Restoring International Diversification

Many equity investors believe they are diversified because they hold the developed markets benchmark. Yet US equities now account for around 71% of that benchmark, up from about 40% at the time of the Global Financial Crisis, and for close to 80% of its risk. What looked like international diversification in 2008 now looks much closer to a concentrated position in a single market. This paper asks what an investor should do about it, and makes four points.

The benchmark carries a view on US outperformance, and we make it explicit. The capitalisation-weighted benchmark is not neutral. Using reverse optimisation, we recover the return expectations embedded in market prices: current regional weights are consistent with US equities outperforming the rest of the developed world by about 2% per year. Following the benchmark is therefore not the absence of a view; it is accepting the market's view and embedding it in the portfolio by default.

A framework to blend the market's view with your own. We provide a systematic regional allocation framework, in which the market-implied view is blended with the investor's own view in proportion to the confidence the investor places in it. Confidence is not an abstract parameter; it maps directly into a tracking-error budget. An investor with no confidence in their own view holds the benchmark; an investor with full confidence obtains the allocation implied by that view alone; intermediate confidence traces a smooth path between the two.

The key practical contribution is governance. Many investors are concerned with the issue of US dominance. However, addressing it through ad hoc regional weights, or through alternative schemes such as GDP weighting, is hard to justify for fiduciaries: such allocations remain opaque about their views on expected returns and carry no formal justification. Our framework instead integrates explicit investment beliefs in a systematic way, so that a CIO can demonstrate to boards and trustees that the allocation reflects a stated belief, held with a stated level of confidence and implemented within an approved tracking-error budget. The deviation from the benchmark becomes the documented consequence of the institution's investment beliefs rather than a discretionary bet.

The key empirical result: even a modestly neutral view moves the needle. In current conditions, an investor who simply assigns the same long-term expected return to US and Developed-ex-US equities would reduce the US weight by about five percentage points under a tight 1% tracking-error budget, and by about thirteen percentage points, from roughly 71% to 58%, under a 3% budget. Applied consistently over four decades, these allocations would have reduced portfolio volatility, and the size of the maximum drawdown, without sacrificing returns.

A growing number of investors are uncomfortable with the heavy US weight of global benchmarks but lacks convincing tools to address this issue and communicate recommendations to their stakeholders. We provide a framework that allows investors to restore international diversification by tilting their portfolios towards a neutral stance on expected returns across regions. Our systematic framework makes return assumptions explicit, accounts for tracking-error constraints, and is fully flexible to accommodate a variety of views on expected returns across regions. This enables investors to restore diversification not through a discretionary call on regional allocations, but as the documented consequence of the beliefs they hold.