Keeping track of currency risk
Whenever pension funds switch from a domestic to a global benchmark, either in fixed income or equities, they take on currency risk that must be hedged.
This is the main reason why Railpen Investments, the investment arm of the UK Railways Pension Trustee Co, moved into both passive and active currency management.
Railpen manages pension assets of £13.5bn (€20bn) in eight pooled funds for some 100 railway employers in the UK. In 2001 it changed its bond benchmark when it moved from UK to global bonds. This meant creating a bond portfolio with 100% passive currency hedging.
It also meant moving in a small way into active currency management. Brendan Reville, head of manager monitoring and research at Railpen Investments, speaking at a UK currency conference organised by ABN Amro earlier this year, explained the process.
“We started doing active currency management because our bond managers had proportions of their risk budget that they were allowed to spend on active currency management; probably about 25% maximum in each case,” he says.
The following year Railpen also changed its equity benchmark. With assets under management of £8.5bn, Railpen’s equity fund is by far the largest of its funds. The fund had used a peer group benchmark for some time. As a result, two-thirds of equity assets were invested in UK equities.
The change in benchmarks led to an immediate reduction of Railpen’s UK equity allocation to 50% (the current allocation is about 45%). To deal with the additional currency risk, Railpen put in place a 50% hedge at the end of 2002.
Reville, who joined Railpen at the start of 2003, moved Railpen into active currency management to make better use of risk in the pooled funds.
“Active currency management would increase the risk in the portfolio only marginally, because of the low correlation between equity and currency. But it would increase the return and information ratio significantly,” he says.
Selecting managers was simpler than expected, and did not involve the use of consultants, at least initially. “It is a market that’s got a remarkably small number of participants, so it’s often easier to do it yourself.”
If pension funds do take the DIY route, Reville says, they should use the internet to check what pension plans in other areas are doing. “You can see who’s gaining assets and who’s losing assets.”
When Railpen started its manager search at the beginning of 2003 it selected 19 managers to screen – coincidentally but comfortingly the same number of managers used in the Russell Mellon CAPS survey earlier this year. It short-listed 11 managers for interview and finally chose two.
Selecting managers was initially a matter of choosing between the three basic currency management styles – technical, fundamental and judgemental. However, this was not as simple as it appeared, says Reville. “The basic styles are not well defined. We found out quite rapidly that there are a small number of purist managers and a growing number of managers who were adopting a wider range of techniques and putting those in their active currency programmes.”
Technical managers are worth serious consideration, he says: “Don’t throw out the technical managers simply because what they do seems like mumbo jumbo. The style may not have any philosophical basis but it will make money.”
Fundamental managers are more difficult to evaluate: “In some ways it’s actually harder to demonstrate that you can value currencies,” he says.
Judgemental managers are almost impossible to evaluate, he suggests. “You will find people who have been making money out of currency for 30 years and they can’t tell you how they do it. In the end we eliminated judgemental managers from our search simply because we felt we couldn’t persuade the trustee that they could appoint them. But I’m sure there’s value there, and it may be something we add at a later stage.”
As with asset management generally, there is a health warning. Past performance is no guide to future returns, Reville says. “Everyone in currency outperforms. There is an inefficiency in the market that everybody can make money out of. So don’t place too much reliance on the fact that the manager you’re looking at has outperformed. Equally the fact that they haven’t got enormously high returns doesn’t mean that you shouldn’t appoint them. You need
to consider things other than pure performance.
“What made up our minds was more to do with people than philosophy. Most of the teams are actually relatively small and have a few key individuals with quite a lot of experience.
“We also thought a dedicated research resource was important. There are a number of managers with no resources behind them and they run the risk their information ratio will get eroded. Currency is an area where collecting and interpreting data is of fundamental importance.”
Choosing managers is not simply a matter of choosing the best but choosing managers that will work well together, he says. “We wanted as many unrelated sources of alpha as possible.”
Railpen called in consultants to validate its initial choice of managers. “We had six managers that we knew we liked and we got a ‘buy list’ from the consultant. This included one manager that we already had in our bond portfolio but who we hadn’t considered separately for currency overlay.”
Railpen also used other tactics to ensure uncorrelated sources of alpha, in particular by hiring a manager
who works in emerging markets currencies.
The most important task, however, was to reassure the UK Railways Pension Trustee Co that active currency management was a suitable use of Railpen’s risk budget, says Reville. “We spent roughly six months talking to the trustee about active management. We needed to lead them through concepts. We got existing managers, who were managing currency for them, to talk to them and tell them what they were doing. And we got the new managers in to explain their processes.
“This was probably the one thing that the trustees found most useful, because it persuaded them that
other people were treating this as a sensible asset class and that it wasn’t irrational to try to make money out of currency.”
The trustees were worried about adding to Railpen’s large number of existing managers – the funds employ 35 managers. “From one point of view they just wanted to control the numbers. So the fact that we’ve ended up with two is in part a pragmatic decision not to expand the programme straightaway to a larger number. There is still scope for doing that later though,” he says.
Cash flows were also a concern for the trustees. “They wanted to know what happens if you have a serious event and you lose a couple of million on currencies.”
Railpen modelled the cash flows in different stress scenarios, to assess the impact of likely drawdowns from the currency, passive or active, separately or together. It has set up a cash pool within the equity pool to deal with all the drawdowns from the currency managers.
“We’ve combined our active and passive mandates. The best reason for that is that you get the passive for free. The two managers each handle half the passive mandate, and we compare them on a daily basis. They should both be generating exactly the same numbers.”
Perhaps the most worrying feature for the trustees, Reville says, was the volatility of currency management. “We have changes in profit or loss on the actual currency programme of up to £5m a day,” he says. “Now you
can spend all day looking at those numbers or you can just walk away from them. It’s probably best to do the latter.
“The reality is that currency management does create large positive and negative cash flows. The most difficult issue with trustees is getting over to them that, on the passive side, a really big cashflow isn’t actually a loss it’s just a benchmark implementation and that on the active side the amount of risk that they’re taking is actually relatively trivial.”
In the end, Reville and his team gained approval for a risk budget for currency that that was half the size they had originally proposed. “The trustees agreed there might be money in it but they wanted to test it first,” he says. “This is quite typical of pension schemes in the UK. They will put a small amount of risk into currency first of all to see whether it works.”
If it does work, they might be given the whole risk budget they requested. “We’ve been told we can go back
to them at the end of this year with a one-year performance history and
talk to them again. So there’s the potential for doubling the risk budget,” he says.
Railpen’s active currency management programme started well, generating £11m within the first two months. Reville can only hope that the figures look as good after 12 months.