.
The paper pointed to a number of academic studies that analysed the performance of porfolios that are based on risk factors.
“Briand, Nielson and Stefek (2009) found using an equal weighting across 11 style and strategy risk premia from 1995 to 2008 had similar returns to traditional 60/40 portfolios but with 65 per cent less volatility,” the paper stated.
Podkaminer (2013) found that a simple factor portfolio “historically achieved a slightly higher level of return than the traditional portfolio while taking on about one quarter of the volatility”, according to the paper.
However, there is a “level of naivety” when it comes to constructing portfolios depending on risk factors – the portfolio is not optimised to target the risk factors relevant for the end client.
