On statistics, uncertainty and spreadsheets ...

Oil IT Journal editor Neil McNaughton provides a heads-up on new functionality on the www.oilit.com website. Inspired by Patrick Leach’s book reviewed in this issue of Oil IT Journal, he returns to some earlier reflections on statistics and uncertainty—noting en passant that current spreadsheet usage, even in a compliance/reporting context, often falls way short of established industry best practice.

We have been working away on the oilit.com website over the last couple of weeks. Users of either or both of our services (the Oil IT Journal site license and our Technology Watch reporting service) coming into the oilit.com site can now search and see all of the content they have subscribed to. To see if things are working for you, please visit www.oilit.com. If you are a site licensee you should automatically see the full text of the current issue of the Journal. If you also have a Technology Watch subscription, a click on the ‘TechWatch’ link at the top of the page will bring up a full list reports going back to 2002—all of which can be viewed or downloaded. If this doesn’t work for you, please email us at info@oilit.com and we will fix up your access.


A word or two on the new search functionality. One of the curious things about Google is that you can’t organize your search results by date (without paying for the Enterprise Edition, that is). One of the questions that I frequently ask myself is ‘did we publish this story last month?’ So it was mostly in self interest, and with help from our search technology provider, FreeFind, that I retrofitted HTML ‘Date’ metadata tags to our articles so they can now be ordered by date. FreeFind also indexes our PDF documents, something the old system failed to do I am ashamed to say. A user of the updated site said, ‘The new system is fantastic.’


Rather longer ago that I would readily admit, I got caught up reading a book—you know the kind that you can’t put down and may even carry on reading through the night. That’s not so unusual. What was strange was that this was a book about statistics! Doing some mindless Googling while preparing this edition I came across the Amazon page for my erstwhile potboiler and found that my passion for the oeuvre was shared by others. The book in question is ‘Facts from Figures’ by one M.J. Moroney*.


Moroney describes the analytical approach to statistics. Writing before computers were around, the only way to address such problems was through math with detailed analysis of the interplay of variables and errors. For more on this you might like to revisit my September 2002 editorial ‘Lies damn lies and futzing’ (OITJ Vol. 7 N° 9).


Patrick Leach, in his excellent book reviewed on page 4, ‘Why can’t you just give me the number**?’ drums home the point that uncertainty is the source of value in business. I came across some amusing support for this fundamental truth in the Financial Times Fund Management supplement—not exactly my favorite reading, but a piece titled ‘No free lunch in downside protection***’ caught my eye. FT columnist Steve Johnson reports on products that purport to offer the upside potential of equity investments with some from of protection from downside risk.


A Barclays study**** quoted by the FT concluded that ‘attempts to curtail the volatility of equities by options hedging are likely to eliminate excess equity returns over time, suggesting the investor would have been better off in less volatile assets.’ In other words, as everybody knows, there is no such thing as a free lunch. Personally I have always steered clear of these products as a) I suspected a trick and b) I have no money. All this would be rather anecdotal were it not for the fact that, according to the FT article, such products are ‘common in pension funds across Europe.’ With long term returns from the ‘free lunch’ products coming in at around 1%, it makes you wonder...

Best practices

More researching on the current issue brought incontrovertible evidence that at least one major oil and gas company uses spreadsheets in what one could call mission critical planning and compliance reporting. Nothing unusual there. Unfortunately, there is probably nothing unusual either in the fact that the usage was demonstrably poor in terms of best practices. If you want to know what spreadsheet best practices are, I would refer you to our report from the 2005 Crystal Ball (CB) Oil and Gas User Group (OITJ Vol. 10 N° 10). Here, Decisioneering’s Steve Hoye reported that some 40% of CB trainees don’t know about the elementary best practice of using range names rather than cell references. There is a delicious irony in the fact that large companies are misusing spreadsheets in the context of compliance reporting!


But enough of the cynicism! Supposing you actually want to do something about your organization’s dodgy spreadsheet practices? Well I came across what appears to be just what you should be looking for in some marketing material from PricewaterhouseCoopers’ (PwC) as follows—‘The use of spreadsheets creates a high risk of error in the data reporting chain. PwC’s tools appraise the effectiveness and reliability of spreadsheets. For many companies, the finance department has an interest in well controlled spreadsheets.’ Indeed!


Rich tapestry!

Way back when I started editing PDM, OITJ’s forerunner, folks warned me—and I half believed them; that the oil and gas IT scene was all but carved up between Landmark and Schlumberger and that soon, there would be nothing left to write about. The truth is that we seem to be moving further and further from a software duopoly. In this edition alone, there are 15 companies that are new to me—and they are not fringe footlers, but already have clients and serious offerings. Oil and gas IT—a rich tapestry!

* Penguin 1952, ISBN 0140202366.

** Probabilistic Publishing 2006, ISBN 0964793857, www.decisions-books.com.

*** Financial Times February 19th 2007.

**** Barclays 2007 Equity Gilt Study.

This article originally appeared in Oil IT Journal 2007 Issue # 2.

For more information or to comment on this topic email here.