Foreign exchange debate: The illusion of normality?
Despite, or perhaps because of, the changing nature of the interbank market, liquidity has returned to most corners of foreign exchange. But uncertainties remain over quantitative easing, Japan and the recovery. The panel gathered at the end of 2009 to find we are still in uncharted territory.
• Emerging markets still offer many of the most interesting trading opportunities
• Dollar versus G3 and dollar versus emerging markets are the two fundamental trades
• China will be a key story in the next 12 months
• Liquidity is back to normal in most currencies but not in longer-dated derivatives
• Cash volumes are down and the macro-economy is uncertain
• The changing structure of the interbank market is fundamentally affecting the nature of liquidity
Foreign exchange debate: Learn more about the panelists
Simon Brady, Euromoney Let’s start with the macro-economy. Two of the key topics are the US dollar and the emerging markets. Stuart?
We think the majority dollar view is tied in with equity markets: it is a risk-on/risk-off trade. We feel also that emerging market currencies against the US dollar are going to continue to offer value, because many have both a yield advantage and the advantage of underlying economic growth – for example, the Brazilian real and Hungarian forint.
In foreign exchange trading you have to be flexible and to trade crosses that you perhaps traditionally would not have. For example, trading Brazilian real against South African rand is an interesting play on relative growth. We are quite happy to put on that kind of trade, although clearly you need to watch liquidity conditions. That has been very much our approach from an investment perspective, and we don’t see that changing as we move into next year , apart from, perhaps, dollar/yen. We don’t feel that Japan can see an economic turnaround with a very, very strong Japanese yen. There are a lot of people in the market who think dollar/yen should be at 70 and I have heard the number 50, which I can’t quite believe as I am in the 110 camp, but this has not proved a profitable trade this year  unless you played it against the Aussie or the euro, which was a range-bound cross.
SB, Euromoney What about Record? What have your clients been doing and thinking about the dollar?
What we do see, certainly, talking to clients and particularly talking to US clients, is a widespread expectation of continued dollar decline: there are a lot of dollar bears out there in the US who take that view because they need to protect against imported inflation and the risk of dollar depreciation. I don’t think many of them necessarily want to speculate against the dollar but they at least want to protect themselves against its decline. My sense over the past month or so is that the tone may have changed to be more questioning of a continued dollar bear position than would have been the case perhaps a couple of months ago.
SB, Euromoney Vincent, what have you seen in the derivatives market?
SB, Euromoney Paul, as a hedger what assumptions do you make?
SF, Threadneedle Also let’s not forget that the euro/dollar high this year is nowhere near the record high even with all this talk about reserve diversification into euros. This is still a dollar-centric world and it is going to stay that way. As for yen, if you are betting the ranch on selling dollars and buying Japanese yen, that is the wrong thing to do because it has been noticeable that dollar/yen has stabilized a little and is certainly not going down when bad news comes out on the US economy. There are creeping fundamentals coming in and next year I think things will change. It won’t be exclusively about selling dollars against everything; there might be some unusual little plays into the US dollar.
SB, Euromoney Do you think all the news about emerging market currencies is also stale?
SF, Threadneedle Yes it is. There has been a lot of pick-up in two-way volatility in the market, which is usually a sign that the market is getting frustrated and that there is profit-taking. If you have had the trade on all year, let’s say long Brazil, even long Norway, and short dollars, you are going to bank some.
VC, HSBC I do think a key macro trend will revolve around China. They definitely want to internationalize the renminbi. They need to let it appreciate at some point and they need to do it soon. What they want to see, though, is a pick-up in exports.
JWC, Record The other topic we haven’t discussed within developed market economies is the impact of quantitative easing and particularly its unwind. It is clear there has been a massive distortion of short-term rates by governments and the consequences of the unwind are going to be pretty significant.
DB, Deutsche The big trade over the next 12 to 24 months will be the rate the taps turn off on a relative basis country to country.
