Wednesday, April 1, 2015

Fermi Calculations for XBRL

CompSci Resources, LLC

Washington, DC – Recently, our team of engineers has been working on new ways to analyze and make eXtensible Business Markup Language (XBRL) even more accessible and useful for the common investor. During these brainstorming sessions, it occurred to us that the standard has a rather limiting attribute that can be expanded to vastly improve XBRL’s accessibility.

Currently, the XBRL standard itself permits precision to be specified in two ways: with significant figures, and with decimals. The U.S. Securities and Exchange Commission (SEC) requires the use of decimals since this is the most intuitive of the two, and the most commonplace in practice.

To open the door to a more expansive interpretation of standard financial statements, a new way of specifying precision for XBRL Calculations is proposed here. This method is one that is likely to resonate with financial analysts, engineers, and accountants alike: Fermi.

This approach would enable a more accurate representation of the precision required of XBRL itself. This would also allow financial analysts a much broader range of reporting precision from which they can draw a variety of new conclusions and recommendations.

To better illustrate the idea, please review two drafts of the balance sheet for a theoretical public filer, Bioguys, LLP →

So while it might be more natural to see that everything in column A foots, we are pleased to highlight that everything in column B Fermis.

Nathan Summers was the lead developer on this project, so we wanted to give him an opportunity to shed a bit more light on Fermi Calculations. He noted that, “While the technical details behind this proposal are arcane, the general idea is something so simple that even a rocket scientist could understand it: the calculation is considered consistent as long as it's within an order of magnitude or so of what you would expect.”

As far as finding a practical application for Fermi approximation within XBRL, Summers added, “Well actually, this approach appears to be well-suited for financial reporting … although we may be well over a decade late for WorldCom and Enron. It is likely to be welcomed by inattentive billionaires as well.”

There are some limitations to introducing this approach, though. For instance, we do not further recommend employing Fermi to approximate the existence of creative accountants, lest we introduce yet another paradox.

What is your preferred approach to representing precision in XBRL? Would Fermi be a preferable option to decimal and significant figures?

Anyway ... we hope that you enjoyed this publication in honor of this first of April, two thousand and fifteen. For further reading, here is our link to our blog from one year ago.

Also, if you'd like to read a serious, very good publication on XBRL Calculations, read "A Calculated Risk", by CompSci's Shawn Rush. Have a great day!

Fun Facts About this post:
  • Authors’ Combined Age: 10
  • Number of Authors: 1
  • Number of Words: 100
  • Number of Terrible Jokes in this Post (incl. this one): 10
  • Authorship + Editing Time: 1 hr
  • Number of Days Until April Fools’ Day 2016: 100
  • Number of Terrible Jokes in this Post (incl. this one): Still 10

Nathan Summers serves on the XBRL Standards Working Group and is a CompSci software engineer. He also authored a popular article explaining XBRL Negation and is a co-author of patented XBRL technology.