The environment since 2008 makes the case for dynamic hedging. Martin Steward asks which strategies and which instruments are best-suited to the risk-management process.
Most institutional investors get the argument for currency hedging, especially as the de-risking trend of the past 10-15 years has made unwanted currency exposure a greater and more visible part of total portfolio risk, at the margins. “Our response to that has been to advise clients to fully hedge their currency exposures as far as practically possible,” says Sorca Kelly, managing director in consulting and advisory services at Russell Investments.
This wasn’t always the way hedging was done. Many investors experimented with ‘dynamic’ currency hedging in the years running up to the financial crisis: instead of hedging completely or setting a static hedge ratio of, say, 50% to remove some currency risk while minimising the regret of missing out on FX rallies, that hedge ratio could vary according to a manager’s alpha-seeking models or views.
There was always some resistance to that idea in the UK and the rest of Europe, and when a number of currency managers ran into difficulty during the financial crisis, dynamic hedging was dealt a blow – leaving us with the fully-hedged solution described by Kelly.
“If we want currency exposure, it should be entirely on our own terms, and that process should be kept separate as a funded mandate with a benchmark that isn’t related to the incidental exposures you have,” she says.
Gavin Francis, a managing director in client currency portfolio management at Insight Investment, concurs that European currency-hedging tends to be done passively. But while he fully supports the idea of separating risk-reducing from risk-seeking activities, he argues that “dynamically” controlling risk in a hedge is “quite distinct” from trying to extract added value from moves in FX markets.
At State Street Global Advisors, head of currency management Colin Crownover similarly acknowledges that the big move of the past 10 years has been from active hedging to separating hedging and alpha-generating processes.
“But now people are returning to the passive part of that and asking if there are smarter ways of doing that, having covered active somewhere else,” he adds. “It’s still about asking the question, ‘Is there a smarter way of reducing risk?’”
The key is to recognise that hedges introduce new risks associated with the cash flows required to fund them when you are profiting from your assets’ FX exposures. In this respect, 2008 makes rather than breaks the case for active hedging: fully-hedged euro or sterling investors faced huge cash calls as the flight to safety pushed the US dollar through the roof at the same time as their equities crashed – equities that they may have been forced to sell to raise that cash.
“On the one hand you want to eliminate large currency-related losses but, on the other, you want to avoid large negative cash flows,” says Francis. “These are conflicting objectives.”
Crownover says that investors are taking one of two routes in this return to dynamic risk management. Some are trying to build optionality into their hedging programme – ways to capture big appreciations in their base currencies while opting out of scrambling for cash when there are big moves the other way. Others – like SSgA – introduce some kind of systematic ‘tilt’ to their hedging strategy.
SSgA tilts towards what it counts as under-valued currencies, hedging out those that are over-valued. This might sound like the sort of active management that investors have rejected. But Crownover makes the case for its risk reduction properties with two points. First, the value tilt is entirely systematic, rather than based on forecasting. Second, the primary objective of the value tilt is not to add value but to provide tail-risk protection.
“You tend to have the biggest drawdowns when you are invested in over-valued currencies,” he explains.
Nonetheless, a value tilt can introduce meaningful risk: it is notoriously difficult to judge over what time horizon re-pricing of a currency to its supposed fair value will occur, or how far the currency can continue to diverge from that fair value.
“Fundamentals can say whatever they want, at the end of the day the risk shows up in the price action of currencies,” says Maria Heiden, investment adviser in currencies at Berenberg, which prefers the optionality approach in its clients’ active hedging programmes. “Just look at 2008 – US fundamentals looked terrible, but that didn’t stop the US dollar from appreciating strongly.”
Heiden observes that short-term basis risk matters even to long-term investors, not only because they face regular reporting requirements, but because the biggest currency moves tend to happen very suddenly.
“An un-hedged US dollar investor would have seen two years’ worth of returns destroyed by a couple of months’ movement in FX markets in 2008,” she points out. “The move in the Japanese yen this year has been similar.”