SF, Threadneedle It could be reflected in the curve as the bond markets take the pressure in the long end and you just get constant roll-down in the short, three-month contracts. That is what happens. I don’t think that is particularly healthy, but at some point it does kick in and help currencies because the long end of your market is yielding a lot more than it did a year ago. It does become an attraction; that is just one angle we have looked at – a US curve-steepener.
DB, Deutsche The other big thing that is going to come into play in the macro theme is relative demographics. Countries do really well once 35- to 55-year-olds make up more than 60% of the population, because there are relatively few old people to feed, there are relatively few young children to feed; everyone is working. Japan went through that in the 1970s and 1980s; the US went through that in the 1980s and 1990s, and Europe is peaking now. As the productive parts of the world become the emerging markets, that is going to make big differences to the currency markets. Europe is peaking about now; the US peaked in 2000 – timing for the first stock market peak; Japan peaked in 1990.
SB, Euromoney We come on to a subject that normally creates a fair amount of controversy, which is liquidity. Frédéric, how do you think the liquidity situation is at the moment?
FJ, SG I have been amazed at how quickly liquidity has come back and how quickly the market has normalized. Go back a year: we were all watching, checking each trade and each counterparty to make sure that things were going to settle properly. Now look at how spreads have come in and look at the competition. Liquidity is a function of competition and the banking sector was badly bruised by the crisis and yet we now have liquidity conditions which are, I would deem, normal. Liquidity is back.
DB, Deutsche We are all moving one more decimal place down, so liquidity spreads have to be narrower than they were a few years ago because we’re down to the fifth decimal point not the fourth. Volumes are a little lower this year compared with last year so in that sense liquidity is not as great as it was last year. The volumes that we saw at the tail half of last year were absolutely phenomenal. But the trend still seems to be up, and spreads are narrower, so I think the FX market is extraordinary.
SB, Euromoney In the derivatives market, Vincent?
VC, HSBC I would add a couple of things. First, I think it is remarkable that liquidity has come back with a vengeance in a lot of emerging market currencies. It is not true of all of them but look at the Brazilian real or the Korean won and we are definitely back in terms of volume and spread to pre-crisis levels. Central bank intervention increased liquidity and decreased volatility and therefore encouraged new positions as people became more comfortable with the market trend.
In FX derivatives liquidity has come back, possibly not to the same extent as in cash for a number of reasons: first the corporates are using the derivatives market in a different way now. Structures they used to do long-dated – six months to a year – they are either not using or they are using them in a shorter-dated manner. And a lot have decided not to hedge this year after the pain of last year. That is one reason you have less liquidity and certainly the supply of volatility for slightly longer dates has been greatly reduced.
In addition to that a number of players have disappeared from the derivatives market. Some of them were professional market makers and some of them more sporadic players – local banks in EMEA, with the exception of Turkey. They used to provide liquidity in FX options and they are not active any more.
DB, Deutsche To Vincent’s point on the long-dated currency options – those out past 10 years; a lot of people aren’t playing in that space any more. So there has been a hangover from books that have that risk and that have had to sit in the market, particularly in long-dated structured yen options where there is a definite lack of liquidity.
Also on the same point I would say that the liquidity providers have become smarter in the e-space. They have developed auto-quoting engines, they have reduced latency and they manage risk better with auto-hedging tools. Because of that, overall liquidity has increased tremendously and I wouldn’t be surprised to see spreads narrow even further from current levels.
VC, HSBC That is a valid point. I would not go as far as saying that you can only execute small amounts in good liquidity – we can all debate what is small for this currency or that currency – but the point you are raising is about the depth of the market. There were times last year when, regardless of the currency, there seemed to be an endless pool of liquidity. There was real depth to the market, with minimum slippage in executing orders that could be greater in size than the daily average. And in FX options this was also true, though the big trades here do not go through the market, they are warehoused by the banks. There used to be quite a vibrant inter-bank market in FX options, including in emerging market currencies, until Lehman last year. This has disappeared. If you have a very diversified business model, you are in a good place. If you don’t have that then you will find it very difficult to execute any kind of big order for a client.
