Amy Middleton and Sylvia Huang investigate the diversification benefits to bond and equity portfolios of both unhedged and hedged foreign exposure
On the equity side, the allocation to international assets which was found to best maximise the diversification effect varies, for example Clarke and Tullis (1999) found it to be 20% to 30%, while Lewis (1999) estimated 40% and Michaud et al (1996) stated 90%. With respect to international bonds, the amount to allocate is even less clear.
Further complications are added because foreign assets are not only subject to price risk in local terms, but also to risk associated with currency fluctuations.
Yet despite this, many fund managers still choose to ignore the impact of currency within their international portfolios - but are they wise in doing so? Essentially this is a case of asking how large an international exposure needs to become before the additional risk created by currency starts to matter. Many involved with the management of currency argue it should always be dealt with regardless of the size of international allocation, however, little consensus appears to exist within the consultant space.
In this article we summarise the results of an empirical study on the effect of international diversification and the impact of currency risk on the foreign asset allocation process.
Our analysis used equity and bond portfolios based on the Morgan Stanley Capital (MSCI) and the Salomon Smith Barney (SSB) country indices respectively, for four bases currencies: USD, EUR, GBP and JPY. The period under study was January 1984 to March 2007 .
Domestic versus foreign assets
The returns of an international asset comprise of three individual sources of risk: the asset's variance in local terms, the foreign currency variance, and the covariance of them both. As currency tends to be more volatile than bonds, it will contribute a greater proportion of the total risk in an international bond portfolio than in an equity one. Therefore, the diversification effect achieved in an international bond portfolio will be short-lived compared to the equity portfolio. To assess the magnitude of this, we quantified the risk of a series of bond and equity portfolios in which the allocation to foreign assets was gradually increased from 0% (i.e. purely domestic portfolio) to 100% (i.e. purely international portfolio).
For equities, the diversification effect was substantial and maximised when the allocation to international assets was c.40%; this concurs with the results of Clarke and Tullis (1999) and Lewis (1999). On the bond side, however, the diversification effect was lower and disappeared quickly, with risk being minimised with a c.10% foreign allocation.
Hedging an increasing proportion of international exposure reduces portfolio risk for both equities and bonds. The impact is greatest on the latter though, for example, moving from unhedged to fully hedged decreased the total risk for bond portfolios by c.60%, whereas the reduction for equities was 20%. But can the minimum risk portfolio be improved upon by varying both the amount allocated to international assets and the hedge ratio mix? In order to test this, we created portfolio combinations in which the percentage of international assets and percentage hedge ratio were varied simultaneously in increments of 2.5%.
In the case of bonds, we see that when the international proportion is small, then the marginal benefit of increasing the proportion of the foreign assets hedged is negligible. However, once the allocation exceeds roughly 10%, currency risk takes effect and hedging is essential if the original level of portfolio risk is to be maintained.
For example, consider the case of a USD-based bond manager with a 25% allocation to foreign bonds. If he were to increase his foreign asset allocation to 50%, then the total portfolio risk would also increase and shift into a different heat map banding. However, the same level of original portfolio risk could be maintained if approximately 20% of the exposure was hedged.
This simple example demonstrates how hedging can facilitate an increase in allocation to foreign assets without inadvertently acquiring any additional risk.
For bonds, the minimum risk portfolio occurred when the allocation to foreign assets was at least 50%. Hedging facilitated a markedly higher allocation to foreign assets and even improved on the minimum risk portfolio found in the non-hedging case.
For equities, only when foreign equity approached 40% did currency exposure begin to add to total risk. In general, for all bases, the combination of international assets and hedge ratio that minimised total portfolio risk for equities and bonds was very similar, i.e. a high level of international assets and a high ratio of hedge coverage.
Risk from international assets
Analysis of the non-hedging case revealed a striking difference in the impact on risk when adding international assets to bond and equity portfolios. We found that over the long term horizon, diversification was greatest, i.e., total portfolio risk lowest, with a 40% allocation to foreign assets for equities and 10% allocation for bonds.
Once these levels were exceeded though, currency risk outweighed the diversification benefit and a hedging programme would be required to maintain the original levels of risk.
The introduction of hedging not only permitted a larger allocation to international assets (average 70% for equities and 85% for bonds), but it also succeeded in improving upon the minimum total risk portfolio found in the non-hedging case.
Whilst the results of this study may be sensitive to shorter time horizons, our analysis nonetheless provides some insight into the risk of currency in international portfolios and highlights the need for it be monitored closely.
An innovative funding structure has been agreed for Croydon Pension Fund. However, there are some concerns about the arrangement. Stephanie Baxter reports
Some 52% of red flags raised by schemes on suspected scam pension transfers involve advisers or unregulated introducers, a report by the Pension Scams Industry Group (PSIG) has claimed.
The Norfolk Pension Fund has been successful as the lead plaintiff in a class action case that went to jury trial in California involving securities fraud.
In this week's Pensions Buzz, we want to know whether bosses should have to pay into the same staff DB scheme as their workers rather than their own executive pension fund.