Sebastian Hofmann-Werther
Sebastian Hofmann-Werther

Automation and Volatility: What did we learn?

I would argue that real money firms learn less about the efficiency of their workflows and the performance of their technology during quiet markets, when the overall cost of trading is low, than they do during stressed market conditions when volatility suddenly spikes

Sebastian Hofmann-Werther, Head of EMEA at 360T, reflects on some recent lessons learned regarding FX automation in this most unusual and unpredictable year.

I would argue that real money firms learn less about the efficiency of their workflows and the performance of their technology during quiet markets, when the overall cost of trading is low, than they do during stressed market conditions when volatility suddenly spikes. This might help explain why, following the volatility that we saw in March and April, there seems to be a renewed focus from this client segment on something that was already a hot topic in the industry: automation.

While the functionalities that differentiate various automated trading tools are becoming better understood, what is worth paying attention to now is what we actually learned about them when it matters most – in volatile conditions. With that in mind, here are five important takeaways regarding FX automation evidenced by the events of this year.

1. Automated trading worked even in extreme conditions

We did a study comparing the trading activity of firms that are currently consuming the 360T Swaps Data Feed (SDF) to execute in an automated manner and those instead execute via RFS point-and-click trading. Looking at the June IMM for EUR/USD we created a chart showing the bid/ask spread and the market midpoint – the latter of which comes from the SDF – across the three most volatile days in March. And looking at this chart a very interesting story emerges.

It is immediately noticeable that the orange dots, which represent trades that are not being auto-executed, start to drop off and eventually completely disappear over the course of this highly volatile three day-period. 

Why? We can’t know for certain, but it could be because these firms are taking more time to conduct price discovery or are being forced to negotiate their trades through opaque, non-electronic channels which could make it harder for them to achieve and prove best execution to their investors.

By contrast, firms that were executing their trades automatically during these three days were able to trade continuously, and the fact that this chart shows their trades hugging the SDF midpoint for the most part indicates that these firms were generally able to execute at the market midpoint despite the volatile conditions and wider spreads.

2. Having the right data is crucial

Data is the engine that powers automated trading, but not all data is created equal. For example, EUR/USD spot market data is plentiful and widely available from numerous sources, unlike swaps market data that can be harder to access given that more of these products are still traded via voice channels.

Moreover, even when FX swaps are executed electronically it is via RFQ, meaning that the price only exists during a very short window of time and it is specific to the requesting counterparty based on their credit, profile, associated yield, etc. Therefore, any data feeds built off these prices will not actually represent where the market is at any given point in time. 

By contrast, our Swaps Data Feed is integrated directly with the pricing engines of more than 20 top FX banks and from these, once the relevant aggregation and sanitisation methodologies are applied, we are able to construct an accurate midmarket price for FX swaps that are completely disassociated with the various factors that dictate RFQ pricing. An inaccurate data feed is clearly not something that you want to build auto-execution functionality around and so firms need to be looking under the hood to check what data sources are being utilised.

News
Source: 360T internal platform data

3. There’s no “one-size-fits-all” solution

The importance of customisation has been a recurring theme in our conversations with real money clients around automated trading. These firms increasingly require tools that can be configured to adhere to a set of highly bespoke rules for auto-execution in order to meet their specific trading requirements and to ensure that they can be adapted to changing market conditions. Auto-execution tools that only have basic parameters around currency pair and the size of the order are blunt instruments, when real money firms today need to execute with precision, especially in stressed market conditions.  

It’s also important to note that automation is not an all-or-nothing proposition, and we’ve actually seen many of our real money clients derive significant value from automated solutions within their workflow even while they choose to continue executing their trades manually. 

4. Oversight and compliance remain key concerns

We all know how important it is today for real money firms to have the right oversight and control functions in place on their trading desk, and that the idea of automation itself can initially spark anxiety as people equate removing manual processing with removing human control. However, the reality is that the rules-based nature of automated trading actually means systems can be configured so that they literally cannot break this set of predetermined rules. So the element of human oversight is still very much there, it’s simply that it takes place before the trading is even conducted.

Our clients have found this particularly advantageous this year because even though they have been forced into a decentralised working environment the rules for their automated trading activity can still be configured centrally and they have been able to maintain a complete audit trail of everything.

5. Technology needs to continue evolving

The FX market is not static, and neither are the firms that participate in it. Real money clients are constantly seeking to streamline their operations, improve their trading capabilities and deliver greater returns to investors. That’s why they need technology solutions that are equally capable of evolving and delivering new functionalities which can help them achieve these goals. 

It’s probably safe to say that there are no automated trading solutions for FX today that will completely satisfy all of the current and future workflow and trading requirements of every real money firm. Bearing this in mind, what becomes important are the mechanisms that are in place by which a technology partner is able to adapt and enhance their automated trading offering based on the various and changing needs of their clients.

Conclusion

The trend in FX is clearly towards more automation, not less. However, conventional wisdom has largely held that in volatile conditions more of the market would generally shift back towards manual execution channels as human traders would look to assert control. But what has been fascinating to observe this year is that in many cases this shift never really occurred.  

This is at least partially because real money firms have realised that automated trading does not have to involve relinquishing control, and that in fact this technology is designed to empower traders. Moreover, now automated trading has been proven as safe, secure and significantly more efficient even in the most volatile market conditions, we expect the demand for this technology to only increase further.