RO, Oanda What I find remarkable if you just do the numbers is that while the pure spot market is today something like $1.2 trillion, that means that per second we have only $12 million of transaction volume going through the system. That is 6 million buyers and 6 million sellers – which is nothing.
DB, Deutsche If you wait an hour it is a lot.
RO, Oanda But if you ask your market-maker how long he will warehouse the risk there are very few banks which want to warehouse risk for anything more than 10 seconds or half a minute. If you then look at that, then there is not a lot of liquidity. What would be interesting for me is your estimate of how large an order it takes to move the euro/US dollar exchange rate by 0.5%? How much does it take to move the euro/Canadian dollar exchange rate by 1%?
SF, Threadneedle What is a big ticket now in euro/dollar for the banks? How much? Can somebody make a statement about what is a big ticket now for the euro/dollar? €300 million? €500 million? Mid-market, 12 o’clock?
DB, Deutsche You can clear 500 [million] euro/dollars pretty easily.
SF, Threadneedle Is that a big ticket?
MW, Credit Suisse Above a billion is a big ticket.
VC, HSBC It was not the case a year to 18 months ago.
RO, Oanda What is the spread that you would charge?
MW, Credit Suisse You wouldn’t come out with a two-way spread. Usually you would work the order into the market and this way you would get a much better fill than quoting a two-way price for the client on such a large order.
RO, Oanda Wouldn’t it be interesting just to find out what is today’s spread?
MW, Credit Suisse Now, perhaps 30 pips for $500 million, for a yard, 12 o’clock London time.
RO, Oanda And for $100 million?
MW, Credit Suisse $100 million is probably six pips.
VC, HSBC Yes, six or seven pips for $100 million. It is a huge improvement compared with where we were six months ago when it was 30 pips for $100 million.
PD, ShellI would be interested to know if many people come to you and ask you to execute half a yard in one amount?
DB, Deutsche People do, but it is very rare they ask for a price these days.
PD, Shell OK, so you are not taking the risk, so the client is, or you wouldn’t be charging that sort of spread.
MW, Credit Suisse Perhaps, ‘no worse than price X’.
JWC, Record There are obviously a number of different ways that one can execute and it is incumbent on the user to execute in the way that gets the best terms from the bank. There is something of a two-way street to that, we wouldn’t try to impose yard sizes – in fact our process specifically tranches up deals and cuts them into slices where we expect our footprint in the market to be invisible. We think that is a duty we owe our clients. The threshold at which we have had to do that may have shifted somewhat over the last year or so but frankly we have always been sufficiently far below that threshold for that not really to be an issue. I take Paul’s point, the number of participants who want to be able always to get tight two-way prices on deals of that size should be limited – it is an odd set of circumstances where you don’t have the opportunity to execute that in a more constructive fashion.
PD, Shell There was less liquidity just after Lehman went bankrupt, but we have seen a big improvement in recent months. The biggest concern for us was in the emerging markets because there definitely wasn’t liquidity supplied by some of the international banks. During the worst days of the crisis we managed to keep dealing by using our local bank relationships.
I am still perplexed, though, when I see a large movement in the FX rate because someone has done a very large deal ticket. Perhaps it is because we are willing to run a bit of risk and we are doing it every day – if it was the only deal you had done for a month and it was very large I could see why you might want to transfer that risk to the bank, but as we are dealing FX every day it is very simple and efficient to average the rate in.
SF, Threadneedle Are you transferring risk to the bank? My argument is that we feel that the banks have reduced risk exposure in foreign exchange and that they are not prepared to warehouse risk and that if a trader has a position then they want to get out of it straightaway. We do need the banks to take risk and we feel that the credit crisis has been transferred to the foreign exchange side of the banks and the risk tolerance of the banks on the FX side has been slashed. That is an issue and that has taken a lot of liquidity out of the market.
SB, Euromoney But these guys are saying spreads are tighter than ever yet you are saying the banks aren’t taking any risks and aren’t trading FX – the two statements seem not to match up. Paul is saying you can click your mouse 10 times, do 100 million and get no slippage, which to me sounds like liquidity as well. Which is true?