At Insight, which also favours the optionality route, Francis makes much the same point. “The typical variables like interest rates, inflation and current account balances that go into a macro model don’t change fast enough to explain all of the volatility in currency markets,” he says. “And, of course, our clients have hired us to manage that shorter-term volatility as part of the hedging mandate.”
So investors arrive at a recognition that the two key problems they want to solve with hedging are: first, the avoidance of large losses from foreign currency weakening (or domestic currency strengthening), which often happen very suddenly over short horizons; and second, a limit to, or at least stability in, the cash-flow demands that result when the opposite happens – or when nothing much happens at all.
For the first problem, as we have discussed, optionality seems like a good solution – it is short-term oriented and systematic. For the second, FX options seem, on the face of it, like a good instrument.
An FX option offers exposure to the strengthening of your base currency without saddling you with the exposure to the weakening of your base currency that results in those unpredictable cash flows. In exchange, you pay an up-front premium determined by the implied volatility of the exchange rate.
So the trade-off is ultimately between certainty around your cash flows and the cost of that certainty. Is it worth it?
“Right now, a three-month GBP/USD at-the-money call is about 1.4% and a 12-month call is about 3%,” says Kelly at Russell. “With forwards you can get most hedges done for about 10 basis points.”
An option priced at fair value is one where the accumulated premia offset the accumulated pay-off, argues James Wood-Collins, CEO at Record Currency Management.
“The fat tail you’re concerned about hedging is a pronounced weakening of your exposure currency or strengthening of your base currency,” he explains. “If you buy an option that’s a call on your base currency that will become much more valuable. But the premium you’ve paid will probably have taken into account that very risk. Over the long term, when the currency exposure mean-reverts and nets-out to zero overall effect, by definition, a rolling option programme is going to cost you to the extent of the premiums you’ve paid to keep it in place. I think that performance drag explains why institutions don’t systematically hold options positions.”
Neil Staines, head of trading at the ECU Group, which offers both return-seeking and currency-hedging programmes, cautions against following this efficient-markets critique unquestioningly. He points out that the longer the tenor of an FX option is, the more its valuation is driven by the interest-rate differentials in the currency pair, rather than the simple volatility of the exchange rate.
“The value is not just a function of the options market pricing in where spot is going to be, it also takes account of interest rates, volatility, skew, risk-reversal and the current market price,” he says. “So the cost of an option can depend on, for example, whether the interest rate differential is to your advantage or not. If it is, the option can be incredibly cheap. There are definitely situations where you can lock in better gains and cheaper optionality by using interest-rate differentials or the different volatilities in one tenor versus another – but assessing that is, by definition, quite a complex process.”
However, even if options structures of varying complexity can reduce the performance drag associated with cash-flow certainty to within acceptable limits, there remains a question even against that assumption of cash-flow certainty. The argument in favour of options is that once they are in place you know your costs. But do you know what the cost will be to roll that position into a new contract in three or six or 12 months’ time, to maintain your hedge?
“When the contract expires you face an uncertain and uncontrolled cost associated with implementing a new one,” says Francis at Insight. “The client is completely exposed to whatever volatility there might be in the market – and, of course, the time when they need protection the most is precisely the time when it’s most expensive.”
In addition, options can prove pretty inflexible should you need to change your hedge. An investor faces a choice between writing a corresponding option or entering into a negotiation with its counterparty to amend or close positions – probably at a punitive cost.
So FX options themselves may not be the optimal way to implement optionality in an FX hedging programme.
A number of managers choose, instead, to replicate option pay-offs using trend-following currency-forward strategies – increasing the hedge ratio once the base currency has entered an appreciating trend, and decreasing back towards the hedge ratio benchmark once it has entered a depreciating trend.
Record and Insight hedge in this way, and a momentum strategy underlies the programme Berenberg runs for its clients, too. It has five trend models, working to different time horizons, which each control 20% of the potential total hedge. In general, the programme would move a client to fully-hedge or fully-unhedged, given a currency move equal to about half of that currency’s volatility. For example, because USD/GBP has a volatility of about 10%, a roughly 3-5% movement in either direction would turn a USD/GBP hedge fully on or fully off in five stages as the trend develops.
