| FXMarketSpace’s standard fees | ||
| $blns/month | Passive | Aggressive |
| 0-10 | $4.00 | $6.00 |
| 10-20 | $3.60 | $5.40 |
| 20-40 | $2.88 | $4.32 |
| 40-60 | $2.30 | $3.46 |
| 60-80 | $2.07 | $3.11 |
| 80-100 | $1.87 | $2.80 |
| 100-120 | $1.77 | $2.66 |
| 120+ | $1.68 | $2.53 |
| Source: FXMarketSpace | ||
The plans of Reuters and CME joint venture FXMarketSpace to bring exchange-like efficiencies to the spot foreign exchange market appear to have been dealt a huge blow. Core to FXMarketSpace’s proposed business model is the use of CME Clearing as the central counterparty (CCP) to trades.
The use of a CCP makes trading anonymous and lowers credit risk, as the clearer is the counterparty to all transactions. FXMarketSpace had been widely touting the benefits of its model, saying that it would open the forex market to a new range of participants because credit was removed as a discriminating factor in sourcing liquidity. Exchanges also tend to net deals, and forex participants had been waiting for clarification on whether or not the new platform would adopt this initiative.
Netting results in fewer trades being sent for final settlement, reducing the ultimate cost of trading for some market participants. The ramifications of FXMarketSpace’s launch have been discussed widely, and of particular interest was the impact it would have on CLS, the forex settlement system.
Is CLS too costly?
Few in the industry would disagree that CLS has worked very well, removing Herstatt or settlement risk as a factor; CLS is accounting for a growing percentage of the market’s volume. However, CLS is considered by many in the market as expensive, with some even calling it a bad tax. It charges on a per ticket basis for settlement and does not allow pre-netted trades to be submitted for settlement. According to Jonathan Butterfield, EVP Marketing and Communication at CLS Bank, when the system was set up, banks signed up for it on the basis that they would submit gross trades for payment versus payment settlement.
In response to criticism that it is expensive, Butterfield says CLS is a utility owned by 71 financial institutions that has fixed infrastructure costs. These cannot be avoided and the only way that charges can be reduced, on a per ticket basis, is to increase the number of trades put through the system.
Butterfield points out that as a result of increasing volumes, CLS has been able to reduce its charges to many participants. “The cost of settling trades can’t be completely eliminated. The way to make it cheaper in CLS is to put more through,” he says.
Bringing down prices
Butterfield adds that earlier this year CLS reduced its ticket cost for trades at or below $1 million or €1 million, which now account for around 55% of the volume it processes. These now start at 40 pence per ticket and can decline to 25 pence for incremental volume.
For other trades, settlement costs range from 80 pence to £1.10. When CLS went live in September 2002, its target cost per ticket was £1.13. This was achieved in January 2006. However, although it has reduced its costs, with forex prime brokerage now said to start as low as $2 per million dollars traded, it seems some sell-side liquidity providers are running what looks like a marginal business. The attraction of pre-netting and therefore reducing the flow of tickets into CLS is obvious.
But clearly this would have ramifications, given CLS’s fixed costs. The result would either be higher charges, or the end of the system itself, something the regulators and the industry are extremely unlikely to consider.
There has been some debate about how CLS would respond to a reduction in ticket put-through. Butterfield appears to have clarified the issue, at least for the time being. “FXMarketSpace will not be netting and then settling through CLS. CLS raised its hand and as a result FXMarketSpace and CLS have agreed a process utilizing the standard CLS gross settlement model,” he says.
Mark Robson, chief executive of FXMarketSpace is believed to have admitted that the platform will not pre-net at an informal presentation made to industry participants after the November forex Joint Standing Committee meeting held at the Bank of England.
Disappointing, not devastating
However, he has downplayed the importance of this, arguing that deal netting is just one part of FXMarketSpace’s overall value proposition. “I think it’s true that we had hoped that we’d be able to have a net settlement model at the time we were announcing the joint venture. We weren’t committed to it but we have explored the possibility with CLS,” he says.
Because the CME has the ability to settle netted forex trades through CLS, FXMarketSpace might well have believed that it would also be given the go ahead to do the same. But this has not proved the case. Robson says it is disappointing, but not devastating.
“We don’t believe any customers have signed letter of intent with us on the basis of net settlement,” he says, adding: “The argument we’ve been making is we believe we can bring new volume into the market space which wouldn’t be there if we maintain the existing status quo.”
Robson points out that the emergence of high frequency traders, many of whom square their positions every day, has brought a new dynamic to the market. For these, the ability to net might be a more important factor overall, rather than settlement risk.
“CLS has already achieved critical mass. It’s not about trying to pull the rug out from under CLS but about how much more business we can bring into the forex market if we make it more attractive for one of the key factors for growth, the high frequency traders,” Robson argues.
Butterfield broadly agrees with these sentiments, admitting that one of the biggest issues CLS has had to face is the emergence of the high-frequency traders. “Realistically, there are 12 prime brokers in forex and they’ve all had to face compression in what they can charge,” says Butterfield. He says that 18 months ago, the PB range per million dollars traded ranged from $7 to $25. Now the minimum is thought to be as low as just $2.
He argues that the benefits of netting can vary and that it sometimes results in a large, single transaction being sent off for settlement. He questions whether this is really the sort of risk market participants want to be potentially exposed to.
Other routes to netting
Butterfield concedes that although banks signed up to settle on a gross basis, the technology exists for many of them to submit netted trades and that some are looking at using it to reduce the number of deals they submit to CLS. Sources suggest an initiative is about to be unveiled in the cash market.
If this happens, it will be interesting to see how CLS reacts. Unlike with FXMarketSpace, there might be little it can do about it. However, industry participants will have to decipher if this will lead to greater savings, because if fewer tickets are put through CLS, costs will undoubtedly rise and there is evidence to suggests settling outside CLS results in far higher operational costs than within.
For FXMarketSpace, CLS’s refusal to accept netted trades is undoubtedly a blow in its fight to gain liquidity and differentiate itself from the existing cash market. Now, arguably even more so than before, it will have to prove that it does have a convincing technological and cost advantage. The gloves are really coming off.