MW, Credit Suisse Liquidity has got better.
FJ, SG Exactly. And if you think about FX as a business for the banks, then it makes sense. Yes, you have settlement risk and you have counterparty risk but in terms of maturity the risks are short-term. Sometimes you have sophisticated products, but you don’t have that lack of transparency which you had with the very sophisticated credit products. So banks are happy to invest in the business and assign capital to it.
VC, HSBC You will be surprised to hear, Stuart, that a lot of banks have reallocated resources and value at risk to FX. This is a very different situation from a year ago or even six months ago. We are willing to increase the risk in our foreign exchange business for clients, which translates to better execution and more competitive pricing. We are not the only one.
SF, Threadneedle You are talking about G7 currencies. But if you move outside G7 currencies?
VC, HSBC I am talking about everything.
SF, Threadneedle You won’t get a price from a bank for dollar/Brazil outside the hours Brazil is open. You won’t get a price in dollar/rand outside the hours that dollar/rand is open; you won’t get a price in noki/stoki overnight Tokyo. I have held up the market on one side in tiny amounts; it will not be dealt and I am just resigned to the fact now. I don’t even argue with a bank, I just say: "OK. We won’t get anything done"; that is a fact of life. We have accepted that as a client now; that is the just the way it is because in Japanese hours nobody is going to warehouse a risk in Norwegian kroner or Swedish kronor whereas a year ago you would have warehoused that risk – as has been our experience – you would have taken it on and that is the part of the market that has disappeared. The G10 currencies are there, we accept that – liquidity is good, the pricing is good – but with anything that is remotely exotic, it is difficult.
DB, Deutsche That is definitely the case in the emerging market currencies, not the G10. This is about rationalizing resources. It is pretty expensive for me to put a noki/stoki trader in Tokyo who is bored 364 days of the year. Inefficient resources get trimmed out and markets have certain opening times and Brazil is one. It doesn’t surprise me that at 8am or at 9am London time you can’t get a great price in dollar/Brazil.
VC, HSBC But that will change if the business is there. So we have noticed that Asian clients want to trade those crosses because of the links between Asia and Brazil. We have resourced our Hong Kong office with people who can talk Brazil and trade Brazil and who will make a price in Hong Kong time. The more minor currencies may fall under the radar, but any G10 trader sitting in Hong Kong/Singapore/Tokyo can make a price in noki/stoki. That is not particularly complicated; you don’t need to employ somebody to do it.
DB, Deutsche It is wide though.
SF, Threadneedle You said you would make a price any time in dollar/Brazil? Yes you will but I am not going to trade on it because it’s too wide.
VC, HSBC Naturally it is going to be wider than during São Paulo hours.
SF, Threadneedle Yes, because you don’t want to warehouse the risk, you want me to warehouse it for you; that is my point. As an asset manager I am an investor and I have to do the best for my customers. For me it is about getting the best price so I will, as I do, put prices and orders into the markets and that is how I trade. That is the only way I can get on terms with you guys.
VC, HSBC We all know 95% of the dollar/Brazil real volume goes through when the BMF in São Paulo is open. So if you want to execute a dollar/Brazil real trade in Hong Kong because something has happened over the weekend or overnight and all the emerging market currencies are 3% weaker or 3% stronger, you can do so with us. It is quite normal, given the uncertainties, that the price is wider. It simply reflects the liquidity. It is not a question of unwillingness to warehouse the risk, just the uncertainty between the Tokyo hours and the opening in São Paulo.
SF, Threadneedle Yes, but your dollar/Brazil trader should know the market a lot better than I could ever know it because they are the market maker. He or she should know where that market is going to open at the time it opens. That should be reflected in his or her price as a professional.
VC, HSBC Certainly the trader should know the market and know where the opening is. But you would only trade dollar/Brazil real or dollar/South African rand in Asian time because you have had an event which has now dislocated the market, otherwise what is the point? And that means there will be uncertainty and that is reflected in the price.
