Accounting Standards Update (ASU) 2016-01, Recognition and Measurement of Financial Assets and Financial Liabilities, became effective for financial statements beginning after December 15, 2017. One of its key provisions was a change in the treatment of equity securities classified as available for sale. Under prior guidance, the change in unrealized gains and losses for available for sale equity securities was recognized in equity as a part of other comprehensive income. Under ASU 2016-01, the change in unrealized gains and losses on all equity securities is recognized in income. For companies that have large portfolios of equity securities that were classified as available for sale, this change may significantly affect earnings. In the insurance industry, for example, companies with large portfolios of equity securities classified as available for sale securities showed dramatic increases in both earnings volatility and earnings surprises for 2018 and 2019, the first two years of implementation.
Accounting Standards Update 2016-01 (ASU 2016-01), Recognition and Measurement of Financial Assets and Financial Liabilities, became effective for financial statements beginning after December 15, 2017. In his annual letter to Berkshire Hathaway shareholders in early 2018, Warren Buffett criticized the new accounting treatment under ASU 2016-01 and warned of increased volatility in earnings: “I must … tell you of a new accounting rule … that in future quarters will severely distort Berkshire’s net income figures and very often mislead commentators and investors” (https://bit.ly/347Kpt0). Two years after adoption of the new guidance, Buffett continued his criticism of ASU 2016-01. In his annual letter to shareholders, Buffett commented: “Berkshire earned $81.4 billion in 2019 … including a $53.7 billion gain from an increase in the amount of net unrealized capital gains that exist in the stocks we hold … that $53.7 billion gain requires comment. It resulted from a new GAAP rule, imposed in 2018, that requires a company holding equity securities to include in earnings the net change in the unrealized gains and losses of those securities … the adoption of the rule by the accounting profession, in fact, was a monumental shift in its own thinking. Before 2018, GAAP insisted—with an exception for companies whose business was to trade securities—that unrealized gains within a portfolio of stocks were never to be included in earnings and unrealized losses were to be included only if they were deemed “other than temporary.” To support his argument, Buffett added, “Berkshire’s 2018 and 2019 years glaringly illustrate the argument we have with the new rule. In 2018, a down year for the stock market, our net unrealized gains decreased by $20.6 billion, and we therefore reported GAAP earnings of only $4 billion. In 2019, rising stock prices increased net unrealized gains by the aforementioned $53.7 billion, pushing GAAP earnings to the $81.4 billion reported at the beginning of this letter. Those market gyrations led to a crazy 1,900% increase in GAAP earnings!” (https://bit.ly/3g7z8LP).
Similarly, Steven J. Johnston, president and chief executive officer of Cincinnati Financial, commented in its earnings release for the 4th quarter of 2019: “Our GAAP net income rose 596% to $1.997 billion for the year, in large part because of strong equity markets in 2019. While our swing in net income may be surprising, it is the continued result of the accounting rule changes implemented by the Financial Accounting Standards Board in 2018 … the volatility this rule change introduced, by requiring unrealized investment gains and losses to be recognized as part of net income instead of on the balance sheet, may distract investors” (https://bit.ly/32Bewsv).
To determine whether these concerns were well founded, the author reviewed the results of a sample of insurance companies to assess whether ASU 2016-01 affected earnings volatility and earnings predictability. The sample compared the two years since the adoption of ASU 2016-01 to the five years prior.
Change in the Treatment of Equity Securities
Prior to the 2007–2008 financial crisis, FASB and the IASB began a joint project to improve and converge the accounting for financial instruments. The result of the project, ASU 2016-01, addresses certain aspects of recognition, measurement, presentation, and disclosure of financial instruments. One of the main provisions of the new guidance was to require recognition of the changes in unrealized gains or losses in income for all equity securities accounted for under the fair value method. Previously (under SFAS 115), management intent was a key factor in determining where the changes in unrealized gains and losses were recognized. If management’s intent was to maintain a buy-and-hold strategy, the equity securities were classified as available for sale (AFS equity securities); otherwise, the securities were considered trading securities, with equity securities under both classifications valued at fair value. The corresponding changes in the value of trading securities, as short-term holdings, were recognized in income. Conversely, the changes in fair value for AFS equity securities were reflected in other comprehensive income (OCI) and stockholders’ equity.
Although the new guidance in ASU 2016-01 converges with IFRS 9 in certain areas, IFRS 9 allows entities to make an irrevocable election at initial investment to present subsequent changes in fair value in OCI. Thus, FASB’s approach represents a more aggressive move towards fair value accounting because the changes in fair value for all equity securities (with available market values) affect income.
FASB concluded that measuring changes in fair value through net income improves the decision usefulness of financial statements by presenting the exposure to financial instruments in net income rather than OCI. The board concluded that fair value presented in net income is a more relevant measurement attribute for equity securities, because the total realizable value of most of those investments is primarily realized by selling the investment. The board contended that the returns for equity securities differ from debt instruments because the value for debt instruments may be realized through the collection of interest and principal. As a further argument supporting the new guidance, previously management could elect to sell equity securities to boost earnings for a particular period even though much of the increase in the value of the securities occurred in prior periods.
