Spreadsheets and the Emperor’s New Clothes!

The article explains how human errors in designing and operating spreadsheets can have serious financial and reputational consequences. Enhancements to ERP systems and shared service centres are enabling organisations to rationalise and control spreadsheet usage.

Spreadsheet applications remain one of the most common management tools in modern business. They have persisted as the de facto ‘system’ for executive decision making, long after the multi-million dollar ERP systems promised to deliver a ‘single version of the truth’. 
In comparison to corporation-wide information systems, spreadsheets still offer significant flexibility to those business partners that prepare information for individual divisional managers. However, the flipside to flexibility and individuality can be anarchy; and with it the possibility of serious, possibly terminal, damage to the organisation.  In this article, Ian Herbert of Loughborough University explains how the humble spreadsheet is now attracting scrutiny from both the IT department and the C-suite.
‘Every other day I’ve probably heard someone say “Oh, that number’s wrong!” she added, “if you have these errors that go undetected, they could be life threatening to the organisation!”’
CFO on spreadsheets  

Some well-known problems with spreadsheets

Until very recently, few people felt confident in challenging the spreadsheet generated information that they receive. Prowess in doing clever things with spreadsheets was seen as a key skill for up-and-coming managers. Casting doubt on the reliability and the validity of the results is difficult due to the sheer scale and complexity of many spreadsheets, and in any case can appear uncollegial, akin to asking questions about the emperor’s new clothes! 
In our research into shared services, managers were keen to talk about the increasing risks that might arise from an overreliance on spreadsheets, both in terms of their design and operation. A common concern is that spreadsheets skills are usually self-taught and thus, users are not likely to be steeped in the behaviours and protocols embedded in those professionals who have trained in an IT environment. Even when an individual spreadsheet is audited and passed as ‘logically correct’, its input data may be derived from a whole series of interlinked spreadsheets further down the information ‘food chain’. In the past few years, a number of high profile errors are starting to make the ‘clothes’ of the office spreadsheet guru to look a little transparent. For example;
JP Morgan Chase (JPM) found a formula error in a spreadsheet that had undervalued risk by a factor of 50%, contributing to a total trading loss of $6.2 billion in 2012. The problem turned out to be quite simple, but resulted in the dismissal of traders, and top executives being hauled in front of the US Congress. Even more damaging, investigations were launched by several regulatory bodies, including the FBI. In short, the spreadsheet-based model used to monitor the traders’ risk levels was essentially a manual process which involved copying and pasting data from one spreadsheet to another. As Gandel (2013) noted; 
“One key measure was added when it should have been averaged.” [This meant that] “…risk officers at JPMorgan believed the credit derivatives bets were half as risky as they actually were.”. 
The use of spreadsheets owned and operated by individual traders for such a critical activity appears to be a sharp contrast with the $2 billion JPM spent on corporate IT systems in 2012 (King, 2012). 
Harvard Professors, Kenneth Rogoff and Carmen Reinhart’s thesis "Growth in Time of Debt" was hailed as a key influence in the development of austerity programmes of several stricken governments following the 2008 financial crisis. The authors had argued that countries with more than 90% debt to GDP are unlikely to be able to grow and that public debt should be reduced through austerity measures. However, their spreadsheet model had excluded five of the countries in the overall data set that did not support that conclusion. The coding error was found by student Thomas Herndon of the University of Massachusetts Amherst. But, this was not just an embarrassment for the researchers, their mistake might yet hold serious consequences for the economic policies of several countries. More recently, similar concerns about spreadsheet analysis have been raised about the work of economist Thomas Picketty (see Giles, 2014).
British Nuclear Fuels Limited operates in one of the most rigorously controlled environments. In 2001, the Sellafield plant in Cumbria UK started producing Mixed Oxide fuel pellets, as a source of fuel for nuclear reactors. But, even before the plant had reached full production, controllers found that engineers had falsified safety records about the precise size of the pellets being shipped. Apparently, a single set of results had been ‘copied and pasted’ into other spreadsheets, resulting in the quality approval of what might be incorrectly sized pellets; potentially causing a nuclear meltdown similar to that at Fukushima, Japan (Thorn, 2012). A defective consignment had been sent to Japan and this had to be returned to the UK with significant financial and reputational cost. Gaia Hoerdner of the Centre of Nuclear Information in Tokyo was quoted by as saying: 
"The Japanese companies have totally lost face because they received this defective quality-controlled fuel and it is very unlikely that they are going to resume trade with BNFL (British Nuclear Fuels Limited) again."  (Thorn, 2012)

The end of the desktop revolution?


