New technology is challenging the pre-eminence of Excel in financial modelling and analytics
- New technologies target specific uses while Excel has a billion users with a broad range of needs
- Excel tasks will be done elsewhere such as in analytics platforms and cloud-based modelling tools
- Corporate analytics teams have reached the extremes of Excel and are using new technologies
By Ian Bennett FCA
Microsoft Excel turned 30 last year. This makes it older than half of my colleagues in PwC’s Deals Modelling team, a team for whom Excel is as clay is to a potter.
Excel was released the same year as the film Back to the Future. Perhaps the film’s director Robert Zemeckis wasn’t into spreadsheets, but he chose not to incorporate it into his vision of 2015 — or maybe the scenes with Hill Valley’s finance and accounting community ended up on the cutting room floor — but it’s fair to say that Excel has a bigger impact today than hoverboards or self-lacing boots.
Microsoft launched Excel in the same week as the first version of Windows and it now has over a billion users. That’s one in seven people on the planet. Which is quite a lot.
I’ve been financial modelling in Excel for more than 15 years and for the whole of that time spreadsheets and models were synonymous with Excel. But it’s not just the advanced financial models we build and review in our team. The role of Excel has broadened to every conceivable use.
Accountants have a unique perspective on Excel. Love it or hate it, it’s an important part of every accountant’s day. Have you ever stopped to consider what would happen if you woke up one day and it was gone?
Excel at the big end of town
About a decade ago we helped some of Europe’s biggest financial institutions write and finalise their spreadsheet, EUC (End User Computing) or UDA (User Developed Applications) policies. We wrote the policies with confidence that all key spreadsheets would be systemised within about five years and Excel would no longer be part of key business processes.
Five years later I was working with Australian banks doing the same thing. This time it was different. We knew the proliferation of spreadsheets was here to stay. With complete systemisation no longer an option, the strategy was just to keep track of the important spreadsheets and not worry about the breeding and cloning of the rest.
Over the past five years financial institutions and major corporates have enhanced, expanded and tightened their spreadsheet/UDA policies, with almost no reduction to the number of spreadsheet errors.
The permanent proliferation of Excel files in large organisations, and the pressure in the US from the Fed, has led to the development of major technological solutions like ClusterSeven to help us manage spreadsheets institutionally.
Some financial companies have a mind-blowing number of separate spreadsheets. Felienne Hermans, one of the world’s leading academic experts on institutional spreadsheet risk, tells the story of when she took her spreadsheet scanning software into a Dutch bank.
“They confidently told me they thought they had about 10,000 spreadsheets,” Hermans says.
“I set the tool running and within an hour it had already found one million sheets. In the end they had 2.5 million spreadsheets, and this was a 1,500 employee company.”
Because spreadsheets are created by every analyst in an organisation, the associated business processes are being consciously or unconsciously designed to cope with the design constraints, technical limitations and operational risks of Excel.
These limitations are so fundamental to organisations that advanced Excel or modelling training is actually less about understanding how to use the advanced features of Excel, but rather to help analysts best cope with the constraints Excel imposes.
Many organisations will continue to beef up their policies but there’s a new view that the unique risks presented by spreadsheets/UDAs can’t be managed by traditional controls frameworks. In response, operational risk teams are embracing financial model design standards like PwC’s 15 Model Design Best Practices and the FAST Standard (which PwC Australia is associated with) as a way to improve spreadsheets and lower spreadsheet risk.
Dead or alive?
Excel’s death has been heralded many times during my career — by a revolving door of vendors of alternative spreadsheet software packages, and by those who’ve spent all night fighting it to get something to a client and declare they’ll never use it again.
I always believed that reports of Excel’s death had been greatly exaggerated. But I am beginning to question that view.
But before we write it off, we must understand what it’s used for.
Excel needs to be all things to all people, and as Microsoft enhances one aspect, another user group is up in arms that their needs are being ignored. New technologies are emerging all the time which target specific Excel use cases and, just as it is in many other traditional industries, the pace of disruption is faster than it’s ever been.
Who uses Excel?
The ubiquity of Excel means that its billion users span a broad range of needs and expertise. At one end of the spectrum are those that use Excel to make lists or store data and at the other are people like the Microsoft Excel MVPs and the four individuals who can call themselves Financial Modelling World Champions.