Because changes in unrealized gains and losses were previously recognized in OCI, comprehensive income remains comparable under the current and prior treatments. FASB’s rationale that fair value recognition in income is a more relevant measurement suggests that the users of financial statements are less able to incorporate changes in OCI, as well as comprehensive income, into their decision making.
Impact of ASU 2016-01 on Decision Usefulness
Useful financial information permits users to forecast future cash flows, evaluate performance, and compare performance across time and entities. New accounting guidance, if it is to be effective, should improve predictability and the evaluation of company performance. If, however, a change in guidance produces confusion or misinterpretation, or reduces comparability, then decision usefulness is not improved.
Under prior guidance, the earnings for insurance companies with significant AFS equity securities were comparable to others in the industry because earnings primarily reflected operating activities, whereas comprehensive income reflected the change in unrealized gains or losses on securities. By reviewing comprehensive income, investors and analysts could factor in the change in AFS equity securities into their evaluation of a company’s performance. Under the new guidance, investors and analysts must be cognizant that movements in the equity markets affect earnings for insurance companies with significant equity holdings. As a result, comprehensive income is the most comparable measure for periods prior to and after the adoption of ASU 2016-01.
Investors interpret and react to new information as it becomes available. Most often, earnings releases are the first source of financial information available for a given period. As companies announce quarterly earnings, the financial press, investors, and analysts compare the reported earnings to analysts’ forecasted earnings and prior periods. Under the new guidance, forecasting earnings may become more challenging for companies with a substantial portion of their assets in equity securities. If analysts’ ability to accurately forecast earnings is diminished, the likely result would be an increase in earnings surprises (i.e., a large percentage difference between actual EPS and consensus forecast EPS). Furthermore, if the change in the treatment of equities securities causes investors to misunderstand the earnings communicated in the earnings release, or the comparison of earnings to prior periods, they may incorrectly react to the earnings announcement.
Many companies in the insurance industry hold significant amounts of equity securities, following a buy and hold strategy; and these equity portfolios were previously classified as AFS. Because Warren Buffett is a vocal critic of ASU 2016-01, and his company, Berkshire Hathaway, is an insurance company, the author chose a sample of companies from the insurance industry to assess the impacts of ASU 2016-01 on earnings. Based on SIC code, the author selected all companies classified as insurance carriers and reviewed the financial statements for the five years prior to 2018 (the first calendar year impacted by ASU 2016-01) and two years (2018 and 2019) after implementation. The sample consisted of 41 companies with available information for all periods tested. To examine the impacts across the industry, the author further separated the sample companies into four groups based on their AFS equity securities as a percentage of total assets as of 2017 year-end. As shown in Exhibit 1, the four groups were equity securities as a percentage of total assets of less than 1%, 1–5%, 5–10% and greater than 10%.
Sample of Insurance Companies
As of the end of 2017, the percentage of AFS equity securities varied from a low of zero to a high of approximately 28%. As shown in Exhibit 1, a large percentage (48%) of the companies in the industry had less than 1% of total assets in AFS equity securities, whereas the two groups of companies with over 5% of their assets in AFS equity securities represented a combined 31% of the companies.
The author reviewed the standard deviation of earnings per share (EPS) prior to and after the adoption of the new guidance to test the impact of ASU 2016-01 on earnings volatility; the results were compared for each of the groups based on the percentage of AFS equity securities. The author performed a second analysis of the impact of ASU 2016-01 focused on the predictability of earnings using analysts’ forecast errors, the difference between actual EPS and consensus forecast EPS, with the results analyzed for each of the four groups.
Impact of ASU 2016-01 on Earnings Volatility
Exhibit 2 shows that the volatility of earnings was comparable across all groups prior to the adoption of ASU 2016-01. While the volatility ranged from an average of .32 to .59, the volatility was slightly lower for companies with a higher percentage of assets in AFS equity securities. Since implementation of ASU 2016-01, there has been a significant increase in the volatility of EPS for insurance companies with higher percentages of their total assets in AFS equity securities (see Exhibit 2).
For the group of companies with 5–10% of their assets in AFS equity securities, volatility increased from .32 to 1.13. For the group of companies with more than 10% of their assets in AFS equity securities, earnings volatility increased from .42 to 1.07. Whereas earnings volatility more than doubled for both groups of companies with more than 5% of AFS equity securities, earnings volatility for companies with lower levels of AFS equity securities modestly decreased (see Exhibit 2).
Impact of ASU 2016-01 on Earnings Surprises
Volatility in earnings is not necessarily a negative consequence of new guidance if those earnings are understandable and predictable. To gauge the impact of ASU 2016-01 on the predictability of earnings, the author reviewed the accuracy of analysts’ forecasts (analysts’ forecast error) before and after adoption of the guidance. Large earnings surprises, the differences between actual EPS and analysts’ consensus forecasted EPS, may indicate either unexpected results or an inability of professional analysts to accurately forecast earnings.
