FX debate: Is currency the perfect asset class?
Euromoney Limited, Registered in England & Wales, Company number 15236090
4 Bouverie Street, London, EC4Y 8AX
Copyright © Euromoney Limited 2024
Accessibility | Terms of Use | Privacy Policy | Modern Slavery Statement

FX debate: Is currency the perfect asset class?

Participants Katie Martin, Dow Jones Let me start by setting the scene a bit. The most recent study from the Bank for International Settlements found a one-third rise in daily trading volumes to $1.9 trillion a day. Financial accounts and hedge funds formed a big part of that boom, with their share of the volume rising to about a third. That segment is understated too as prime brokerage relationships allow clients to execute under their banks' name. So it appears that the battle to convince investors that currencies do provide a meaningful source of returns has been won. Monica, why has this asset class sprung to life as it has?

Monica Fan, RBC Capital Markets If you look at the performance of traditional asset classes over the past four years, they've produced poor returns. Naturally people are looking for positive absolute returns and they can find those in the FX markets. Increasing investor education has encouraged people to move into so-called non-traditional asset classes, and the willingness of consultants to advise their clients to invest in absolute return funds has also played an important role.

Harriet Baldwin, JP Morgan Fleming Katie called FX an asset class. I think it's more accurate to say that FX is a medium of exchange between different asset classes. You can create an asset class out of FX returns, but it's not an asset class in the same way as fixed income or equities because currencies have a unique characteristic: they don't have a theoretical expected return, they're just a way in which we denominate other asset classes. That said, the growth we've seen is a function of two things. First, increased global diversification: when you invest internationally you get currencies as a side effect and you have to think about what you're going to do with them. Second, as Monica points out, the expected returns on traditional asset classes have been much lower. So any kind of alpha – excess return – that you can get is valuable and FX does give you opportunities to get excess return.

Michael Shilling, Pareto Partners I think there has been a change in investors' view of what is the active risk in their portfolios. Larger-cap equities, for example, have been disappointing and the active management of them has consumed investors' risk budget without adding any additional value. So investors have started looking at active sources of return separately from the underlying asset classes and been splitting out their allocation of risk. This has led to the debate on the difference between beta risk and alpha risk. Currency has come to the fore because active currency management is one of the pure sources of alpha risk – there's no beta associated with it – so it's pure transportable alpha. That has attracted people's attention.

Bob Jolly, Gartmore Remember too that people have found through the years that currency managers do actually generate this alpha and that return is typically uncorrelated with other asset classes. As a consistent diversifier FX generates more return for a given unit of risk, which is the name of the game.

Henir Foch, BNP Paribas Don't you think there is too much money chasing too few opportunities? Returns are declining and access to FX liquidity has drawn many more people into the markets removing opportunities?

MS, Pareto Partners Well we are talking here about scaleable alpha. When you look at all the money that's pouring into hedge funds you start to expect the return to come down because opportunities are going to get arbitraged away. But with currencies, if more money goes in it helps, it doesn't dilute the alpha that's available.

HB, JP Morgan Fleming I would pick you up on that Michael. Hedge funds like to follow trend-following strategies. For that inefficiency to exist, insufficient people must notice it in time. If more participants follow this model, then does that destroy the strategy or does it simply exacerbate the trends?

MS, Pareto Partners If everybody followed the same strategy, that would be true, but they don't.

HB, JP Morgan Fleming I think this comes back to Bob's point that currency managers seem to be systematic winners. It's hard to find a currency overlay manager who hasn't made money. For that to happen, there have to be systematic losers and in the FX markets they are central banks and also the global equity managers moving from one stock to another across a border and ignoring currency. These two sources of systematic loss will, I think, continue.

BJ, Gartmore All the evidence is that in the medium and long term people make money out of trading currency, whether they be hedge funds, CTAs or currency overlay managers. FX is one of the more consistent alpha sources and it's one of the more consistent diversifiers. So with FX in a portfolio construction framework you generate more return for a given unit of risk, which is the name of the game. To argue that this will not continue to be true you have to argue for some fundamental change and a permanent shift to a different regime which involves ranges rather than trends. I can't see that.

