Data on the Balance Sheet

Data is unquestionably playing a larger role in driving business value; a May 2017 Economist article asserts:

Data are to this century what oil was to the last one: a driver of growth and change. Flows of data have created new infrastructure, new businesses, new monopolies, new politics and—crucially—new economics.

This applies not only to tech firms but to non-tech companies as well. ‘Digital transformation’ is the watchword of the moment as companies of all stripes work to become more agile and technologically adept. Not only are an increasing number of companies working to become more data-driven, but more and more areas of the business are working to incorporate data into their decision-making and business processes.

If we can agree that data is an increasingly important driver of business success, should a company’s data be included as an official financial asset? My colleagues have argued as much.

“We’re going to see datasets increasingly recognized as a serious, balance sheet-worthy asset.”
– Steve O’Grady, 2010

“Data is clearly still not a well understood or significant investment category – brand “goodwill” is better accounted for, but there is no doubt that markets value companies perceived to be data rich with higher evaluations than other companies. Data is a moat.”
– James Governor, 2017

However, from an accounting perspective is including data on the balance sheet feasible?

General Principles

Before we dive in, it might be helpful to first step back and review the goals of financial statements. Accounting standards (like U.S. Generally Accepted Accounting Principles or International Financial Reporting Standards) are designed to help companies report financial metrics using the same basis. While not all companies follow these standards, they are intended to provide:

1. relevance: financial statements should provide timely, material information that provides either feedback about past operations or predictive value for future operations
2. reliability: measurements on financial statements should be verifiable and replicable by independent third parties
3. comparability: financial statements from different enterprises should report similar types of information in a similar manner
4. consistency: companies should apply similar methodologies to their reporting over time

Balance sheets in particular are about illuminating what the company’s owns and owes at a specific point in time. The assets side of the balance sheet ostensibly shows the tools the company has at its disposal to run their business, and the liabilities/owners’ equity side shows how the company has financed these assets. That ‘ostensibly’ above is important, so let’s dive a bit more into the term ‘asset.’

Colloquially we use ‘asset’ to mean something that is useful, as in “she is an asset to the team” or “that skill will be an asset to your career.” In the context of this usage, data is absolutely an asset to a company. However, the meaning from an accounting perspective is more specific.

In accounting, an asset reflects future economic benefits that are expected to result from past transactions or events.* The asset category of interest to this discussion is intangible assets. Intangible assets do not exist physically (as a building or inventory do) but are also not financial instruments (like stocks). Some common examples of intangibles are:
– trademarks
– copyrights
– license agreements
– patents
– customer lists
– software development
– goodwill

At this high level, a company’s data seems to fit the definition of an intangible asset. As stated above, we expect the accumulation of data (past events) to be powerful drivers of new technologies and competitive advantages (future economic benefits). However, there are some significant constraints to recognizing intangible assets under current accounting rules that impede our ability to recognize data as an asset.

Current Limitations of the Balance Sheet

Can all things that drive value for the company be reflected as official assets? Put another way, do all the assets a company owns actually end up reflected in their balance sheet? Unfortunately not, especially in the area of intangible assets.

One of the biggest limitations of the balance sheet is that it “omits many items that are of financial value to the business that cannot be recorded objectively.”* This includes things like employee knowledge and experience (apologies to Mr. Hamilton), company reputation, and perhaps most relevant to this discussion, internally-created intangibles. Accounting firm PwC explains, “in the US, investment in intangible assets has surpassed investment in tangible assets since the late 1990’s. However, if those intangible assets are generated within an ongoing business, they generally do not appear on a company’s balance sheet.”

In terms of the accounting principles above, there are trade-offs between what may be relevant to a company’s performance and what can be reliably, comparably, and consistently measured. Historically these tradeoffs fall conservatively in favor of prudence, meaning that accounting standards as they are currently written exclude most of a company’s internally-created intangible assets. This means that, again per PwC, “some companies’ most vital assets do not appear on their balance sheets.”

There are rules and guidelines that impact how and when a company can declare an asset. Here are some of the rules as they currently stand that would make recognizing data as an asset difficult.

Balance sheets reflect changes in ownership

As discussed above, balance sheets tend to overlook internally-created assets. An item’s historic cost is the preferred anchor for most assets because it is reliable (a transaction establishes a verifiable value for the asset), comparable (there is a common methodology for determining fair market value for assets), and consistent (the asset’s valuation does not fluctuate based on short-term market movements).

There are two implications of this:
1. While there are exceptions (notably internal software development costs), most assets are recorded on the balance sheet when they are purchased. This means that the ‘intangible asset’ line on the balance sheet is not fully reflective of all the company’s intangibles, only the ones that have been purchased or are the result of M&A.
2. In an acquisition, all assets and liabilities of the purchased company are updated to reflect the value of the purchase; this valuation becomes the historic basis that the acquiring company uses for it’s balance sheet. This means the amount of assets reported on the balance sheet can be underreported if the market value of the asset diverges significantly from the historic price paid. (While James argues that data ages like wine, in practice most assets on the balance sheet do not appreciate in value. See further discussion in the impairments section below.)

How would this impact adding data to the balance sheet? Under current rules, data that is generated from the company’s operations (aka an internally-created intangible) would not be reflected on the balance sheet. Furthermore, the value of any acquired data would be limited to its purchase price.

Assets have useful lives

A fundamental tenet of accounting is matching revenue with its corresponding expense. Because investing in a long-term asset helps drive revenue for the company over an extended period of time, the cost of acquiring a long-term intangible asset is spread out over the useful life of the asset using amortization. While assets with indefinite lives are not subject to amortization, assets with finite lives are.

