“IRS Data too stale?” I Say Absolutely Not True! by Dave Miles, CPA, CVA
In the business of valuation, experts have to make tough choices about data. Consciously or not, these decisions are usually based on the core tenants of sufficiency and relevancy. I am dismayed to report that most folks think IRS Corporate Ratio data stinks. “It’s too old” is a common refrain bantered between valuation experts and for seemingly good reason. Data from July 2017 to June 2018 was published almost 3 years later in 2021. For many valuators, they fear the showstopper question “Why aren’t you using the most recent data?” when considering using IRS data. There are some good and defendable reasons to use older data. This article provides some thoughts, observations, and insights on using older data in the context of the sufficient and relevant motif.
Beginning in 1916, congress mandated the annual publication of statistics related to the “operation of the internal revenue laws.” Over the past thirty years, ValuSource has been acquiring, cleaning, and processing that data about form 1120. (Yes, you could get it for free, but good luck processing that data in a reasonable amount of time). After 2017 when the Corporation Source Book was replaced by the Corporate Income Tax Returns Complete report, data is now only represented by either size or industry, not both. ValuSource acquires the data and then creates ratios and common-size statements. It is interesting to note that a number of providers of benchmarking data use this same IRS data as one of their sources. The take-home message is that there is a lot of credibility, accuracy, and transparency in IRS data.
Every expert must make a judgment as to the sufficiency and relevancy of data used in an engagement. That judgment is paramount. Facts and circumstances, the context of the engagement inexorably drive that judgment. ValuSource sells three different financial ratios and benchmarking databases to the industry: RMA, IRS, and Bizminer. Each of these databases has strengths and weaknesses and none of these are the one fits all choice. No chef would carve a chicken with a paring knife – and just the same way the tools of a chef change based on the situation – the best data set to use in a particular instance should be based on the fact pattern of the engagement. Based on the fact pattern one database may be better than the others. Here are my reasons that the IRS may be a better choice.
All other things being equal, the more current data is the better it is. However, other things are rarely equal. I call this the currency fallacy. Just because data is newer does not usually mean it is better. If you follow Twitter during a historic event, the data is almost instant. A picture will come through the feed, but can you say that it is accurate, it has the right context, or that it is unbiased? Newer data is not necessarily better data. The point of financial ratio analysis and benchmarking to industry norms is to see if it’s strong, average, or weak compared to the industry. Let’s dig down on the IRS data.
How long does it take to publish? RMA data is collected for twelve-month periods ending between April 1 previous year and March 31 current year and published in November. IRS data is collected from July 1 of four years ago to June 30 of three years ago and published in June. A public corporation can take three months to publish its data. For sources of benchmarking data that use multiple sources of data and combine them, the concept of “collected date” becomes very fuzzy. On the surface, the collected date is what a lot of valuators rely on to discount the IRS Corporate Ratio Data. Is data from over a year ago better or worse than data from three years ago? What happened in those 24 months?
Under the hood, how different is today from three years ago with respect to financial analysis? Are gross margins different? Are current assets as a percentage of assets different? Is the return on equity different? If you have done your industry analysis correctly and can see there are no great changes, then ratios three years ago might a good comparison. I would also add the subjective argument that it is the ability of the business to adapt to change that makes it valuable and not simply that it is strong in comparison to an industry.
What else is not equal? A variety of things can impact confidence in a data set to accurately represent a population, the biggest factor happens to be sample size. With a larger sample, a larger bucket, one outlier, one drop of poison, matters less and less as the sample size or bucket grows more and more. There were over 12 million tax forms used in the IRS data published in 2021. Such an unprecedentedly large sample size will represent the entirety of the industry and size better. Other data sources with a much smaller set of data might not paint as complete a picture.
What else is not equal? The ability to replicate. The IRS Corporate Ratio Data is transparent. Statisticians publish the files of average tax form 1120. You can replicate the ratios easily. Replication adds to credibility. Other databases, with proprietary and undisclosed processes, show only the end ratios. It should be noted that the IRS Corporate Ratios data is based on actual data and as a result, data may be missing or left empty.
What else is equal. Any data used to benchmark will have a great effect on the value of the company. Choosing a database that is complete, accurate, transparent, and replicable is important. The strength of the IRS data is also its weakness. It is based on tax returns. Do tax returns have enough information to calculate some ratios? ROE for instance. Other Issues include whether the return was accrual-based or not. Items may be missing on the return. The next weakness is that the format may not let you get granular; it is size OR industry. Some nuances are lost. Sometimes being small in an industry is different than being large in an industry.
To summarize, all data has strengths and weaknesses. There is no bad data, only bad uses for the data. Experts need to understand those virtues, or follies and choose data that fits a specific fact pattern so that they can use the data in an honest way. Experts must choose the most sufficient and relevant data for a particular engagement, and no two engagements are alike. Choosing to underestimate IRS Corporate Ratio data simply because of a “stale data” theory is a mistake as there are many advantages to using this data and, after reading this article, defending another benchmarking database may actually be more difficult than defending ValuSource’s IRS Corporate Ratios.