Reactive to Proactive: Managing FX exposures with next generation hedging solutions

Automation has significantly advanced both the risk and liquidity management in FX, says Matti Honkanen, director and head of next gen FX at Nordea Markets. The logic is simple. While risk management requires a lot of thinking that can’t be automated, most tasks are purely manual. For machines, they are easy and fast to handle. For humans, they are time consuming, error-prone and demotivating. 

Consequently, the division of labour is based on giving strategic tasks to human beings and repetitive execution to computers, says Honkanen. “For example, humans still have to decide what kind of risk policy to follow but the execution of the policy can be automated.

The capture and consolidation of FX exposure data is one of the tasks where automation is far better than human labour, says Honkanen. “Once the principles are set for how the consolidation should be done, and the connections to data sources are set up, reading and consolidating data is very easy to automate. This means people can focus on setting the principles and controlling the process, and intervening if necessary. This time saving leaves more time for brain work.”

However, the automation is mostly run by quite old and mundane systems. This means that new technology has a limited role. “For example, in automating the execution of risk management, very few utilise any AI to a significant degree,” says Honkanen.

The more advanced FX treasury platforms have the ability to continuously monitor markets and execute pre-defined hedging strategies instantly to ensure organisations can react rapidly to market volatility. But according to Honkanen reacting to sudden market volatility is not the first priority for most corporates. 

“Usually, the focus is in monitoring the customer’s business and making sure they can manage their risk and liquidity as soon as there is a need,” says Honkanen. “This means they are more or less fully covered when the market volatility spikes. If you need to react to market volatility, you are typically already late. However, naturally corporates also rely on automatic alerts whenever market moves above or below certain thresholds. They can even have limit orders that are executed automatically in those cases.”

“AI can do a great job in finding things that a human being doesn’t pay attention to,” says Honkanen. “Its power is in its superhuman ability to go look at so many places that would be undoable for any human. AI is also very good at analysing and solving certain kinds of complex problems. That’s why it beats every human being in for example chess. However, predicting currency movements is not something AI can do well. FX market is the most efficient market in the world where it is practically taken impossible for any corporate to beat the market. If anyone developed a model that could predict the currency movements, they would make fortunes with it and not sell it to anyone else,” says Honkanen.   

Professional hedging

If risk management includes many repetitive manual tasks where computers outperform humans, in accounting this is even more true, says Honkanen. “There are many questions around hedge accounting that require human expertise, but there are also very many tasks that can be let to computers after humans have set the specifications. Many corporates copy data manually from trading systems to accounting systems still today.”

However, says Honkanen, increasingly many corporates fetch the data automatically from accounting-friendly APIs, and let the accounting system further enrich the data automatically before booking it in accounting in an automated way. “FX has traditionally been a hard nut to crack for hedge accounting, since it is not such a universal topic for all corporates as are for example handling sales and purchase invoices. Therefore, there has been room for improvement in automating that flow, and we try to be in the forefront of that change,” says Honkanen. 

Innovative new treasury technology can also enable dynamic, scenario-based hedging strategies to support the balancing of risk and return, provided that companies are adequately prepared, says Honkanen. “Most treasuries don’t want to let automation decide and execute hedging policies in all the imaginable scenarios. However, the best treasuries prepare in advance for different scenarios, create action plans for each of them and make sure they can automatically run the plan whenever the scenario materialises.”

Risk and return of FX movements and strategies are not easy to calculate without deep knowledge about the business of the corporate, says Honkanen. “FX movements can have vastly different implications for corporates that have similar FX flows, for example because one company have better possibilities to renegotiate their commitments. Therefore, you can’t have one automated tool out of the box that calculates risk correctly for all treasuries. This is a place where human expertise is still invaluable.”

Since treasuries are quite different, their tools for improved monitoring, reporting and execution should also differ, says Honkanen. “Most crucial is to understand the biggest pain points of each treasury, and find the right tools for them. Therefore, we also have plenty of experienced experts that are able to understand customers’ situation when they are in contact, and offer them the right solutions. The digital solutions for treasuries are not so universal that you could just download the right one from a webshop and turn it on. You still need human beings whose judgement you can trust, so that you end up having the right tools in the right use.”

Significant differences

When it comes to discussing the role of automation in FX risk management, there needs to be a distinction between large multinational corporates and SMEs, says Niki van Zanten, FX & liquidity solutions, TreasurUp, because the challenges for both are significantly different. 

