The conventional approach to asset allocation — which began with the staple 60/40 policy portfolio of stocks and bonds, before morphing into highly leverage-sensitive alternatives — is approaching a generational inflexion point in my view.
As it continues to serve as the industry default setting, I recently conducted a dissection of the characteristics of the 60/40 policy portfolio across 26 countries and long periods of time. The key takeaway from this research is straightforward — the 60/40 policy portfolio is riskier than most think, perhaps dangerously so. There are a number of grounds on which I assert this claim.
First, the real returns to a 60/40 portfolio exhibit dramatic time variation or regime persistence, which renders long-term (unconditional) average returns a frail reed to lean upon. Put another way, it is the variability around the average, not the average itself, that is the striking feature of 60/40 portfolio returns.
For instance, average returns to a domestic 60/40 portfolio for U.S. investors have barely kept pace with inflation across seven of the past 11 decades. The other four decades have done all of the heavy lifting in generating meaningful returns for investors, although importantly, these periods were associated with unrepeatable or rare events, namely post-world-war recoveries in the 1920s and 1950s, and the one-off secular decline in inflation and bond yields spanning the 1980s and 1990s.
On a more worrying note, a 60/40 portfolio has produced negative real returns over a rolling 10-year holding period for almost a quarter of a 111-year period in the U.S. market, commencing in 1900. (Note also that U.S. capital market returns are biased upward in a global context, as the U.S. has never lost a war on its own shore.) An examination of other developed and emerging countries reveals that severe and lengthy 60/40 portfolio drawdowns are commonplace. The real value of a domestic 60/40 portfolio for a Japanese or Irish investor is today worth the same as it was in, respectively, 1987 or 1993.
Analysis of the underlying statistical, valuation and economic risk factor exposures to a 60/40 portfolio are similarly disconcerting.
On statistical grounds, there is serious instability in the diversification properties of stocks and bonds (in only one of three macro conditions are stock and bond returns negatively correlated), as well as highly unbalanced contributions to portfolio risk (for instance, 95% of overall portfolio variability is contributed by the 60% equity tranche). These two findings render bonds as inefficient hedges in my view.
Valuation risks are also a problem for the policy portfolio today. Low or negative real bond yields across the U.S., Japan, Germany and U.K. now impair the ability of fixed income to act as a shock absorber as in the past. It is worth highlighting that real Treasury returns were negative for 45 years from similar valuation levels as today. (The risk parity literature should be viewed with trepidation for this reason.)
With respect to stock valuations, the secular 1980s and 1990s equity bull market was driven by a fall in real bond yields, a decline in inflation uncertainty, and a rise in the corporate profit share of gross domestic product — none of which can be repeated from today's starting point. I forecast a paltry 1% per annum in real terms from a 60/40 portfolio in the U.S. over the next decade, well shy of long-term averages (around 4%) and pension fund requirements (around 5%).
Finally, on macroeconomic grounds, the 60/40 portfolio is in effect 100% short inflation and sovereign risk. A long-only stock and bond portfolio is ill-equipped to navigate a stagflationary environment similar to the 1970s. It is worth pointing out that over the course of the 1970s — a period bearing some semblance to today — a 60/40 portfolio lost 3% per annum in real terms across the G8 markets as both stocks and bonds declined.
In short, my research finds the 60/40 policy portfolio unreliable, inefficient and ill-equipped for the challenges ahead. The institutional investment industry must do better.
A “second generation” industry response to some of the challenges posed by the 60/40 policy portfolio has simply been to expand into more assets, following from the insights of modern portfolio theory. Leverage-sensitive alternative assets, such as hedge funds, real estate, private equity and infrastructure have been the primary beneficiaries. This is a highly questionable solution, as most of these assets are short a common risk factor, shared with equities, in the form of systemic liquidity risk. This soft underbelly was revealed in dramatic circumstances in 2008. Note for instance that 14 out of 18 hedge fund strategies suffered their worst-ever drawdown at the same time, despite them pursuing what appeared to be different strategies in different markets.
Against this backdrop, I expect the asset allocation process adopted by the industry to evolve in a number of directions in coming years.
First, diversification by common risk factors and risk premiums will replace conventional asset class silos as the key building blocks in portfolio construction. Rather than rely on historical correlations observed over benign sample periods, a more robust yet intuitive risk-factor-based approach, focused on the underlying sources of asset returns, would have signaled the alarm bell for various investment strategies long before the crisis hit.
Second, the biological concepts of regime persistence and adaptation will increasingly feature in risk management processes. This reflects the well-known empirical finding that asset returns are not randomly or normally distributed over time, but rather highly regime dependent — that is conditional on the recent past.
Finally, the line of demarcation distinguishing strategic and tactical asset allocation will become blurred. Best practice from the endowment model, with a traditional focus on long-term risk premiums, and the macro hedge fund community, which tends to focus on path dependency and tail-risk hedges, will be fused together. This will likely be reflected in “barbell” portfolio structures, designed to harvest risk premiums in good times, while explicitly protecting capital in more adverse conditions.
Bradley A. Jones is a Hong Kong-based macro investment strategist for Deutsche Bank AG. His commentary is adapted from his research report “Rethinking Portfolio Construction and Risk Management: A Third Generation in Asset Allocation.” His complete report is available at P&I's white paper section.