MS, Pareto Partners People who believe that the market has changed and opportunities will be fewer and smaller are focusing too hard on the short-term conditions we've seen in the market. But we've had periods like this before. The key to making money, if you have a strategy which makes money out of one or other of the environments, is not to lose too much in the other environments.

HB, JP Morgan Fleming Speaking for the currency overlay industry, there are three main areas currency overlay managers look at. One is trend following; the second is relative interest rates – the carry trade; and third is currency relative value – are we at extreme levels of over- or under-valuation? Those three sources of excess return have very different characteristics. Technicals will give you lots of little losses, but they'll have some very big gains when you're in a big trend and that's why they're favoured by hedge funds. Carry trades will give you lots of little gains on average, but there'll be some very sharp periods of downturn – the big obvious ones were the break-up of the ERM in 1992 and dollar/yen in 1998. And the third strategy tends to be very episodic. Extremes in valuation are rare and tend to overshoot, so you have to go through a period of under-performance before you get the move back. And so that strategy tends to be favoured by much longer term investors.

At present there is a disproportionate amount of short-term trend following money in the market. There you can ask: "Is this a persistent inefficiency?" But I don't think carry trades disappear, because central banks around the world all have different interest rates; and I don't think fundamental mis-valuation disappears.

Dr. Robert Meijer, Meijer & Co Currencies tend to trend in two out of three years, so it is completely normal that there are years in which currency overlay managers make little or no money. But it depends which kind of manager. This year while technical managers were not making money, fundamental and discretionary managers were. That's a perfect argument for diversification.

KM, Dow Jones Is the demand for currency overlay services – as opposed to hedge fund investment – growing?

RM, Meijer & Co Only 10% of international assets are covered by a currency overlay mandate. So there is plenty of scope for growth and demand is growing. One important driver is funding ratios – many pension funds are running deficits and the IFRS regulations force companies to reflect that in their balance sheets. This drives up gearing ratios and means that not only is the pension fund more risk averse, but also the sponsor. As sponsors look at risk management on top of pension fund management that will increase the need for currency overlay.

MS, Pareto Partners So initially this new demand for overlay is about controlling risk in the fund?

RM, Meijer & Co Yes.

BJ, Gartmore I think in overlay the shift is towards alpha, it's not towards risk. It started off being very risk-oriented and then as it became clear that added value could be generated by currency management because it's uncorrelated with the other asset classes, everyone started to jump on the alpha bandwagon.

HB, JP Morgan Fleming That's a fair statement, but I do think it depends on your starting point. If you're a UK pension fund, you've concluded that you have been wrong to wait as long as you have to tackle your FX risk. But a typical US pension fund, with a relatively small proportion of assets outside the US, may be able to move to a portable alpha strategy quite rapidly.

MS, Pareto Partners I think there are two very distinct activities: there's alpha generation, which is a risk-spending activity – creating risk in order to add value; and then there's a risk-controlling activity, which is taking a pre-existing risk and reducing it. And you need both. You can't decide just to have an alpha strategy, because where are you going to get that risk budget from? Or are you just going to increase the overall risk to the fund? The reason people are now more interested in currency as an alpha activity is that they've started to look at their funds from a risk-budgeting perspective. They're looking at areas such as active equity management and saying: "We're spending our risk in the wrong place. Where can we spend it and actually get a return?". Currency is one answer.

The other side of that question is: "How can we reduce risk that we're not being rewarded for?" And currency is one answer here too. So you create risk budget by reducing your pre-existing currency risk. Once you've created that risk budget, you look at where to spend it and currency is one of the candidates – but these are two completely different activities.

Persuading the sceptics

Simon Brady, Euromoney If currency overlay has always made money why is it so difficult to sell it?

HB, JP Morgan Fleming The typical UK plan sponsor would give three reasons for not doing it. One: "Well, sterling always falls anyway"; two: "Currencies are a wash in the long run"; and three: "Well, no-one knows anything about currencies here, I'm an equity person myself." There's also the natural inertia of a committee-driven system, compounded by the fact that this is a very consultant-led market. They only bought into currency overlay fairly recently after they'd carried out long-term studies which proved the alpha exists.

MS, Pareto Partners Also the return from active currency overlay is around 100 basis points. During a period when equity markets are going up 15% to 20% it's not such an easy sell. Once returns from the equity markets fall to 7%, it's a lot easier.

