Goldman Sachs goes live with smart NDF algo

Initial trades have already taken place with the algo, which draws on internal and external liquidity

Goldman Sachs - New Jersey - Getty - web.jpg
Goldman Sachs: the dynamic hybrid algo launched with USD/INR and USD/KRW

Goldman Sachs has introduced what it believes is the first “smart” algo for trading non-deliverable forwards (NDFs), as it eyes growth in liquidity on external markets and a pickup in interest in the instrument.

The first trades have already taken place with the NDF dynamic hybrid algos, built drawing on Goldman Sachs’s experience with its existing algos for deliverable currencies. Like those products, the new NDF algo can pull from the bank’s internal liquidity – its matching pool and e-book risk transfer – and gain access to multiple external venues from EBS as well.

Algos for spot FX trading are well established and there have been stirrings of interest in bespoke ones for traders. As NDFs are not viewed as being in high demand, they have not been considered a likely candidate for development. But David Wilkins, head of global e-FX sales at Goldman Sachs, says the bank has experienced a rise in activity and is responding to a “heavy amount of client demand”.

“We’re seeing a real sea change in the amount of NDFs that are being traded electronically in the market from our clients,” Wilkins tells FX Week.

Most of the early users of the NDF dynamic hybrid algo are asset managers and there are also some sovereign wealth funds involved.

Goldman does already have a time-weighted average-price product for its clients, says Ralf Donner, head of client FX algo execution at the bank. 

“The demand was very much for the smart kind of algo where you’d essentially be able to buy on the bid, very opportunistically aggress [and] don’t have a fixed end time. All of those smart features of algos that they like in the developed market space, they were looking for in NDFs as well,” Donner says.

The dynamic hybrid algo launched with USD/INR and USD/KRW. The bank plans to extend the service to USD/BRL and Latin American pairs, and also to less liquid Asian pairs, such as USD/TWD, USD/PHP and USD/IDR, depending on how the first currency pairs go.

…the most well-developed order books are in currency pairs like dollar-India and dollar-Korea, so that’s where we started
Ralf Donner, Goldman Sach

“Among Asian NDFs, if we look at the liquidity we see on external markets, then the most well-developed order books are in currency pairs like dollar-India and dollar-Korea, so that’s where we started,” says Donner.

The Indian rupee NDF is popular on the Singapore Exchange, but Goldman’s algo does not use it as an external source. The company might reconsider when it starts looking at Latin American pairs in particular, according to Donner.

He noted there are several differences when sourcing liquidity for NDF algos, compared with their spot counterparts. For the majority of pairs, the Brazilian real excepted, the one-month date becomes a replacement for spot in a developed market. This one-month date is therefore traded electronically and liquid, and rolls to other dates post-trade. 

Liquidity focused

Another difference is that an absence of fragmentation means external liquidity is fairly focused on a couple of venues, says Donner. This makes having access to robust internal liquidity, and combining the two in an intelligent way, important.

“We’ve been pleasantly surprised by the fill rate, and the liquidity that is available through the combination of internal and external liquidity in our product. We’ve seen tickets that could be considered sizeable go through in a matter of minutes and with no visible market impact,” he says.

“Clients are attracted to an algo offering that uses franchise liquidity intelligently to reduce the market footprint they have going external,” he adds.

As for what comes next, Wilkins says Goldman Sachs is preparing several big releases in the algo space this year. However, it is unlikely that the company will follow up with algos for options and swaps.

Additional reporting by Karen Friar

  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have a FX Week account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: