IT Faculty issues draft Excel good practice code

Vincent Colin/iStock/Thinkstock

The ICAEW IT Faculty has published for public comment and debate a manifesto containing 20 principles for good Excel practice.

For several months, IT Faculty technical and development manager Paul Booth compiled suggestions and moderated debates between a dozen or so spreadsheet experts to come up with the outline 20 principles document.

But that was only a starting point. The faculty is keen to promote a wider debate around the document to strengthen its core message and broaden the adoption of the principles within the profession. The principles are listed below and the full document available for download to any interested party, so AccountingWEB asked Booth to explain what the project is trying to achieve and how far along that road it has progressed...

Continued...

» Register now

The full article is available to registered AccountingWEB members only. To read the rest of this article you’ll need to login or register.

Registration is FREE and allows you to view all content, ask questions, comment and much more.

Comments

Barclays accidentally bought more of Lehman assets than they int

Cantona1 | | Permalink

That is hilarious!

I think we should start from the very basic premise:

A spread sheet, in it is very basic form is not more than a scientific calculator.

If I want to get a result of 75, it is 100 minus 25, but if I mis type one of the inputs, I get a wrong result. If I think my homework is very important, I will ask some else to double check my work.

I wonder how many people who worked on Barclays’ spread sheet checked the calculation before it was sent to the decision makers.

Paul Booth's picture

If it's important, ask someone to check my work

Paul Booth | | Permalink

Indeed so! That's Principle No 4: 'Work collaboratively, share ownership, peer review'.

Unfortunately there are many more ways of causing spreadsheet errors besides mis-keying a figure. That Barclays error was (with the benefit of hindsight, and for those of us looking on from the outside) ridiculously simple, but it was no doubt part of a large, sophisticated spreadsheet.

London Insurance Company

reggiep | | Permalink

I produced a detailed risk reporting workbook for them. Someone linked to it with a pivot table and used the result to produce some of the data for their so-called "board pack". Of course the number of rows expanded somewhat, but the range this fool used did not. Nobody noticed until he asked me about it, and he never told anyone. I had been paid long before, so nobody lost (?).

How any more examples would you like? I have plenty.

listerramjet's picture

a topic worthy of debate!

listerramjet | | Permalink

I agree with Cantona1 - and I would postulate that 99% of spreadsheets are no more than scratchpads for calculations, letters, doodles, or other day to day stuff.  And unlikely to require any consideration of spreadsheet risk whatsoever, beyond the normal human condition of idiots doing stuff for which they are not qualified - or perhaps just having a bad day.

But for the ones that don't the question is, what does 20 principles offer that is actually useful?  I would suggest not much. After all - Don't see the institute offering 20 principles for auditors, or tax specialists.  There is a presumption of competence, which surely should apply equally to accountants using spreadsheets.  And equally there is a requirement on accountants to only undertake assignments for which they are suitably qualified.  Surely woe betide any accountant producing a spreadsheet for a financial forecast if they don't have the necessary spreadsheet skills?

Having said that, I do welcome the debate.  I do think the list of 20 is a mixture of policy and systems development best practice, and I think it would be better so split into two accordingly.  I also don't think spreadsheet design is something that lends itself to "best principles".  If you read these and learn something new, then you should probably not be designing spreadsheets!  But development principles - such as get (SMART) requirements, test, Keep production and development separate, etc are usefully presented like this.

A Condundrum

Fastlane | | Permalink

The spreadsheet is such a versatile tool and used for a myriad of applications (I often liken spreadsheets to a whiteboard or artists canvas - a medium upon which to create whatever you like) that it is difficult to be prescriptive for all situations. Counter-intuitively, it is easier (more appropriate?) to be prescriptive about multi-sheet/multi-component structured financial models(e.g. capital project evaluation) than it is for a single sheet spreadsheet. Sure some principles/standards will be appropriate for both applications, but many will not suit the smaller one. However, as spreadsheets are now ubiquitous in most business organisations, and having seen most users learn from the person sitting at the desk next to them, I do support the development of a set of standards and/or documenting of best practice - but in doing so there has to be a recognition that there are horses for courses.