BJ, Gartmore Exactly. The level of returns from conventional asset classes is not expected to be in double digits and so overlay, which before seemed like a lot of hard work for not much incremental return, now looks attractive.

Also until recently the correlation between equities and fixed income was positive and returns were good. Fixed-income yields were falling but equity markets were rising, so the liability side of the balance sheet goes up, but the asset side goes up faster creating big surpluses. Now the correlation is negative – liabilities are rising faster than assets and returns have collapsed. So funds need something to generate return that allocates risk effectively. That means something which diversifies and is consistent and that's currency. So that's why it's being embraced now and it's not just UK pension funds, it's global funds.

Models versus man

KM, Dow Jones Clients can choose between model-based or discretionary funds. Which do better?

MS, Pareto Partners Well I'm not entirely sure of the difference. Even a discretionary style has to work off some sort of model. If you have a consistent style then it's a model; the only difference is the information sources you are using. Are you using fundamental economic data, which, by its nature, changes very slowly and therefore is only valuable for a longer-term perspective on the market? Or are you looking at daily or minute-by-minute ticks?

Nicolas Boitout, Crédit Agricole Asset Management Yes, but the advantage of models is the back-testing behind them.

MS, Pareto Partners But everyone has to do back-testing.

KM, Meijer & Co I have done a study on fundamental managers where some allowed for human intervention and some didn't. And where management were allowed to second-guess the model, their results were sub-optimal.

BJ, Gartmore Well that's interesting, because my experience is quite the opposite.

NB, Crédit Agricole Asset Management Yes, mine is too.

BJ, Gartmore In general a model gives you the benefits of removing emotion and identifying and collecting the key drivers for the asset class you're trying to forecast. But you have to take some things out otherwise you get an overfitted model. Now, if the manager of the strategy is aware of the strengths and weaknesses of the model and of the environments in which it does well and in which it does badly, then you will find that they can knock off the rough edges. Certainly the experience at Gartmore is that the information ratio is higher where we use some discretion than it is just following a box.

NB, Crédit Agricole Asset Management I completely agree with that. I am managing a trading system and we are not automated, we are systematic but we left some scope for discretionary decisions and I see that I can add alpha in that way.

SB, Euromoney But what are you adding that you couldn't programme into the model?

NB, Crédit Agricole Asset Management There are macro structural effects in the market that you can't integrate in the trading system.

MF, RBC Capital Markets Doesn't this raise a problem though that in the same way that a lot of active equity managers are accused of being closet benchmarkers, so the same accusation could be made that some model-driven currency managers are really discretionary managers?

Jeremy Armitage, State Street Is there something wrong with that?

MF, RBC Capital Markets There is if you override the model with what you call "discretion" – that is, unspecified variables that are articulated mentally in "discretion" but not explicitly articulated mathematically in the models. Then clients can end up buying a product that is different to the one they've been sold.

MS, Pareto Partners I think if you are trading a fund or a pure alpha strategy, then discretion is probably a valuable approach and you live or die by your performance. If you're managing an exposure, as in traditional currency overlay, then I think clients want a better understanding of what drives decisions. First, because you need to explain what went wrong if your numbers are down and second because the clients' question is: "To what extent am I dependent on an individual who might leave?" Overlay is a long-term strategy and clients may want to know that they are relying not on an individual but on a strategy that is independent of that individual.

BJ, Gartmore The key is the continuity of the people involved.

MS, Pareto Partners But they have no control over the continuity of the individuals, even if they were to identify them.

BJ, Gartmore If the people that have been managing currency overlay over a long period change and the client believes that there is a lot of discretion involved, they will review their use of that firm.

JA, State Street How do you scale your business if you have a significant amount of discretion involved? It's fine if you're running one trading portfolio, but when you're running multiple accounts, different base currencies, multiple constraints, how do you maintain a consistent level of discretionary input across that whole business relative to a pure model-based approach.

BJ, Gartmore You have rules of engagement with your model that you adhere to. If your models say euro-dollar's going to go down and you actually think that it's just the fall-back before it moves up again, then you use whatever your pre-agreed discretion is either to put on a stop or reduce the size of the trade or whatever. And if you're right and the rate just dips and then carries on going, then you note it and are allowed to keep that level of discretionary input.