“Large corporates typically have multiple ERP systems through M&As and operating in different jurisdictions, often with different geographical or legal settings,” says Van Zanten. “That makes it difficult to align the data that is need or to add streaming, market information and bring everything together. The technology really enables you to not only aggregate the data and to clean it up, have a hygiene factor and link different elements of data together to produce something sensible. For example, correlations across different asset classes, but also real-time exposure changes to anticipate this and immediately take action.”

For the SMEs, the big challenge is that they will rarely have dedicated treasury or FX professionals in the field, says Van Zanten. “The technology can really help to bring the knowledge in-house rather than having to hire FX professionals.”

Automation has a key role to play in managing data, however, says Van Zanten, while technology allows you to handle larger data sets, data is not information. “There is always a conversion that needs to take place. Accounting data is not the same as the market data sets you need for executing trades and FX transactions. Technology can help you to do that in a clean and efficient way if it is designed well.”

The cleaning of the data and the alignment is very important but you also want to be able to expand data sets from exposures and add real-time market data and volatility numbers if you are looking to be more sophisticated and linking into real-time market information, says Van Zanten. 

“It gives you a better understanding of exposure and the risks facing companies and allows you to make a shift from looking at exposures and managing risk and taking action to mitigate risk and not just blindly executing policy settings and saying you have hedged your risk adequately. In the end, you’re not hedging exposures but mitigating the risk that those exposures bring,” says Van Zanten.

As FX treasury platforms continue to develop, they will allow companies to adopt more dynamic and sophisticated hedging strategies.  “Most SMEs do not have a very dynamic hedging strategy, it is based on static policies that are determined by the type of exposure, balance sheet or cash flow exposures or volatility carry. The technology allows you to accurately set the policy but more over it will automatically trigger or execute those policies which has a huge benefit for those corporates in terms of being 100% policy-compliant, provided that policy has been well-thought through,” says Van Zanten. 

While the use of AI and data analytics can enable companies to forecast currency movements and adopt predictive models for better hedging strategies and more proactive risk  management, the technology will not take away fundamental market risk, says Van Zanten. “In all cases, assuming market risk remains, the improvements on hedging and mitigating risk associated with currency movements should really be seen in forecasting and more proactive and dynamic execution.”

For instance, if the volatility in a certain currency is increasing and the carry goes beyond a limit that it actually makes sense to hedge, you can use AI to run a lot of scenario analysis to test the parameters or bandwidth of your, says Van Zanten. 

“The basis of a good hedge programme, especially if you’re looking at cash flow hedges or longer tenors is really how you forecast and AI can definitely bring a benefit in terms of what you put into those forecasts but also how it can improve in the future,” says Van Zanten.  “The last and most logical user case for AI is in backtesting and making sure your policies and your strategies make sense and that people can play around with those policies and see the results of those decisions.”

Hedge accounting challenges

Hedge accounting is a difficult topic, says Van Zanten. “Most US-based companies use different standards to companies in Europe and even within Europe there can be some deviation between countries. FASB or IAS standards differ significantly. Also, the information that needs to be tracked prospectively and retrospectively and provided to auditors can vary a lot. That makes it a difficult but very useful case for automation. You can do scenario analysis on the cash flows that have taken place and also make sure that you meet the requirements for hedge accounting, you need the exposures that you hedge in most cases and you also need to validate retrospectively whether that makes sense. Also, it adds a lot of value in terms of providing documentation,” says Van Zanten. 

New technology can also empower companies to adopt more dynamic and scenario-based hedging strategies, says Van Zanten. “Most corporates tend to have a very static hedging policy. They primarily look at the type of exposures, cash flow or balance sheet, and then have a layered strategy or a 100% hedging of risk. Some corporates differentiate between high and low carry currencies. But very few corporates look to economically hedge those exposures, it is primarily an accounting game which makes it slightly easier to accurately measure the effect of hedging. For dynamic hedging, there’s a lot to be said for using that because typically financial markets do not behave in the process way that static hedging would behave. To make things work economically, you need technology expertise but also large quantities of data, typically a dedicated resource to do that,” says Van Zanten.

However, it is the only way you can balance the risk and return because financial markets are unpredictable and do not work in a processed manner, he says. “If you have a process that is based on predictable outcomes and the counterparty is the financial markets, you might burn yourself. The new technology can help corporates work in a more dynamic way by bringing together different data sets, automated execution, and backtesting to really get to those scenario-based strategies without the abundant resources that were needed in the past.” 

