Accounting rule makers have long debated whether companies should recognize intangible assets on their balance sheets. At the heart of this debate is whether recognized values can predict future income and cash flows, or whether the high degree of measurement uncertainty embedded in intangible asset values precludes such predictive ability. While many agree intangibles like patents, trademarks, and goodwill contribute significantly to value creation and profitability, separately identifying and reliably measuring intangibles is costly. In fact, due to measurement-related concerns, accounting rules generally do not allow companies to recognize internally generated intangibles on their balance sheets. Intangibles acquired externally in business combinations are recognized, however, at their fair values.
Against this backdrop, our study examines whether recorded fair values of acquired intangible assets predict future operating returns and how operating payoffs to acquired assets vary across acquisition components: tangibles, identifiable intangibles, and unidentifiable intangibles (i.e., goodwill). Overall, we find intangible asset values recorded in acquisitions are reliable predictors of future performance, with the highest returns accruing to investments in goodwill. However, in settings where expected measurement error is high, returns to intangibles are much lower and are often not detectable, implying measurement uncertainty is an important element in understanding the relationship between intangibles and future payoffs.
* * * * * *
Our study’s theoretical framework is simple. Absent measurement-related uncertainty, the operating returns to acquired assets should be a function of their riskiness and the economic rents a company can earn from owning them. However, measurement error in this framework would result in decreased returns to the recognized values of acquired assets, on average. From a policy perspective, if acquired intangibles are predictive of future operating returns, then the costs of recognizing them is likely justified. However, if recognized intangible asset values are poor predictors of operating returns, this would call into question the wisdom of requiring companies to assign them values.
We use the returns to acquired tangible assets as a benchmark for studying the returns to acquired intangibles. In principle, one might expect the returns to intangibles to be higher than tangibles because they are riskier and a source of competitive advantage. However, compared with tangibles, intangibles are rarely traded, and their estimated and recognized values are thus subject to significant measurement uncertainty.
Using a unique dataset of roughly 4,000 acquisitions completed between 2003 and 2014 (provided by Houlihan Lokey), we find a positive relation between the total acquisition price and future income. Each dollar spent by the acquirer is associated with incremental future operating income (i.e., post-acquisition EBITDA less pre-acquisition EBITDA of the acquirer) of between 7 and 9 cents per year.
When we examine the components of the acquisition price, we find that total intangibles (both identifiable intangibles and goodwill together) are a reliable predictor of future income, yielding 8-10 cents per year per dollar invested. These returns are significantly higher than returns on acquired tangible assets (around 5-6 cents per year). These higher returns are driven by goodwill—identifiable intangibles have similar average returns to those of acquired tangible assets. This is consistent with goodwill being riskier and yielding higher rents for acquirers relative to other assets and suggests measurement error in the recorded value of acquired intangibles is not sufficient to result in the values assigned to them being poor predictors of future payoffs.
We next consider the impact of measurement error more explicitly. We first focus on goodwill by examining a setting where we expect significant measurement error: acquisitions with negative announcement returns (i.e., bad deals). Here investors expect the acquisition to be value-destroying, which implies the acquisition price is too high and goodwill is likely overstated. Not surprisingly, we find that for these acquisitions the operating returns to goodwill and other assets are lower than acquisitions with positive expected NPV. However, goodwill still predicts future income even in bad deals, with returns roughly equal to tangibles. Thus, despite measurement concerns, goodwill is still a reliable predictor of future payoffs, even in settings where measurement error is likely.
We then examine settings where we expect significant measurement error in identifiable intangibles. Mismeasurement of identifiable intangibles could either be: a) non-strategic, arising simply due to valuation uncertainty, or b) strategic, arising from managers’ incentives to misallocate the acquisition price. For non-strategic measurement error, we predict that in settings with more reliable measurement of identifiable intangibles, there should be a more reliable correspondence between recorded intangibles and future profits. We examine three such settings: larger acquirers, low-volatility acquisitions, and experienced acquirers. We generally find that the positive returns to recorded identifiable intangibles are concentrated among these types of acquirers. Among acquirers without these characteristics, we generally find no significant association between identifiable intangibles and future payoffs.
Turning to strategic measurement error, consistent with prior research we predict that when managers face higher concerns over meeting future bonus targets, they undervalue identifiable intangibles to avoid amortization charges reducing future earnings. This should result in more measurement error in identifiable intangibles when bonus incentives are higher. Indeed, we find returns to identifiable intangibles are higher than tangibles when bonus incentives are relatively low.
We also examine the impact of measurement uncertainty across types of identifiable intangibles and find positive returns to customer-related intangibles such as customer contracts, customer lists, and backlog. These returns are higher than the returns to non-customer intangibles such as in-process R&D, patents, and trademarks. In fact, we find no significant association between future income and non-customer intangibles. This finding is consistent with the values assigned to customer-related intangibles containing less measurement error than the values assigned to non-customer intangibles, likely due to the close temporal proximity between customer-related assets and sales.
Our findings offer practical takeaways for standard setters. The FASB is considering changes to the accounting for and disclosures about identifiable intangible assets, including allowing or requiring certain intangibles to be subsumed into goodwill. While our study cannot specifically address the costs of separately measuring and recognizing intangible assets, our results indicate that in a large sample of acquisitions with public acquirers, a relation between investments in identifiable intangibles and future income exists, is economically significant, and differs between goodwill and identifiable intangibles. Thus, separate recognition provides information to financial statement users.
This post comes to us from professors John M. McInnis at the University of Texas at Austin and Brian Monsen at Ohio State University. It is based on their recent article, “The Operating Returns to Acquired Intangible Assets,” available here.