We love Excel. There are 750 million installations globally, which would make it the third most spoken language globally. I have seen a financial planning and analysis (FP&A) analyst go as far as recreating a honeymoon beach picture in Excel by using the color fill feature on small cells, like a Georges Seurat painting.
So why did Forbes say that Excel “might be the most dangerous software on the planet”? And why is my inbox filled emails from EPM/CPM vendors telling me that I need to get out “Excel hell” before the next system crash? Clearly there is a discrepancy here.
Back to the Forbes article—on the asset side of the ledger is power that Excel gives to each end-user to create their own models and analyses. Forbes gives an example based on bank trading floors: “If the spreadsheet, or Excel, didn't exist, then a lot of what the financial markets do couldn't be done. There would be no collateralized debt obligations, (CDOs), no credit default swaps (CDS), indeed much of the complexity of the financial markets would simply disappear in a puff of smoke. For if you cannot model these things (however badly they are modeling them) then you simply could not be trading them as they are.”
On the liability side of the ledger is the very incident that led to this article—JP Morgan reviewed the how its processes broke down after the London Whale incident, in which a rogue trader cost the bank $6.2 billion dollars in losses and an additional $0.9 billion in penalties. The bank’s internal investigation identified the risk that tens of billions of dollars in trades were evaluated and modeled in offline spreadsheets that had multiple formula errors, and where data inputs were manually cut and pasted. There are numerous other examples of billion dollar errors caused by wrong signs, formula errors, or using old versions of spreadsheets.
Part of the discrepancy between the two views is that Excel is such a successful product and is ubiquitous in finance. It is a common language that allows models to be exchanged easily across teams, departments, and oceans; it also allows for people to walk easily from one job to another. It is flexible, accessible, versatile and powerful, and it allows for “end-user self-servicing” where everyone can design a model as needed without relying on an administrator or the IT department.
Because of these traits, spreadsheets are not treated with the same rigor as specialized applications. They frequently lack documentation and do not have the same level of validation, version control and backups are more complicated, and data governance is in the hands of each user.
Since we are living with spreadsheets, we must do it well. Understand the limitations of Excel so that you can minimize them—follow good data practices by creating data maps, establish single versions of the truth, and put change management practices in place for model updates. If you are going to use Excel as an enterprise tool, treat it as such. If the model will be shared, use appropriate collaboration features and security such as locking sheets or cells, limiting inputs through validations, and giving people read or write access through passwords or network controls.
At some point, when your company is large enough and the business model is stable, you need more capabilities and it will be time to cut over to more sophisticated specialized tools. Even then, you will probably maintain Excel for offline analyses, sandbox development, a front-end input source or an ad hoc analysis tool via that “Export to Excel” button. Be sure to use it wisely!
Bryan Lapidus, FP&A, is a contributing consultant and author to the Association for Financial Professionals. Reach him atBLapidus@AllegianceAG.com.
For additional insights on FP&A, subscribe to the AFP monthly newsletter, FP&A in Focus.