The role of the algo
SB, Euromoney But I understood that the trading desks are increasingly algorithm-driven. Doesn’t that give you trading capacity outside hours?
MW, Credit Suisse Exactly. In the multi-portal world we cannot rely on ‘n’ number of traders working 364 days of the year without enough business, so you have algorithms to do the work. It just takes time to build up the right algorithms to do a solid job with the more exotic stuff.
DB, Deutsche When I started in this business there was a blackboard on which you wrote all the orders, so you knew where the market was going to open. As things are being traded electronically those big order boards aren’t there any more, they are out there in cyberspace. You don’t know where the pockets of liquidity are and so sometimes you hit them and you can sell a billion euro/dollar and it doesn’t move the market at all and other times you can’t because you can’t see the liquidity as you used to be able to.
RO, Oanda I have a different view. A market-maker’s job is to provide liquidity 24 hours and, in my view, seven days a week. We have hands-on experience: we make a market over the weekend. On Sunday if you decide to get out of a position, you should be able to get out of it. We, as liquidity providers, should be able to provide you with the technology. Today that doesn’t exist.
SF, Threadneedle No, but I will say one thing in the banks’ favour: we are supposed to be the risk-taker, so that is our responsibility. We don’t blame the banks for things that go wrong or anything like that, that is our responsibility. What we do like, or we wanted, obviously with Mifid, is best execution and I don’t feel that sometimes we get best execution post the credit crisis.
HE, UBS There are a number of different issues here. One obviously is if the markets are more volatile, then for any given trade the bank is taking on a bigger quantum of risk. That is especially true if you want to trade on a Sunday afternoon or a day before the market opens.
Two, a lot more technology is being deployed in the markets. There is much more algo-driven execution in both the interbank market and on the buy side. In theory at least, this should make the FX market even more efficient as the algos shred and atomize risk and every action is transmitted almost instantly across the entire market. In practice, as many spot traders will tell you, it can make it harder to manage significant risk positions without the market running away from you.
So put these two things together and banks are having to adapt in the way they deal with risk in terms of keeping positions and providing prices in unusual situations. That reflects the changing structure of the markets, not simply a credit-driven contraction in risk appetite.
SF, Threadneedle One big change is that you don’t have as many players in FX as there were 12 months ago. The other day I asked a big bank, "do you still have a prop desk?"; the reply – "no, they left six months ago". That straightaway is a chunk of depth and liquidity out of the market so I can’t see how liquidity has increased. Also through M&A the number of banks has decreased significantly over the past 20 years and hence the number of people you can turn to for a price.
DB, Deutsche Spreads for G3 or G7 currencies are narrower. If your price for dealing 10 [million] euro/dollar is lower than it was 12 months ago then liquidity has increased. Over a billion it might be a different issue, so it depends on what you define as liquidity.
VC, HSBC Depth is the issue as opposed to liquidity. Size rather than bid/offer.
DB, Deutsche Exactly. Volumes are 25% to 30% lower this year.
VC, HSBC What I find quite interesting is that trade is down 40% this year, yet if you look at some of the volume numbers, perhaps not for all currencies but for a number of currencies, they are not that bad or are even back to pre-crisis levels. The Korean won is an interesting example. There is something happening. Clearly central bank intervention provides market depth in Korean won but spreads are also narrow. It really depends on the currency, especially when we come to peripheral currencies. Hungary, Poland, are they very liquid? Not really. Is Turkey liquid? Yes. What’s the difference? In Turkey there is still significant involvement from corporates and banks because they didn’t get hurt much – or not at all – during the crisis. Poland and Hungary are different and a lot of liquidity supply coming from the banks or the corporates through hedging in those countries is gone. The problems we are talking about are extremely complex and it is not possible to paint all the currencies with the same brush.
SB, Euromoney Stuart, from your previous comment you would much rather see the banks maintaining big prop desks?
