Currency risk out of the closet
Many may not be familiar with the concept of a ‘currency benchmark’. It results from the explicit recognition of currency risk in a portfolio, and, for example, might be ‘World ex UK equities 100% hedged’. The ‘100% hedged’ is the key point here: the process of choosing benchmarks is in part setting a ‘norm’ against which managers can be measured, but more importantly is expressing a strategic preference for a particular asset class or asset mix. A fund which has a 50% hedged benchmark for its international equity portfolios is expressing the view that is wishes to partially eliminate currency exposure from its international equity portfolio.
In making an assessment of the most appropriate currency benchmark, or benchmark hedge ratio, most funds will ask their consultant (or other adviser) to include the question within a strategic asset allocation review or as part of an asset/liability study. Many consultants have now developed the analytical tools and data to do this.
The principal measure that is targeted is the overall equity portfolio volatility; in particular (for a defined benefit scheme) the relative volatility of the equity portfolio versus a fixed income benchmark (representing the liability profile).
I have taken 20-year UK data to illustrate the approach. In Figure 1, the blue line shows the relative volatility (in this case versus index-linked gilts) of a portfolio of varying proportions of domestic and international equities – 100% domestic on the left and 100% international on the right. The international is unhedged. The graph clearly shows the advantages of international diversification, and indicates a domestic/international mix of about 50:50 gives the lowest relative volatility.
This analysis, however, ignores currency risk. What happens when a further line, showing the effect of fully hedged international, is added?
The answer, as the red line shows, is that it has a hugely material effect for a fund which has already moved to an optimum, or near optimum, domestic/international mix.
Using hedged international equities rather than unhedged moves the minimum risk position volatility down by 0.5%, and the optimum mix from 50% international to 65%. A 0.5% volatility reduction may not sound like much, but it represents the volatility reduction from moving from 25% unhedged international to 50%.
On the face of it, this argues very strongly indeed for a fully hedged benchmark. But, as is often the case, this is not the end of the story.
When a fund undertakes currency hedging of its international assets, it has to do so by entering into – via its manager(s) or a specialist currency overlay manager – forward foreign exchange contracts to sell foreign currencies for delivery at some future date. But the foreign exchange market only has liquid markets out to one year forward (six months for some currencies), and the equities will be held for much longer than that. This means that the currency contracts have to be regularly settled in cash (ie, closed out by a reversing deal when they mature), and then resold for a date in the more distant future. This creates a regular stream of cash payments and receipts, which are by their nature uncontrolled (since they depend on the movements in the foreign exchange markets), and which could be large.
Figure 2 gives the 12-month rolling cash flows expressed as a percentage of the international portfolio for the portfolio in Figure 1 over the past 20 years. These (large) numbers may look like the manager’s problem, not a benchmark problem, but this is not so.
Most serious students of the currency markets now accept that the expected return from currency risk, after adjustment for interest rate differentials, is zero. This means that the expected returns for hedged and unhedged international can be the same only if the cost of implementing the hedging is also zero.
We estimate that the cost of continuously rolling a passive hedge to replicate a 100% hedged benchmark is less than 10 basis points a year including manager fees. However, the cost of re-investing and disinvesting the cashflows shown above are likely to be an additional 15bps a year using conventional equity markets. A performance drag of 25bps is material and will be discriminated against by an optimiser.
Is there a way around this impasse? The answer is a twofold yes – one in effective benchmark design, and one in active cash-flow management.
The purpose of a benchmark hedge ratio greater than zero is to reduce or eliminate in the benchmark avoidable risk at lowest possible cost. Since we have established that the high cash flows associated with a 100% hedged benchmark are costly, can we still achieve significant risk reduction with lower benchmark hedge ratios? Figure 3 shows the answer.
At allocations near the current UK average (30%), over half of the currency risk can be eliminated by a 25% hedge ratio. At international allocations between 30% and 65%, some 90% of the currency risk can be eliminated by a 50% hedge ratio. Note that the optimum benchmark hedge ratio purely from a risk-reduction perspective, is 75%, not 100%. This reflects the complex interplay of currency and equity markets, and 0the ‘dollar content’ embedded in the UK and European markets.
Assuming that the choice is a 50% hedge ratio, this reduces all the costs by 50%, bringing them down to nearer 10bps again, even without further management effort.
Costs can be brought down much further by specialist currency managers. We have developed the hedge design expertise to reduce the costs of a passive hedge to under 5bps of the international equity allocation, including fees. We also have an active management process that will significantly reduce negative cash flows, and add 1% expected value added.
From a pure risk-reduction perspective, the argument for eliminating currency exposure in the international benchmark is compelling. However, passive currency hedging, unlike other ‘asset classes’, is a dynamic process, which incurs transactions costs and creates cashflows.
Effective benchmark decision-making requires the fund and their consultant to understand not just the theory of currency risk reduction in benchmark-setting, but also the practical and cost implications of implementing a hedged benchmark.
Specialist currency overlay managers are in many ways the best placed to provide all the information both funds and consultants need to make this decision effectively.
Neil Record is chairman of Record Treasury Management in Windsor