Record splits the total potential hedge into 12 rather than five, and its 12 trigger points for hedging or de-hedging are set by implementing a new forward position each month at the prevailing spot rate. This means that, at any one time, there are 12 different exchange rates acting as triggers for each currency pair the client wants to hedge. Each position that remains above today’s spot rate will be open and each position below it will be closed, with the result that one-twelfth of the fully-hedged position will be switched on or off as the spot price moves through each trigger level. In other words, Record’s programme’s sensitivity to the strength of trends is path-dependent on the currency pair’s movements over the preceding 12 months, whereas Berenberg’s programme’s sensitivity is determined by the long-term average volatility of the currency pair.
This sensitivity to trend strength is vital because it determines the costs of the programme – the equivalent to the up-front premium that you pay for an option. With a forwards-based trend-following programme, you achieve a similar result as with an option, in that you get protection against big cash-flow demands from significant appreciation of your base currency. But, because the costs depend upon how your programme responds to realised currency volatility, your cash flows will always remain uncertain. Put simply, a long period of range-trading in a currency pair which is choppy enough to switch triggers on or off without establishing a decent trend will result in higher cost for little hedging benefit.
There are ways of managing these costs that are certainly easier than building complex offsetting option structures – Record will over-ride new trigger points if they would be placed to close to existing ones, for example. And in the last resort, any trend-following programme’s losses can be capped by simply closing positions to return to the benchmark hedge ratio until markets trend again. But advocates of systematic trend-following say that the costs will tend, in any case, to be lower than those associated with options.
“When the base currency is depreciating, you can compare the size of the option premium you have paid with the costs you incur from the trend-following model,” says Maria Heiden, investment adviser in currencies at Berenberg. “Only real-money performance can truly settle the debate, because it depends on the pattern of realised volatility over the term of the contracts, but there are a couple of fundamental reasons why options should tend to be more expensive.”
The first of these is the costs associated with rolling into new option contracts. The second is that a trend-following strategy, by definition, responds to, and therefore its trading costs are defined by, realised volatility, whereas options are priced according to implied volatility – and realised volatility tends to be lower than implied.
Wood-Collins adds another explanation. Most currency pairs do exhibit momentum, and anything that exhibits momentum also exhibits higher volatility over longer time horizons. Because the costs of implementing a trend-following programme are incurred over short time horizons (in Record’s case, hedges are switched on or off according to daily movements in spot prices), and profits are harvested over longer time horizons (Record uses one-year forwards), the positions exploit this natural arbitrage between long and short-term volatility.
“We have 30 years of evidence that the premium paid for a trend-following strategy is less than the pay-off,” says Wood-Collins. “We have found that it adds about 100 basis points of value over a full currency cycle.”
Some remain sceptical of the optionality-based approach to hedging in general. Crownover says that SSgA does some option-replication in its active strategies, but does not do so in any passive or dynamic hedging programmes.
“You can cheapen the delivery of an option-like pay-off, but that doesn’t change the fact that there isn’t a free lunch,” he insists. “Transaction costs will not be as high as a purchased option but they will still be higher than hedging with a passive ratio in forwards.”
But that doesn’t seem like the appropriate benchmark for the costs of a trend-following programme. In fact, it can only be so if one genuinely feels there is no benefit gained from building optionality into a currency hedge – a position that is difficult to sustain.
“It’s true that in sideways markets the trend-following process comes with a cost associated with opening and closing hedges,” as Heiden puts it. “But this is not the risk. When it comes to risk management, pure performance is not the aim of the strategy. The aim is to make sure the programme is triggered when there are big, significant trends in the risks you are exposed to.”
Those risks are twofold: missing out on base currency appreciation; and facing major cash flows from your hedges against base currency depreciation. Passive hedging leaves you fully exposed to one or the other, or sub-optimally exposed to both. Despite the current lack of interest, there is still a place for dynamic hedging in an investors’ consideration of how to manage foreign-exchange risk.