An essential tool
Excel’s ease of use, combined with the numerous combinations of analytical computations has resulted in an almost unlimited capacity to conduct quantitative analysis. Over the past 30 years traders have adopted Excel as the go-to analytical tool in every area of FX, whether it is to calculate prices, create a complex financial model or a VAR calculation. For many users of Excel, its flexibility and convenience has enabled them to develop customised reports quickly and without needing to consult with IT departments.
The problem with Excel
Fundamentally, although users expect Excel to act like a database, it simply cannot manage the large datasets that today’s electronically traded markets require.
The strength of Excel is its flexibility, but this can also be its downfall, especially when used by financial firms trading and managing billions of dollars of risk.
Excel spreadsheets are decentralised, with all the logic, risk and management often dependent on the person that created them, and this effect is compounded by the increasing complexity of pricing models. If the trader who created a specific pricing spreadsheet leaves the firm what happens? Do firms continue to use a spreadsheet which is not fully understood, or do they create a new spreadsheet, which takes time and resources?
Excel allows firms to perform complex calculations across connected spreadsheets, but with figures being regularly added (often manually) and formulas edited, there are concerns about the integrity of data and reports. This is compounded by Excel connecting to multiple spreadsheets, some of which may be changed without the knowledge of downstream users.
The lack of internal usage policies and change control mechanisms often means that Excel’s use for critical business processes goes unmonitored. Excel often bypasses IT departments, but its extensive use and its embedded nature make this an enterprise-wide challenge.
FX workflow automation
For trading businesses, spreadsheets may be viable for small numbers of transactions or in markets with a reduced need for automatic pricing. But as trade volumes increase and FX markets become more automated, spreadsheets are no longer a viable solution and may become more of a hindrance, with calculations taking longer in the face of bigger transaction amounts and growing transaction data.
As FX workflows become more automated throughout the trade lifecycle, more data than ever before is being captured, analysed and acted upon. Analysing large datasets to gain market insight in search of better trading results means that analytical systems must be capable of managing vast amounts of data.
When we talk to traders about pricing FX Swaps, many of DIGITEC’s clients understand that the market is becoming more electronic, where update frequency of price data will ultimately evolve to the point of using tick data. Systems need to be robust to manage this volume of data. In addition, new pricing models for FX Swaps need to be developed which have less reliance on broker feeds, meaning that pricing tools require more computing power and can no longer rely on Excel.
The need to replace Excel
As financial firms face increased regulatory oversight, Compliance teams are demanding that Excel is replaced by applications which support data controls, specified workflows and approvals, and change management processes to ensure complete visibility and transparency across all functions, fully supported by an audit trail.
Excel has insufficient governance and control, with difficult to understand workflows across different teams, systems, and locations. There is a lack of transparency and auditability, and Excel limits the availability of timely information and makes real-time analysis more difficult, hence impacting decision-making.
Spreadsheets result in human keying error and can heighten security concerns and hide control issues. Security features are lacking, meaning that even if spreadsheets are password protected they are relatively easy for hackers to access.
Remember the “London Whale”?
The use of Excel in Financial Markets has led to some high-profile losses.
Perhaps the most famous was J.P. Morgan’s “London Whale”, the trader who lost $6.2 billion in 2012 in CDS positions.
After the event, J.P. Morgan’s internal investigation identified issues related to a VaR model, which used several Excel spreadsheets. These were updated by copying and pasting data from one spreadsheet to another, introducing the potential for keying error. The most concerning finding, however, was that the VaR spreadsheets contained a significant error, where instead of dividing by the average difference in prices, the model instead used the sum of the differences. This reduced volatility by 50% and resulted in a significantly lowered VaR.
Other examples include Goldman Sachs using a spreadsheet that erroneously overstated Tibco’s share count in 2014, which led to a miscalculation of Tibco’s equity value, and a $100 million loss.
Also, in 2002 John Rusnak, a trader at Allfirst Bank, manipulated spreadsheet models used by the bank’s internal control staff, which ultimately cost the bank $700 million in losses.
The right time for change
The problem for FX departments is that Excel is deeply ingrained in their infrastructure (and in the infrastructure of the whole FX industry), and any attempt to move away from spreadsheets is likely to cause major upheaval.
However, as FX workflows become increasingly automated, reducing manual errors and increasing efficiency, there is no room for applications like Excel. We regularly see clients replace Excel with DIGITEC’s D3, a centralised application for all pricing, which can be accessed by users who can work in parallel, across desks or in different timezones. This robust solution enables simple and fast auditing and compliance, and supports the downstream distribution of system outputs and market data.
For firms requiring a more bespoke solution, there is a need for No Code applications, like D3 Curves, a new module and extension of the centralised D3 system. With D3 Curves, clients are able to build their own more complex models and curves that can lie behind the D3 pricing engine, without needing a Quant or having the ability to code.
In the past there was the argument that using centralised applications and No Code modules were too expensive for anything other than core products, but the recent adoption of applications deployed on the Cloud makes these robust systems more affordable and accessible for many more firms – meaning that at last there is a credible (and superior) alternative to using Excel in FX.