JA, State Street But you must be putting that discretion on earlier than the trade, no, Bob? Because you must have, I don't know, 20, 30 trades.

BJ, Gartmore It varies by situation, but if you're asking: "Does one client get favoured over others?" the answer is no, because you're doing it in a systematic way. In theory you could turn it into a model. But one of the problems about a black box is that you are building something on the basis of history. If there is some kind of change in the behavioural characteristics of the market, then there is a risk that you will suffer a period of poor performance.

MS, Pareto Partners Versus discretionary managers that live forever!

BJ, Gartmore Valid point, but some clients are wary of black boxes.

MS, Pareto Partners I accept that, but I would've thought that for a client the most opaque black box of all is the person who just says: "I know what to do. Trust me".

BJ, Gartmore But you don't win money by saying: "I know what to do. Trust me".

MS, Pareto Partners No, you've got a track record.

BJ, Gartmore In general, even if you have a track record you do not attract capital unless you can clearly explain your investment philosophy and process. At the end of the day the objective is to make money and the only significant change has been the evaluation of how effectively managers make money. So there is much more attention now paid to Sharpe ratios and information ratios [alpha divided by tracking error] than to total return because the information ratio is leverageable, whereas total return may not be. Efficient creation of alpha is the objective and we believe that you can generate a more efficient alpha by using an intelligent model implemented intelligently by people with an element of discretion.

MS, Pareto Partners Another complication is that in a mix of model and discretionary it is at the most critical times that the emphasis is on what the human does. When you get into periods of great volatility and uncertainty, isn't that when the temptation is greatest to override the model and that potentially has the most impact?

BJ, Gartmore Well it all depends whether you're making money or not really, doesn't it? I think what humans are very good at doing – and this is my experience – is reducing the depth of draw-down. I think that is where discretion is most effective.

JA, State Street There are two levels of discretion. One is the trading decision – the timing of the trade – where there's information in the market the model can't capture that a trader will be aware of. The other is changing the investment decision discretionarily – that is overriding a factor or re-weighting the factors. That concerns me more, because that's a much more complicated decision.

MS, Pareto Partners Where should the combination of systematic and discretionary occur? Should it occur at the manager level or should the client say: "What I need is a discretionary-based manager and a model-based manager because that will give me two diverse sources of alpha, because they're accessing different sources of information? If I rely on the manager to combine both, do I get the same type of diversity?".

BJ, Gartmore One of the currency consultants said that what really surprises them is that although managers with forecast models all use very similar [data] inputs, the excess returns they generate are reasonably uncorrelated. So, even in systematic models where a client should be able to say that returns will be similar, they're not.

SB, Euromoney So to get back to the original question: which approach has done better?

NB, Crédit Agricole Asset Management Obviously when you look at the back-testing or the Parker Index, you see that systematic trading has outperformed discretionary trading. And after this year, where pure systematic trading has been hurt, we can expect a shift to a mixture between the systematic and discretionary.

The art of measuring returns

KM, Dow Jones So how should clients measure the overlay service? There doesn't seem to be any consensus.

MS, Pareto Partners Investors are changing their emphasis and looking at the sources of alpha separately from asset classes. That means they now have a series of long/short strategies which don't require the investment of capital. An alpha overlay doesn't require the investment of capital, it requires the allocation of contingent capital. So the question now is not how much is being invested, it's how much could the investor lose. It's against that measure investors need to assess returns. And because different styles have different degrees of implied leverage, it's no good just saying there is a notional amount which can be used when measuring the return against the calculated information ratio. We need some way of defining the contingent capital being allocated and measuring the return on that. That is a true risk budgeting approach. That ratio – return on contingent capital – was christened the Pareto Ratio by Rob Liesching– after the creator of the Pareto distribution which models the tails of non-normal distributions.

The information ratio and value at risk and so on are modelled on normal distributions. However, we know in currency markets that distributions are not normal, they have fat tails; so you need to model the tail of the distribution in order to assess what contingent capital you're putting at risk.

MF, RBC Capital Markets So do you think that's going to be the new primary risk measure?

MS, Pareto Partners I think it makes a lot of sense to measure returns and risk in those terms for investment strategies that are not investing in a hard asset, yes.

