Asset management, just like many other businesses within the financial sector, runs on the bedrock of the Excel spreadsheet. It’s been working that way for years — ever since Microsoft first introduced the program in 1987 to challenge the dominance of Lotus 1–2–3. Excel spreadsheets are now ubiquitous, and asset managers almost universally rely on them to store the vast majority of their data. But as the amount of newly generated data continues to grow at a staggering rate, so too does the risk that an ever increasing number of these spreadsheets will fail due to data mismanagement and formatting errors.
They will fail because the Excel spreadsheet is not an effective data storage platform. The biggest challenge I faced when working with asset managers was that I always spent far more time verifying critical data than I ever spent analyzing it. That’s because it was almost impossible to trace the source of any given piece of data. And if that data had been changed from its original form, there was no log explaining why, or by whom. These problems were compounded by the fact that there was no easy way to access the data in one spreadsheet from another. In short, there was no sure-fire way of trusting the data.
This is not to say that Excel is a bad program. As James Kwak has pointedly stated, Microsoft’s Excel is certainly one of the most powerful and arguably one of the most important software applications of all time. Nevertheless, the program does have its limits: It can easily get you into trouble if you use it in the wrong way, sometimes with catastrophic results. Here are just three examples:
1: In their seminal 2010 paper, Harvard professors Reinhart and Rogoff claimed that when a country’s debt hits 90% of GDP, that county’s economy slowed dramatically. However, serious doubts grew over the accuracy of their findings when a graduate student discovered some anomalies in the dataset used: A serious coding error within their Excel files meant that the top five economies in the world had been omitted from their calculations.
2: JP Morgan’s widely reported and analyzed London Whale fiasco resulted in a $3.1 Billion dollar loss due to VaR inaccuracies — the cause of which was simply the wrong sign entered within a formula on one of their Excel spreadsheets.
3: MF Global, a derivatives powerhouse, went bust due to a series of perceived liquidity problems. Among the details to emerge from the ensuing postmortem were the discovery of critical errors in some of their Excel spreadsheets. As a result, many of their models for risk management had not been updated with the latest data.
These are three particularly dramatic examples, but these kind of fundamental mistakes are not at all rare. A recent study identified that 88% of all complex Excel spreadsheets have serious errors. In the past, I’d always assumed that the Excel spreadsheets I was given to work on were accurate, until I was proved wrong.
The simple truth is that Excel was never designed to be used as a robust data storage and management platform. In addition to the deficiencies I’ve outlined above (no way of tracking data sources, monitor changes, users etc) it’s almost impossible to quickly verify the accuracy of any of the data — which is a particularly big problem if you’re working with an asset management firm.
Sophisticated data management can bring incredible benefits to today’s asset managers, but only if every function of the asset management system uses the same validated data, thereby creating one single source of truth which delivers the most up-to-date data for analysis, and increasing transparency and credibility for the investors.
In seeking to build a strong investor base, Asset Managers need to have serious discussions about not only what data is relevant, but also how they can they can best store and access it. Just throwing everything into an Excel spreadsheet and hoping for the best is not the answer.