SF, Threadneedle Not necessarily big, but certainly taking some risks so they can understand the risk and the nature of how the market works. Unless you own a position you sometimes can’t understand what the risk is. Can I ask the banks whether it is correct to say that your algorithms seem to be very short-term in nature. All these trades are going into a silo and the algos are just trying to match them and take the spread – which means they must be very short-term in nature. And isn’t that what is creating the volatility?
DB, Deutsche It is not entirely short-term trading, it is people taking advantage of short-term dislocation between two different markets, and the percentage of people playing that game has increased dramatically over the past 12 to 24 months. And there has been a big decrease in the guy sitting on a prop desk saying: "I’ll sell 100 euro/dollar and see if it goes down". Those guys weren’t successful; these algo guys are buying and selling almost instantaneously.
MW, Credit Suisse They are always square.
DB, Deutsche They are always square so they are not using a lot of risk. The time they have a position is six milliseconds or five milliseconds, so people are investing into that space and divesting from the big macro space.
RO, Oanda This explains why liquidity is spurious. Like in a shop window, liquidity is only on display.
DB, Deutsche The bulk of the currency risk we take is not in the spot market because that happens so quickly. The risk is in our derivative books and we certainly do commit capital to positions in that business that benefit clients.
MW, Credit Suisse What has completely changed in the past five years is that the interbank market as it existed has completely died. In the old days there was Deutsche Bank following UBS, asking for a quote in $50 million and there was a moral obligation to quote on spread ‘x’ ; that is total history.
Today we have all these ECNs, and multi-bank portals and aggregators that together essentially form one big virtual exchange. Then you put algos on top of that and no one knows who is dealing with whom. It ends up with Traiana, they match it all up, a net amount comes out, it goes to CLS and it gets settled; it is a completely different way of doing FX. And liquidity is much, much better.
HE, UBS The technical efficiency of the wholesale market is much better. However, yes, if Stuart wants to deal 100 million on a given Friday afternoon and no one is willing to put up balance sheet to take the other side, then the interbank market can pass it round as much as it likes, but capital is required to make the trade.
DB, Deutsche Martin is right. The interbank market is transformed. If I look at the types of traders that we had on our spot desk 10 years ago, compared with the guys now, the average guy on our spot desk has a PhD and they are taking the risk in a totally different way. The market is now driven by different participants and we are being forced to be smarter about the way we take risk. We can’t just ring the guy at UBS and ask, "do you have a lot of bids down below?"; "Yes, I have a lot of bids"; "Okay, it is going up". It is different now because you can’t see those bids any more.
SF, Threadneedle You don’t need to know that information if you have confidence. I don’t operate like that; I am talking about myself. We look at markets and we make our own decisions and take our own responsibility and we don’t need to know from a bank that there are three bids lower down to buy euro/dollar; we really don’t. We make investment decisions based on our own analysis.
DB, Deutsche You have a month to make that decision. You have a month to say, "I have done all my research, I think stoki is going up against noki, I have done a lot of research". The spot traders are supposed to be quoting 30 to 40 times an hour; they don’t have the opportunity to go into the library and do the research.
VC, HSBC I would agree with you that we have a different breed of spot trader now but it does not mean that they cannot take risk. They just take the risk in a different fashion due to the increased significance of e-FX platforms. They see a lot of information that they didn’t previously have and they can take risk on the back of it.
MW, Credit Suisse In the last year this industry has increased its risk capital base because FX has low risk-capital usage and low earnings volatility. It is a flow business, it is client led, it is short-term, it is a sweet spot to be in for most banks – and if you look around in the industry the number of banks that are right now beefing up their trading and sales efforts and investing in technology is just unbelievable. That makes me believe that, overall, the risk appetite to be in that business has increased and not decreased.
HE, UBS Also we mustn’t forget that in these supposedly good old days the spreads were wider, which gave a trader more latitude to get out of his position if it started running against him. Now when things are down to fractions of a pip and the machines are betting against you, you can now be burnt to toast within seconds.
SB, Euromoney On that note, we break this two-part roundtable. Next month we cover regulation, CCPs, settlement and counterparty risk, derivatives and the future of the market.