RM, Meijer & Co Well I've always called this the put option benchmark – the cost of the put that protects your assets.

MS, Pareto Partners If your strategy is purely long options, then you can say without doubt what your capital at risk is: it's the premium you're prepared to spend; then it would make sense to measure the return on that budget. But that's the only example in which it's absolutely precise. The rest of the time people are trying to assess what the contingent capital is and how much they could lose. Drawdown is a way of looking at it historically.

RM, Meijer & Co Certainly some clients are now using maximum draw-down as the key risk parameter, rather than the information ratio. That could change the terms of engagement.

JA, State Street It's such an easy thing to game though. Even information ratios are gameable. But a single period draw-down limit can easily be manufactured. If I was a client I'd react very strongly to someone selling me a strategy based on maximum draw-down.

MS, Pareto Partners Wouldn't you get some comfort from the fact that there was a limit to the amount that you could lose?

JA, State Street No, because I know how easy it is to manufacture a strategy. I'm not disagreeing with the idea that investors are much more averse to losses than an information ratio would convey, I just think the leap from there to maximum draw-down throws away far too much information. The Sortino ratio [a modified Sharpe ratio] captures a much richer sense of what those negative returns look like. There are other ways: adjust the utility so you penalize losses within a utility function more than you penalize gains, for example.

Putting your money where your mouth is

KM, Dow Jones We've been talking about traditional currency overlay. But what about actually allocating real capital to currency as an asset class?

MS, Pareto Partners Well currency can only be an overlay, it isn't an asset class in itself even if you create a risk budget and spend it on currency.

JA, State Street The issue arises though where there is a capital requirement. In Michael's case I think it's clear there isn't a capital requirement. A bank would be happy to write forward contracts with you against an unleveraged asset base. That may get you 100 basis points of alpha, but it's not going to drive allocations to currency as an asset class – where asset class means a distinct source of return unrelated to the underlying assets. People have identified FX as a source of alpha and from there start to look for an 8% or 9% return, at which point most banks need some capital to work with.

MS, Pareto Partners Well I wonder, do people actually take cash off their clients to run an absolute return currency programme? And is that attractive for the client?

JA, State Street If it does have no correlation with traditional assets, then that would be very attractive.

MS, Pareto Partners Absolutely, and almost by definition it has no correlation with traditional assets.

JA, State Street Yes, I'm just worried about the risk factor.

Peter Nielsen , Royal Bank of Scotland That depends how much octane is attached to the strategy.

MS, Pareto Partners Well, look at it like this. An investor's entire international equity exposure is pretty large. The allocation to currency as an absolute return strategy might be on a tenth the size of that, in which case 80 or 90 basis points is coming to much the same thing [as the 8% to 9% targets mentioned above]. It's a question of how you define your denominator, because there are no actual securities being purchased to define the denominator as there would be with other asset classes.

JA, State Street Well, there are with a currency fund.

MS, Pareto Partners Yes. If it's a fund, then you're buying a unit in a fund and now you've got a security and now you can look at it like any other asset. But you've allocated cash now, so it's not an overlay; you haven't allocated contingent capital, it's fully funded. Then the question is: how much will a bank allow you to lever up?

RM, Meijer & Co How much leverage is allowed in these funds?

BJ, Gartmore Some, a lot.

MS, Pareto Partners It depends on the type of strategy. Those that rely on carry trades have a tremendous amount of leverage.

PN, Royal Bank of Scotland Yes they do.

BJ, Gartmore I don't think that you can say that currency overlay and currency funds are two different beasts, it's just an issue of leverage. If a client says: "I've got $100 million-worth of foreign currency exposure, manage that foreign currency exposure and you can use up to 10 times leverage", then that's the same as saying to the currency fund: "Here's $100 million, generate me 10%". And you don't need the underlying assets, you just need a forward loan with the bank. Now, the bank might want margin on that but if an investor doesn't want to trade on margin the banks tend to roll over.

MS, Pareto Partners But even if they don't give you the cash, they have to define the amount and the amount of leverage because you can't otherwise compare an overlay programme that generates 1% return on a billion dollars of assets with something that says: "I can produce 10% return", because you're not going to give them the notional amount to produce that 10% return. I mean how big a currency position would you have to take to do that? It would be huge.