The determination about whether an asset is subject to amortization, and if so, what amortization schedule is appropriate is based on the expected useful life of the asset. This determination of useful life provides relevance (asset values are reduced over time to indicate an decreasing usefulness as they age), reliability/comparability (companies follow amortization schedules based on their asset types), and consistency (expenses better match revenues over time).

If we wanted to add data to the balance sheet, there would first need to be an effort at the regulatory/industry level to determine the useful life of data to ensure that there remains comparability between how companies amortize.
– Does all data have a similar useful life, or does some data have more longevity and usefulness than others? Is data indefinitely valuable?
– Is data used similarly enough across the industry that you can establish these standards?
– Do companies have an adequate understanding today of what data components are going to be future economic drivers?
– Most importantly, are companies willing to be transparent enough about both the data they have and their business strategies around said data that they would be willing to engage in these discussions?

Agreed upon determinations about data’s useful life would be required in order to include data on the balance sheet.

Asset valuation and impairment relies on comparables and projected cash flows

If a company wants to throw an asset, it will also need to provide adequate support for its ongoing valuation to auditors.

Accountants prefer conservative estimates of value, which means that most items on the balance sheet are recorded either at their historic cost, or at the lower of either their historic cost or their fair market value. Whatever valuation method is used, the valuation is subject to testing over time.

Long-term assets are subject to impairment analysis to ensure that an asset’s carrying value (amount on the balance sheet) does not exceed it recoverable fair value; if the fair value (e.g. future utility of the asset) is lower than the carrying value, current accounting standards require companies to write their assets down. This ensures relevance (companies should not continue to reflect asset valuations that have diminished expected future value) and reliability (third party judgment of asset value).

A few common things that can cause asset impairment are*:
– significant, unrecoverable decreases in the asset’s market value
– significant changes in how the asset is used
– significant changes is the projected forecast of an asset’s economic value

If we want to apply these impairment testing standards to data, we run into issues. In terms of market valuation, is data used consistently enough between companies that we could establish comparables? Data is different from many other assets in that it is replicable and has multiple possible simultaneous use cases: would companies have a handle on their data’s expected uses cases and their data strategy? Can companies forecast cash flow associated with their data across these various use cases?

The question of transparency is again the primary concern for how data could be incorporated into the balance sheet. If companies want to claim data as an asset, they (and their auditors) would also need ways to substantiate its value.

Is there an inherent value of data?

This post is getting lengthy so we’ll save a discussion of software capitalization for another day. However, it’s worth noting that when we capitalize software projects there are rules about when the development work is allowed to be declared an asset. A company cannot declare software development as an asset until the work has reached a point of technological feasibility.

If we want to put data on the balance sheet, will we need to follow similar standards? Just as hours spent developing software are not inherently valuable unless they are in service of creating a specified project, does data itself have no intrinsic value without a monetizable project? At what point does data shift from being a collection of data points to useful, actionable information? What is the critical mass of data that reflects having a viable, long-term asset?

Motivations of Various Parties

If we wanted to quickly summarize:
– data has an increasing value as a business driver
– there are a variety of barriers to reflecting this value in financial terms, including:
* internally-developed intangible assets often do not show up on the balance sheet
* industry consensus would be needed around the useful life of data
* transparency would be needed around substantiating asset valuation
* at what point in time does data have value?

In addition to these barriers, the landscape of current actors indicates that any changes in this area will not be forthcoming soon.

From a regulator’s perspective, adding data to the balance sheet increases the potential for subjectivity in financial reporting. Changing the reporting intangibles will allow for more judgment calls by companies and auditors, and thus more possibility of inaccurate reporting. Regulators like clear rules and conservative outcomes.

From an external accountant’s perspective, there is the opportunity to profit from increased complexity. Adding more assets to be reviewed and audited is an opportunity for more billable hours. However, given that the accounting industry is currently in the midst of implementing the new revenue recognition guidelines, there is likely little incentive at the moment to begin pursuing new changes to accounting practices.

From a company’s perspective, there is little motivation to trade their current opacity for more regulation. Companies are currently operating in an investment world that does not demand transparency in their data (and many other intangibles) because accounting rules offer no way for companies to provide it. A company’s data and their strategy around it are competitive advantages, and there is little incentive for companies to push for greater transparency in this area. This is particularly true for big companies that are unlikely to be acquired; unless the rules also change around internally-developed assets, the majority of data from the biggest players in the technology market will remain off-balance sheet anyways.

From an investor’s perspective, it would be excellent to have more insight into key drivers of a company’s operations. An investment firm’s valuation models are their differentiators; while I assume all of these firms would like more and better variables to feed into their models, they also have a valuation system that at least for the moment works. McGraw Hill Financial indicated that S&P 500 companies traded on average at 2-3x their book value (meaning their market value was 2-3x more than the value indicated by their balance sheets) and posited that intangible assets were largely the cause of the discrepancy; reading between the lines, its clear investment firms already have some kind of valuation system working behind the scenes.

From a political perspective, this is not an environment that is currently favorable to increased financial regulation.

These barriers lead me to conclude that no party is sufficiently motivated to make data a balance sheet line item in the near term. I agree with Steve and James that data is an important business driver, but even more broadly I think there will need to be a reckoning about how we treat intangible assets at some point in the future. As the the economy continues to shift towards intangible assets, maintaining the relevance of the balance sheet with the current accounting approaches will be increasingly problematic. The limitations of the balance sheet as it currently stand will need to be addressed, but I don’t expect them to be addressed anytime soon.

* I spent a lot of time with my old accounting textbook getting this post done. Many thanks to Messrs. Kieso, Weygandt, and Warfield for all the reference material/direct quotes provided by Intermediate Accounting, 11th Edition.

Disclaimer: Please remember I am not an accountant and as such this list should neither be viewed as comprehensive nor as professional advice.