One critical requirement for new platforms is that they are able to integrate with existing ERP and treasury management systems (TMS), says Van Zanten.  “A lot of data comes from the ERP systems – invoices, cash flow forecasting, order management systems etc. These systems are not typically designed to hedge currency exposures or do any treasury management at all. Each TMS has its own benefits in terms of what they are good at and where they might lack a bit of functionality so it is necessary to convert accounting data into useful treasury data. But the next step is aligning that with market data, specifically spot forwards volatility correlations and working together with pre-set validation and execution rules that can help monitor and move the shift from execution to setting up policies and rules and strategies and spending more time on analysis,” says Van Zanten. “That will be the focus for the future – to move the treasurer’s role to one of intelligence and not just execution.”

Intelligent monitoring

GTreasury, a Ripple company, provides SaaS-based automated treasury management software. According to Ben Hipwell, group product manager, a lot of work has been done to make things more automated and to make use of new technology like AI to make things as slick as possible for clients. “This has transformed FX risk management from being a periodic, spreadsheet-driven function to a continuous process with intelligent monitoring,” says Hipwell.

Modern day treasury platforms such as GTreasury’s provide full real-time visibility across global operations and have replaced the manual processes that left corporate treasurers managing stale data, says Hipwell. “This shift has meant that clients are able to proactively manage their FX risk to identify emerging exposures before they materialise losses and can start to use the kind of sophisticated analytics that were previously only available to the largest multinationals.” 

It is also a shift in decision-making, says Farah Lotia, vice-president, product management and quant. “Treasurers have historically been very reactive in their FX risk management because the data has been very fragmented and spread across different systems. What’s changed is the insane level of volatility in FX markets, in supply chains, in geopolitics and, ultimately, in global trade. So the need to react quickly to market events has become even more significant,” says Lotia. 

Another critical step in the evolution of FX risk management and hedge accounting has been the ability for treasurers to have complete visibility of their exposures, says Hipwell. 

Corporate treasurers typically have a network of subsidiaries and counterparties so data systems and flows look very different. Automation has really helped here by enabling the capture of exposure data from multiple sources, whether that’s ERPs or other systems, says Hipwell. “Also the use of AI tooling helps to map to the APIs that are available and able to connect to all of the ERPs and then pull in exposure data and eliminate the manual, spreadsheet-based work that had been used up to now.”

Systems such as GTreasury’s also provide corporate treasurers with a benchmarking facility and the ability to constantly monitor their FX policy, which is another big shift in process, says Lotia. 

“In the old days, when a corporate treasurer wanted to develop an FX trading process, they went to the board and presented a policy that only got looked at when there was a problem in the markets. But with AI and benchmarking and the tools available today, you can be in a cycle of constantly monitoring whether your policy is effective. So not only are you trading within your policy, you are also questioning and amending and predicting what the policy needs,” says Lotia. 

And by having the APIs in place, treasurers are able to execute on those policy changes much more quickly and, ultimately, get better rates and prices for their trades, which, as Hipwell states, is the ultimate aim for companies. 

The advances in technology have also made dynamic hedging much more attainable for corporate treasurers, especially those operating in the mid-tier and without the resources of a large multinational, says Lotia. 

“There are mid-sized firms that have some level of sophistication and have the technology but have not fully embraced it yet,” says Lotia. “For example, they may be using the platform for their trades but are not using the analytics. But when the market starts moving unexpectedly and the board starts asking questions, then the treasurers begin to use the platform more. And once you do that, there’s no going back and it leads to a rapid increase in adoption. So while the market volatility has not been great for everyone, it has been good for the risk management community because it has highlighted the need for more strategic processes.”

And as AI capability continues to develop rapidly, there will be further benefits in terms of FX risk management, say Hipwell and Lotia. “It is less about forecasting currency movements prices and more about understanding the impact of those market movements and preparing for changes in the market rather than being reactive,” says Hipwell. “AI can really help in that way because it allows you to scale up the scenarios you’re looking at, to model multiple scenarios continuously. That’s been a huge shift from the spreadsheet-based approach where you can only really model one scenario at a time.”

Then there is also the crypto and stablecoin conversation which is increasingly relevant as market participants look for ways to mitigate unprecedented levels of volatility in FX risk. “The biggest barrier to crypto adoption at the moment is understanding the risk because these currencies do not follow standard patterns,” says Lotia. “The ability of AI to understand that risk and find those patterns and create an understanding and appetite for that risk, will be a big change. We are still at the starting point of this and there is more work to be done but that is where the real value of AI is, not predicting prices,”  she concludes.