Technology's two faces

KM, Dow Jones One other issue is technology and the role it plays both in hindering the uptake of new strategies – people need new systems and skills – and also later in increasing the use of these strategies once people are happy with the technology. Peter?

PN, Royal Bank of Scotland Technology has boosted volumes and created a new level of transparency in the market. A number of my sell-side colleagues are reluctant to see some of this change happen but at RBS we welcome it as it gives us an opportunity to talk to a wider range of clients. Also, in the past we've worried about losing liquidity to market participants – particularly those who employ trading strategies that are negatively convex (say, trend-followers who always buy something that's going up or something that's going down). Offering large amounts of liquidity to participants all of whom seem to have models that go off at similar times can be difficult to price. As additional participants come in through the use of new technology they create additional liquidity and that is good for the market.

Andrew Kellner, Calyon Another development is that banks are reducing interbank counterparty risk, which helps with liquidity. The advent of CLS has allowed us to expand the amount of interbank capacity that we're willing to take. You have clients trading through the banks now – for example through prime brokerage – and this is also driving development in the high-frequency trading space.

KM, Dow Jones The technology encourages new types of strategies that exploit weaknesses or peculiarities of individual systems. Does that cause you problems?

PN, Royal Bank of Scotland There is still arbitrage of that kind: legitimate currency triangulation, issues of latency with the different portals and so on. But those are short-term problems. I think the industry is doing a good job of getting rid of or preventing those participants from actually disturbing the developments that are legitimate and need to be encouraged.

HF, BNP Paribas We have experienced a major shift on the trading side because market-making no longer exists: you are increasingly just providing liquidity on a streaming basis to a lot of different platforms. The traders are expected to clear those trades and that's it. On the sales side, technology has altered the relationship between client and salesman from being a transaction-based relationship to an advisory one. And I agree with Peter: there is some sniping – clients trying to take advantage of such easily accessible liquidity.

NB, Crédit Agricole Asset Management Does this mean that we will see wider spreads on the e-platforms [for snipers]?

AK, Calyon Usually the spreads we would show wouldn't move. But if we have a particular issue with one client we would put steps in place to try and address that behaviour. Any bank would look at the situation on an integrated basis across all products. If the behaviour continued would it be a termination of relationship on the electronic side or an overall termination of relationship in FX? That will vary from bank to bank.

Transparency up, liquidity down?

NB, Crédit Agricole Asset Management I'm on the buy-side and I find that on a lot of platforms we can completely lose liquidity from time to time. How can you address that issue?

AK, Calyon I think this is linked to another question: is volatility dying? All of us now stream prices to the multi-bank portals. That means there's great transparency for a small range around a particular price at a particular moment; but with a gap one way or the other, the transparency just causes the market to dry up. Then, after a period of time, the market reaches a steady state again and the liquidity shows back up on the screen. So the transparency introduced by the new technology has created a new discrete jump pattern in the markets.

HF, BNP Paribas Yes. Remember I think it was on 6th August, we had a jump of 170 points in the euro-dollar and then it came all the way back down and then was stuck in a range. At that point more than 1,000 tickets transacted just on EBS in one minute at 2.30 in the afternoon. And we are all taking for granted that EBS liquidity is there for a multiple of what we are providing to our customers via our platforms.

PN, Royal Bank of Scotland The key point is that if technology is only making it easier for the people who are taking liquidity out of the market to interact, then there will be a problem. What we'll find though, as it becomes easier to transact, is that other participants will offer liquidity back to the market.

NB, Crédit Agricole Asset Management Who are these new market participants?

PN, Royal Bank of Scotland There are a number of firms who have as their origin the provision of liquidity in the fixed-income and equity businesses – Nasdaq market-makers – offering liquidity to the exchanges that trade those stocks. They are now deploying that technology – which is better than most of the banks have – to offer liquidity back into the system. It's not a large number of participants at the moment, but I think we'll see more of that.

RM, Meijer & Co What kind of players are these?

PN, Royal Bank of Scotland They're independents. They're not hedge funds, they're independent, private trading companies who are accomplishing the same thing that traditionally only the sell-side banks and spot desks used to do. I think that's an interesting development.