Thousands of advanced financial modellers from all over the world enter a commercially-flavoured annual Mensa-style test called ModelOff. It exclusively uses Excel.
Will ModelOff participants in 2020 be using a different technology? ModelOff will undoubtedly use Excel as long as it is the tool of choice of the world’s leading financial analysts, but Excel will have to fight off increasingly stiff competition to retain this position.
New technology: Microsoft and Google
The first hint of changes to come was when Microsoft included Office Online, featuring Excel Online, in the Microsoft Office 365 experience. Excel Online is the cloud-based create/view/edit collaboration version of Excel. As long as you don’t need macros, Excel Online offers real-time editing of spreadsheets by multiple users.
This was a huge step forward and Excel Online has been very successful. But the cloud has had limited uptake by the business community, and it remains to be seen therefore if Google for Work will be more favourably embraced.
Google’s offering is 100 per cent online, with access via a browser only, but the key difference is that Google have aimed Sheets (their Excel equivalent) firmly at the everyday Excel user. They have intentionally not included enough functionality to satisfy the advanced user.
Here at PwC we adopted the Google collaboration suite in 2015. PwC’s Deals Modelling is using the Google suite every day, but currently only for collaboration not for any financial modelling itself. As far as I can tell, Google Sheets will not replace Excel completely.
New technology: The future of finance departments and financial modelling
In finance departments all over the world, the slow innovation by the large business planning players such as Cognos/TM1, BusinessObjects and Hyperion have resulted in a flight to Excel.
The gaps left are being filled by a new breed of cloud-based forecasting and reporting tools. And although Tagetik, Adaptive Insights and Anaplan have all positioned themselves against these business forecasting and reporting systems, in reality it’s Excel that they are mostly replacing.
Anaplan is intuitive enough to be called “Excel in the cloud” but it also brings a single version of the truth, with role-based access and real-time collaboration. It also lifts many of the natural constraints and limitations of Excel, such as changing a formula, and it automatically replicates across every dimension of the model.
When you become technology agnostic you design better models.
Excel is still used for the vast majority of our work, but for the first time we have a choice. For me, a tool like Anaplan is the future of financial modelling, but it’s not there yet.
New technology: analytics
The burgeoning corporate analytics teams, now commonplace in large organisations, are regularly reaching the extremes of Excel and have embraced a suite of new technologies to meet their need to utilise the vast reserves of organisational data.
One obvious example of this is Microsoft’s Power BI. As powerful as its name suggests, Power BI is now a stand-alone application. But it still doesn’t replace Excel or, for that matter, any of the other data sources that it draws from.
The same goes for the other leaders in visualisation or statistical analysis. Tableau, QlikView, R all embarrass Excel at things Excel was never really designed for and, as a result, will become more and more relevant to business. But every single analyst who uses those tools also relies on Excel.
In fact, it’s clear when you look at Microsoft’s recent announcements that Excel is being positioned as the front-end for finance professionals. Through Excel we will access big data, machine learning, predictive analytics and real-time data. Excel’s not retiring; it’s our cockpit.
New technology: the wish list
One area that I would welcome some disruption is my home of M&A financial modelling. These models are built quickly for important transactions for the use of a diversity of stakeholders such as advisors, lenders, management, equity investors and auditors. The models are large, inherently complex and technically advanced. They manage multiple scenarios and almost constantly changing term sheets, commercial driver relationships and business assumptions. Tens of versions are emailed around over the course of a deal.
What’s needed is a cloud-based single-source-of-the-truth model with the advanced functionality of Excel, that’s transparent and easily audited, that all verified users can immediately access with viewing and editing restrictions appropriate to their role in the transaction, from any device and location without the need for licences/contracts.
The technology doesn’t exist just yet, but is it too much to hope for?
Given the current pace of new technology, we should expect considerable disruption and innovation in the near future and things we’ve always done in Excel will be done elsewhere — especially in new analytics platforms and cloud-based modelling tools.
However, Excel will likely be with us for another 30 years, given its death would require a technology that replaces every use case of Excel, as well as being able to allow access to, or replace, the need for every one of the billions of spreadsheets created in the past. Perhaps only one company could do it … don’t count out Microsoft just yet.
Ian Bennett FCA is a deals modelling partner at PwC Australia.