If leading experts can make such errors, what problems might be hiding in ‘everyday’ spreadsheets?
Some managers in our survey questioned the extent to which the desk-top revolution has been just that, a ‘revolution’ that has now come full circle. After years of decentralisation it would appear that there is a return to central control over the design and operation of end-user tools such as spreadsheets.
‘I ask people if they generally trust spreadsheet data. The answer is usually “yes” and I tell them it shouldn’t be’. - Finance manager 
In our research project, managers in shared service centres (SSC) were keen to emphasise the role of SSCs in consolidating and then standardising corporate information systems. A part of this transformation is to identify critical spreadsheets in the business units, then test the formulae and validate the data sources before ‘locking’ down the formulae. In this way, spreadsheets are still available, but in a more controlled manner. In essence, we are seeing a return to the days when business analysts and system programmers operated as a gateway to mainframe based computers. Some SSC managers are going further by striving to eliminate the vast majority of routine spreadsheets through the increasing availability of proprietary analytical packages which ‘sit on top’ of ERP systems and supply more standardised data which executive users can further interrogate and reconfigure. 

Change Drivers


Several managers explained how higher standards of corporate governance now require organisations to identify their spreadsheets by user and assess the business risk inherent in each document. Some cited the Sarbanes-Oxley act as a significant catalyst whilst others argued that spreadsheets are a natural target for attention because their dispersed nature makes them difficult to control which goes against the need for ‘one version of the truth’. 
With back office activities being pooled within an SSC managers can more readily identify where there is duplication of spreadsheet activities, either through two people doing the same thing out of ignorance or more cynically operating shadow systems to support political posturing. As one company director put it;
 “It’s too easy to create digital cottage industries where everyone is doing the same thing.” 
As we enter a new era of big data, smart organisations will spot patterns and trends in data only through sophisticated algorithms and huge processing power that will be beyond the power of the desktop machine for the foreseeable future. 


The findings of the Loughborough-CIMA SSC research project do not suggest that spreadsheets will be disappearing anytime soon. Nevertheless, many organisations are adopting a more intelligent approach to understanding how spreadsheets are used in the business and how a more holistic approach to data management and information production can provide better, less risky, insight for decision making. Indeed, we should not forget that whilst some processes are being recentralised, smart phones are driving a new era of real-time information and outputs and a totally individual level!  
About the author
Ian Herbert is Deputy Director of The Centre for Global Sourcing and Services in the School of Business and economics at Loughborough University. As Director the School’s programmes in Singapore he is a frequent visitor and is happy to discuss his research with interested organisations. 
The SSC research project is supported by the General Charitable Trust of the Chartered Institute of Management Accountants. 
Visit the project website at www.shared-services-research.com and tell us your experiences at ssc-research-team@lboro.ac.uk and ask for our latest report on spreadsheets.
Thanks to Glynn Lowth, Eliot Buckner, Chris Warner, Kathleen McLoughlin and Keshav Champawat.
References and further reading
Alexander, R., 2013. Reinhart, Rogoff and Herdon: The student who caught out the profs. BBC News, [Online] 20 April. Available at: [Accessed 18 July 2013]
Anonymous, 2000. Japan vents fury on BNFL. BBC News, [Online] 18 February. Available at: [Accessed 15 July 2013]
Brady, B., 2013. Revealed: £2bn cost of failed Sellafield plant. The Independent, [Online] 9 June. Available at: [Accessed 15 July 2013]
Gandel, S., 2013. Damn Excel! How the ‘most important software application of all time’ is ruining the world. CNN Money, [Online] 17 April. Available at: [Accessed 18 July 2013]
Giles, C. (2014) Piketty findings undercut by errors, Financial Times, May 23, 
http://www.ft.com/cms/s/2/e1f343ca-e281-11e3-89fd-00144feabdc0.html#axzz3GTdWVKF4 [Accessed 18 October, 2014]
King, L., 2012. JP Morgan overhauls global IT with Agile development and automation. Computerworld UK, [Online] 6 March. Available at: [Accessed 24 July 2013]
Stamp, G., 2013. Financial modeling - is the government wasting millions?. BBC News, [Online] 21 February. Available at: [Accessed 24 July 2013]
Thorn, S., 2012. The Misuse of Spreadsheets in the Nuclear Fuel Industry: The Falsification of Safety Critical Nuclear Fuel Data Using Spreadsheets at British Nuclear Fuels Limited (BNFL). System Science (HICSS), 45, pp.4633-4640.