Pragma, the multi-asset quantitative trading technology provider, has added NDFs to its algorithmic trading platform, Pragma360.
It allows Pragma’s clients to use algos to trade the Brazilian real, Korean won, Indian rupee, Philippine peso, Indonesian rupiah, Malaysian ringgit, Taiwan dollar, Colombian peso and Chilean peso – all frequently traded NDFs.
They can be traded with Pragma’s time-weighted average price (TWAP), volume-weighted average price (VWAP), percentage of historical volume (POHV) and strike algorithms.
Among the NDFs included, the real, won and rupee are the most actively traded globally, between them accounting for just under half (49%) of global NDF turnover.
“This will be increasingly important in light of the upcoming FX global code, which demands greater transparency and best practices in foreign exchange,” he says.
The move has been made possible by the increasing migration of trading activity onto electronic channels.
Curtis Pfeiffer, chief business officer at Pragma, says: “NDFs were a natural next step for us. They have the liquidity, and prices are already streamed to clients electronically, so it has been relatively simple incorporating them onto the platform.”
NDFs are among the fastest growing parts of the FX markets. According to the Bank for International Settlements’ (BIS) 2016 Triennial Central Bank Survey, the NDF market grew by 5.3% between 2013 and 2016, accounting for more than $130 billion of daily currency trading activity.
The numbers are even more notable when it comes to cleared trades.
IHS Markit, an information and analytics provider, says its end-to-end trade processing system, MarkitServ, processed a monthly average of 54,000 cleared NDF agreements year-to-date in 2017, a 920% increase on the same period in 2016. LCH, the clearing house, has cleared an average of approximately $40 billion NDF trades per day in 2017 on ForexClear.
William Black, US head of over-the-counter (OTC) derivatives clearing at Credit Suisse, says: “NDF clearing has seen strong interest as a response to the uncleared margin rules and other market factors.” Those rules came into effect in September 2016, encouraging traders to clear more transactions.
In April, Credit Suisse cleared its first non-member NDF trade with LCH’s ForexClear on behalf of a client, becoming one of the first futures commission merchants to intermediate FX clearing for a non-captive customer.
|Jon Vollemaere, R5FX|
Jon Vollemaere, CEO at R5FX, the electronic trading venue for EM FX, says a lot of traders are diversifying into EMs to make up for disappointments elsewhere in their portfolios, explaining the increased interest in NDFs generally.
“Anyone with dropping G10 volume is looking at ways to prop that up a bit,” he says. “EM is one way to try to do that.”
However, he adds: “[EM FX] is a very different beast and not a simple add on. When we set up R5, we had to build a purpose-built EM platform, with a bespoke central credit and clearing model.
“To really grow EM, we needed something new, so we built a single trading venue that offers all-to-all trading in OTC and prices from different exchanges, and gives clients a choice of whether or not to clear and net.”
This might have particular implications for algos, notes Vollemaere.
“In G10 there is an assumption that there is sufficient liquidity to get into and out of a position,” he says. “That’s not always the case with EM. Also some markets are still mainly voice traded, which limits the use of algos.
“Certain currencies are dominated by a small number of banks and some by exchanges, with liquidity concentrating around certain dates and times. Algos need to take account of all these characteristics.”
In addition, EM currencies often suffer particularly aggressive volatility during periods of market stress when investors shun risk and allocate capital to safe-haven currencies, creating a stampede that can ignore underlying market fundamentals.
The dramatic increase of capital to EM FX via NDFs could therefore be storing up a problem if market sentiment shifts suddenly, for example if rising US rates reaches a tipping point and begins sucking capital out of EMs.
Pragma’s new functionality is offered primarily to customers who trade FX as an asset class, such as hedge funds. These clients have been using Pragma to trade spot FX – and in theory should be among the most sophisticated market participants, and well prepared for any eventuality should sentiment turn.
It remains to be seen what impact the growing use of algos will have on the NDF market itself, but the evidence from other asset classes is that it could lead to smaller average trade sizes, offset by a higher number of trades overall, and a rise in liquidity overall.
Pragma says NDF trading with its algos could eventually become popular with other types of institutions, such as banks, although that would first require electronic communication networks to offer trading in NDFs.
It is also looking ahead to a time when NDFs could be followed by other FX derivatives products becoming tradable with Pragma’s algos.
Pragma’s Pfeiffer says: “The benefits of algorithmic trading will to continue to expand across FX trading. Pragma expects to offer additional FX products, possibly FX options, eventually, but this will take a couple of years.”
And Pragma is not the only institution diversifying into NDFs. Hotspot recently indicated it will list NDFs on Hotspot SEF, its swap execution facility (SEF), previously called Javelin, at an unspecified time later this year. This follows its recent filing of a revised rule book with the Commodity Futures Trading Commission, setting out new technical details for the Hotspot platform.
Hotspot says it is in the process of on-boarding participants to trade NDFs.
R5’s Vollemaere welcomes the market’s growing interest in NDFs, exemplified by the developments at Pragma and Hotspot.
“More NDF trading activity is good for the wider market and EM liquidity in general,” he says.