The ability to add overseas assets to an investment portfolio can improve diversity and return opportunities, and today it is an integral part of most managers’ toolbox.
But what about the currency? Should it be part and parcel of the initial asset allocation decision? Or can portfolio managers simply ignore it, make decisions based on asset performance, and deal with the currency at a later date? We argue here that currency is an important part of asset allocation, and should be considered at all stages of the process
Let us first consider those managers who believe that currency does not matter. How will their allocation process work, and how will currencies affect it? A simple test is to proxy their asset allocation decision by a straightforward portfolio optimisation process (Markowitz). As an example we took a euro-based investor who can chose to divide his portfolio between the DAX and the Dow Jones Industrial Average (DJA). Our data series began in 1990. At the end of each year, we calculated the percentage that should be allocated to the DJA to give the maximum information ratio for the past 12 months. This can then be used as a proxy for the allocation for the following year.
We did this first without the currency component, and then with the currency. If currencies truly do not matter, we would not expect to see much difference between the two. However, as is very clear below, the two processes can yield very different asset allocations. The most recent year is an unequivocal example.
The next case which we consider is that of the manager who decides that he will ignore and hedge away all currency risk with a passive currency management programme. Initially one might think that under these circumstances one could indeed forget about currency when constructing a portfolio. However, once again, it’s not that simple.
Forward hedges will in theory hedge away currency risk, but as the value of an asset changes, the hedges will need to be adjusted on a regular basis. This rebalancing strategy needs to be carefully decided upon, with a good look at the trade-off between tracking error and trading costs. Additionally, for the individual investor, large interest rate differentials may be a source of additional revenue or prohibitive cost as the forward points accumulate. Overall the effect must net out over the global investment community, but this is no comfort to those facing the prospect of prohibitive hedging costs.
But the most significant disadvantage to forward hedging is that at the end of the hedging period, the hedge may not have made money. Indeed, it may have lost money quite seriously. One might initially think that this is hardly a problem as the underlying asset will have gained an almost identical amount in value, and thus one is neutral overall. But a cash flow effect from currency hedging appears, as losing hedges require funding. This cash needs to come from somewhere, and the obvious source of funding is the underlying asset itself. The manager will need to liquidate some asset to cover the cost of the hedge if cashflows are negative and reinvest the surplus if they are positive.
As frequent selling and buying of asset can be expensive and impact the performance of the fund, it would intuitively makes sense for some part of the portfolio to be maintained as a liquid asset or indeed futures. Below we show the distribution of quarterly cashflows that would have occurred for a ‘standard portfolio’ of 50% DJI, 20% Nikkei and 30% FTSE for a EUR-based investor since 1990.
As can be seen, there is about a 20% chance of having to find an amount of 8% of face or more. This methodology could be used to decide how much of the portfolio should be maintained as relatively liquid assets or futures to provide a ‘cheap’ and liquid source of funding for these cash flow imbalances.
It can be seen that there is a lot to consider when designing a passive currency management programme. The manager needs to select a target hedge ratio suitable for his portfolio and mandates. He must decide on the tenor of any hedges – quarterly or monthly is normal – and consider the associated cashflow risks. He must then consider how he will rebalance the hedges between the roll dates – will it be on a monthly or weekly basis, or will he rebalance when the underlying portfolio moves by a specified amount? Once all these questions are answered, he will be able to put in place a well-judged passive currency management strategy, which should yield much smoother, more predictable returns than would leaving the underlying assets unhedged.
A third possibility exists, which we have not considered. This is that the manager is making a decision based on a view both on the currency and the likely asset moves. For this kind of integrated asset allocation, he needs to have experience in both areas and be comfortable making this judgement. Again the currency decision becomes a key driver in his final decision. This is because he will make decisions based on his currency view as well as his asset view. However, this might lead to high performing assets being omitted just because the currency is expected to fall. In this case it would again make sense to separate the asset and currency decision. This means that asset are judged and apportioned on a hedged basis as described above and then the currency decisions are taken separately in order to attempt to generate excess returns. An added advantage of a successful programme is that it will reduce the negative impact of cashflows.
The asset manager who takes on this methodology is doing two jobs, effectively being his own currency overlay manager! There is a growing body of evidence which suggests that this method both reduce risks and add returns, but it is unwise to underestimate the magnitude of the task.
So, the question of whether currency needs to be considered at the early stages of asset allocation can be answered in the affirmative. It does matter, it should not be ignored, and even if the manager intends to hedge the currency risk into oblivion, he needs to think about funding the cash flows arsing from such a strategy and what hedging methodologies are optimal to achieve this. It is not logical to consider currency and underlying asset in one basket – as soon as one develops strong views on both, a currency strategy becomes necessary to achieve them! Currency should be integral to all parts of an asset allocation process.
Jessica James and Henrik Pedersen are vice presidents of the CitiFX Risk Advisory Group in London. The views expressed are the authors’ own and are not necessarily shared by Citigroup or any of its affiliates