JA, State Street Well a huge amount of capital has flowed into this business on the basis of returns. The last seven quarters represent the strongest cumulative period since the indices began in the mid-1990s. However, we are operating in the most liquid market in the world and when clients retrench in terms of risk-seeking, the first place they look is the most liquid markets. Is this return-seeking capital going to continue to flow into the currency markets? A couple of reasons it might not are: first, a return to significant double-digit returns in the equity markets and second a general retrenchment from risk of any kind, which might come through a major shock. My other question is: "How much of today's liquidity is just a function of low interest rates which give an incentive to all market participants to seek alpha as a source of return?"

A permanent shift

PN, Royal Bank of Scotland It's a truism that the most money gets invested in the last quarter of a trend. In my view we're more than halfway through the trend here and there's still plenty to play for. As to your questions: I have had clients suggest that one of the reasons they're deploying money into this asset class, specifically the trend-following element of it, is that it's a hedge to some of their other strategies – volatility plays and convergence strategies that leave them open to accidents. Those clients are here to stay.

BJ, Gartmore Another key driver is regulation. For pension funds, insurance companies and other risk-seeking investors, there has been a shift away from fairly irregular asset and liability analysis to a more regular and tougher evaluation. That makes it harder to run big asset/liability mismatches. This should help the cause of currency as an alpha generator.

RM, Meijer & Co I think so too. Investors still need to generate similar returns but can't do it by playing the equity card the whole time because it generates a lot of volatility around their liabilities. So people will become more immunized in terms of their assets and that helps this asset class, because it is a diversifier.

MF, RBC Capital Markets Conversely, if you do expect double-digit returns in the equity markets then it's inconceivable to expect the same rate of inflows into the currency markets that we have seen over the last two years.

MS, Pareto Partners I don't think that's necessarily true. The flow from equity markets into currency isn't an asset shift, it's a reallocation of risk. People aren't reducing equity portfolios to go into currency. They can maintain their exposure to the equity market – provided it's a passive allocation – because all they're transporting is the active risk taken above the benchmark. So even if equity markets were to start performing again, there's no reason why they wouldn't keep currency as their alpha engine with equity exposure their beta engine. People would only move out of currency if it started losing money and they thought it wasn't a source of alpha any more.

MF, RBC Capital Markets What about a dramatic decline in currency market volatility?

MS, Pareto Partners Well there are different types of volatility. Long-term volatility helps some people and short-term volatility helps others. But I'm saying that currency programmes and other alternative strategies which aren't using up capital allow people to match their liabilities by, say, taking a higher bond exposure, but still close their funding deficit, because the alpha source is divorced now. Before, the only choice they had was to seek the higher return from equities, which means a bigger mismatch with liabilities.

Sell-side derivatives strategies for sale

KM, Dow Jones So what are the next generation of strategies and products investors should be thinking about?

PN, Royal Bank of Scotland One of the significant changes we've seen is that as yield-curve trades have faded, people have moved to employ similar strategies in currency options. And we've seen a pick-up of interest in things like volatility swaps and correlation swaps in the currency markets – ways to capture some of the triangular gains the sell-side community has spent most of the last four or five years capturing. So, to return to a theme, the industry is making it easier for people to deploy and take advantage of strategies that used to be resident within the sell side and to create alpha with them.

BJ, Gartmore Could I ask the banks what they think about how the exotic derivatives market will develop? Clearly they are used more now and by a broader number of participants, and there is a particular reason it's important. Every day you hear about barriers being taken out. When we were in the 1.20/1.26 [euro-dollar] range everyone said there was so much barrier risk it would influence the behavioural characteristics of foreign exchange.

MF, RBC Capital Markets It's inevitable that volumes will grow as people become more sophisticated in terms of the types of risk they want to offset and in terms of the strategies they employ to seek alpha. Exotics are already having a significant impact on the spot market.

AK, Calyon Yes. I think one of the reasons why we see vols at such a low level at the moment is a result of the exotics that have been written over the past six months.

BJ, Gartmore But what about 2005? Is there going to be more or less use of them? If there is going to be more use of them, then will we get unusual vol curves and new, exploitable anomalies?

