How should an investor with a medium-term horizon deal with currency risk? That is a crucial question given that, unlike fixed-income securities and equities, foreign currency holdings do not earn interest or pay a dividend.
In professional jargon, incurring currency risk is referred to as a relative investment position: you go long a foreign currency and you short your own currency. As a rule, it is best not to take relative positions on a structural basis. In developed countries, in particular, economies are too similar to expect a structural reward for exposure to currency risk.
Relative investments are best managed actively. With currencies, timing the market can be additionally difficult due to the significant “monetary dimension” in the form of central bank interest-rate policy and guidance.
Institutional investors often have a long-term investment horizon and are thus inclined to hedge the currency risk in full. From a regulatory perspective, full hedging also looks attractive. Within the nFTK (the Dutch new financial assessment framework) and under Solvency II rules for pension funds and insurers, fully hedged positions qualify for the lowest buffer requirements (all else being equal).
Nevertheless, in our opinion, investors who are prepared to review their currency policy regularly (i.e. every three or six months) are short-changing themselves if they automatically opt for full hedging. A policy review typically includes a) determining whether the current exchange rate is too high or too low given the current performance of the underlying economies, and b) estimating the expected inflation and growth rates for the coming years.
Such a currency valuation model helps to answer the question as to whether a currency is relatively expensive or cheap and to decide whether, and if so to what extent, hedging makes sense. It forms part of a framework for currency hedging which also includes factors such as expected returns, costs and risks associated with the selected policy. We can summarise this as follows:
- Is a currency pair overvalued or undervalued in the medium term?
- Are the costs of currency hedging high or low?
- When does currency hedging have a risk-mitigating effect and when not?
Valuation: the case for relative purchasing power parity
As we have mentioned, economic growth and inflation influence the value of a currency. Purchasing Power Parity (PPP) can be a good starting point for determining the correct value. In its simplest form, PPP means that a basket of goods should be just as expensive to buy in euros in Europe as it is at the current dollar exchange rate in the US. If it costs more in dollars, the USD is too expensive against the EUR based on the PPP theory.
The underlying idea is that price differences are cancelled out by arbitrage in international trade. Economists agree that PPP in its simplest form does not work. Transport costs and tax levies distort the picture. Moreover, the basket of goods is not identical in every currency area. For this reason, we apply a relative PPP: in the long term, the change in nominal exchange rates (of developed economies) keeps step with the difference in inflation.
For instance, the nominal EUR/USD rate was USD 1.17 in Q2 2018. As average inflation (in US dollars) was higher in the US than in the eurozone, the real euro exchange rate worked out at USD 1.13.Exhibit 1: EUR/USD real exchange rate and PPP
Source: Bloomberg, IMF, BNP Paribas Asset Management, as at 30/06/2018
In our framework, we assume that the real exchange rate converges to the long-term average in the medium term (5-7 years). Based on the rates in Q2 2018, this implies that the US dollar should decline by 1% a year on average. Viewed in this light, the US dollar was overvalued. This could be a reason to hedge the currency risk.
Clearly, we must also take account of the uncertainty surrounding the estimates. This is shown in the chart using the dotted red lines. If the real exchange rate is outside the band, this implies an increased conviction that the currency is undervalued or overvalued.
Pricing in the costs of currency hedging
Currency hedging comes with costs. We are not referring to the direct transaction costs, but to the costs that are included in the prices at which currency forward contracts are traded.
At the end of June 2018, the costs of hedging the US dollar amounted to about 2.7% on an annual basis. At the time of writing , these had actually increased to 3.2% after the US Federal Reserve raised its key policy rate. That is historically high: over the past 15 years, hedging costs averaged about 0.5%. Higher costs are a reason to hedge less currency risk.
Does currency hedging always reduce the risk?
Currency hedging does not always lead to a lower risk. From 2008 to 2015, the hedged MSCI World index showed greater volatility than the unhedged index. From a volatility perspective, it would have been better not to hedge the currency risk during this period.
The essential parameter for determining the optimal hedge appears to be the correlation between the local return on the investment and the rate of the relevant currency (versus the euro). If this correlation is below zero, partial hedging is optimal. If it is strongly negative, not hedging is optimal.
To determine the optimal hedge percentage for a portfolio, all correlations within the portfolio must be taken into account. The rule is: the greater the number of negative correlations between the investments and relevant currency pairs, the more often 100% hedging is not optimal for reducing portfolio volatility.
Return versus risk for each asset class
When reviewing the medium-term policy, we advise looking at the expected return without hedging, the hedging costs and the potential impact on portfolio risk.
It should also be remembered that not hedging can offer more diversification advantages for equity investments than investments in government bonds. In times of financial market tension, certain currencies can serve as a safe haven, so that there are great diversification gains to be had from equities precisely when you need them.
 written: September 2018here