HF, BNP Paribas One key point is that corporates, which have been large users in the past, are more and more constrained by regulatory issues. That is perhaps a reason for supposing there will be less volume in 2005. As for impact on the FX market – the obvious one is that the more people who use them, the more they will create technical levels and ranges in the market. I think this will continue.

JA, State Street I think we need to be wary of the systemic effects of some of those trades, for a couple of reasons. One is the phenomenal leverage – the gamma – that these products can create in the market and what that in turn means for its efficient functioning. I can see a situation where a small number of trades end up consuming pretty much all the liquidity in a particular sector of the market and that kind of situation has the potential to lead to a very disorderly market.

RM, Meijer & Co So your main concern is leverage?

JA, State Street My concern is anything that means that the market can't absorb risk to the extent that there's no price for that risk. That would be fine if there was a degree of transparency so that people knew where those risk points were. We don't have that in the FX market.

PN, Royal Bank of Scotland There's also a risk management concern. For example, you can construct portfolios that have gamma going to infinity, and one needs to be very careful about the risk management of that.

Emerging markets focus

KM, Dow Jones Let's move on to emerging market currencies. Is this the next area of investor interest, given the hunt for volatility and returns?

MF, RBC Capital Markets I think we've already seen a huge jump in interest and that can only increase. That is partly to do with the search for returns but it is also because today's emerging currencies are tomorrow's mainstream. And as our clients move into emerging markets in fixed-income and equities so you will see overlay managers having to offer that service too.

KM, Dow Jones And will the key suppliers be emerging market specialists moving into currencies or currency specialists applying what they already know to new regions?

JA, State Street I think the former group are there already; the key question is the latter group.

MS, Pareto Partners Well we have two constraints on which currencies we can actively manage. One is liquidity and the ability to trade high-frequency. The other is data. So even if the renminbi came in as a major currency with great liquidity, it would be hard for people that are model-based to make use of it immediately, because there is no data. With the euro at least you could construct things with the Deutschmark and so forth, but I'm not sure what you would do with the renminbi. It would be a very interesting development.

JA, State Street It would be interesting to see a currency where the traded goods users of that currency were dominant. back to the old days?

HF, BNP Paribas I have a couple of issues as far as emerging markets are concerned. One is converging currencies, like the zloty or forint. Here convergence looks like a done deal but there's a chance of sudden divergence rather than convergence. And there will be some interesting liquidity problems in these new eurozone currencies. Every man and his dog is playing this trade and that will cause problems at some point.

The commodity proxy

KM, Dow Jones There has been a resurgence in the idea of taking views on commodities either alongside or as a proxy for foreign exchange. Do you see evidence of that?

PN, Royal Bank of Scotland There is increasing demand for products and strategies of this kind. Certainly commodities appear to be driving some elements of the foreign exchange business and investors are taking quite sophisticated views on correlations between the commodity, the commodity currency and other currencies, for example.

JA, State Street I think bringing commodities into that mix is something that investors should consider very seriously. One of the great attractions of real assets such as commodities is they offer the most attractive correlation properties at the time you most need them. When markets are in stress, correlations amongst financial assets rise. This means investors don't get the level of diversification that over a longer period of time those assets display. Real assets work in the opposite direction to give a lower correlation or even a negative correlation at those points. And that has an interesting corollary: investors don't want to end up taking physical delivery of commodities so they have to use derivatives. And that means finally rewriting plan sponsor guidelines to reflect these market developments.

SB, Euromoney Is this really about diversification and a proxy currency trade? Isn't it just more of the search for returns?

MF, RBC Capital Markets It's partly the latter but it depends on the type of client base. The macro hedge funds have a greater capacity to trade on FX and commodities jointly, but will their interest last? Investors have definitely moved into commodities as an alternative way of expressing their currency views. But that's not going to be a long-term strategy.

AK, Calyon I'm not so sure. Look at Vega: earlier this year they started up an energy or commodities-only trading fund having already been active in commodities. Among the hedge fund community we are definitely seeing an expansion of asset classes: they're trying to replicate strategies that have worked in fixed income in other asset classes such as FX and commodities. In the commodities asset class though there are two different approaches: there's the trading of commodities purely on a financial basis and then there's the trading of physical commodities and there's a big, big difference.

KM, Dow Jones Sadly that's where we have to leave it. Thank you all very much.

Gift this article