Under the new guidance, forecasting earnings for a company that holds a significant amount of equity securities requires that analysts not only forecast operating activities, but also the anticipated changes in prices for equity holdings. As expected, earnings surprises were more significant, and increased more dramatically, for the largest AFS equity securities holders than for others (see Exhibit 3). The average earnings surprise increased slightly (27% to 31%) for companies with less than 1% of their assets in AFS equity securities, and the average earnings surprise increased modestly (48% to 52%) for the periods after ASU 2016-01 for companies with modest AFS equity security portfolios (less than 5% of total assets). Conversely, the average earnings surprise increased from 52% to 90% for companies with 5–10% of their assets in AFS equity securities. For the group with more than 10% of their assets in AFS equity securities, the average earnings surprise more than doubled, from 44% to 97%.
Disclosure of the Impact of ASU 2016-01 on Earnings
Companies have the opportunity to discuss how accounting guidance, including ASU 2016-01, affects their financial statements through earnings releases, management discussion and analysis (MD&A), and other supporting communications. In reviewing the earnings releases of the companies with less than 1% of their assets in AFS equity securities, none directly discussed the impact of ASU 2016-01 on their earnings in 2018 or 2019. However, some companies identified the impact of the change in unrealized gains or losses on their equity portfolios in their earnings releases. None of these companies specifically focused on the impacts of changes in unrealized gains or losses in their MD&A in their 10-Q and 10-K filings.
The group of companies with 1–5% of their assets previously classified as AFS equity securities had limited direct discussion of the impacts of ASU 2016-01. Some companies did identify the impact of the investment gains and losses in their discussion of earnings; however, many companies combined the change in unrealized gains or losses on earnings with realized gains and losses. Users must examine tables or the financial statements to identify the impacts of unrealized gains or losses.
Similarly, only a few of the companies with 5–10% of their assets in AFS equity securities directly discuss the impact of ASU 2016-01, or unrealized investment gains, on earnings. Users had to examine tables or the financial statements to identify the impacts of unrealized gains or losses. A common practice for companies in these circumstances is to exclude investment gains or losses from their adjusted earnings (GAAP earnings with certain items excluded) and focus their discussion on adjusted earnings instead of GAAP earnings. Depending on the significance of their unrealized gains and losses, some of these companies also discussed the impact of investment gains and losses on their earnings in their MD&A.
For the six companies with more than 10% of their assets in AFS equity securities, their earnings were significantly affected in both 2018 and 2019 due to ASU 2016-01. However, the discussion of the impacts of the change varied from clear and distinct to subdued and mute. All of these companies provided information in their earnings releases and MD&A that allowed users to identify the impacts of investment gains or losses on earnings, however, the description of the impacts and the level of emphasis on their transitory nature was inconsistent.
Warren Buffett’s Berkshire Hathaway, which had nearly 25% of assets classified as AFS equity securities, was quite consistent and clear in the discussion of the impacts of the accounting change both in 2018 and 2019, in their earnings releases, MD&A, and the aforementioned letter to shareholders. Similarly, Cincinnati Financial, which had 28% of assets as AFS equity securities, clearly discussed the impact of ASU 2016-01 on their earnings in 2018 and 2019.
Other companies with significant AFS equity securities (more than 10% of assets) were more muted in their discussion. None of the other four companies in this group (Alleghany Corp., Markel Corp., RLI Corp. and State Auto Financial Corp.) specifically discussed the impacts of the recognition of unrealized gains on equity securities on their earnings in 2019, although all of these companies’ provided tables in which users could identify the impacts. Three of the four companies (Markel Corp., RLI Corp. and State Auto) did clearly discuss the impact of the adoption of ASU 2016-01 in 2018, when the adoption negatively affected their earnings.
Dealing with Volatility
The change in the treatment of AFS equity securities has added volatility to earnings and increased earnings surprises for many companies in the insurance industry. Although most companies’ earnings, particularly outside the insurance industry, are not significantly impacted by ASU 2016-01, it has affected the earnings (and possibly the understandability of earnings) of companies that have large portfolios of equity securities that historically followed a buy and hold strategy. As a result, earnings of these companies display greater volatility over time, making inter-period comparability more difficult. Because the level of equity securities differs greatly across the insurance industry, comparability within the industry has become more challenging after adoption of ASU 2016-01.
Consequently, the discussion of the impacts of unrealized gains and losses has become more important to understanding earnings. For some companies in the insurance industry, however, this discussion is limited, unclear, or inconsistent; as a result, less sophisticated users may misinterpret the results. Users would benefit if companies identified the impacts of the changes in unrealized gains and losses on earnings and clearly distinguished between realized and unrealized gains and losses. As an additional consequence of ASU 2016-01, many companies with significant equity holdings might emphasize their calculation of adjusted earnings, excluding investment gains and losses, and de-emphasize GAAP earnings.
A potential solution for FASB to consider would be to allow companies to make an irrevocable election at initial investment to present subsequent changes in fair value in other comprehensive income, similar to IFRS 9. This change would allow for greater comparability of earnings for these companies (with significant buy and hold equity portfolios) over time and reduce the need for companies to isolate the impacts of unrealized gains or losses in order to improve the understandability of their earnings. In the absence of new guidance, it is incumbent upon companies to ensure understandability by clearly discussing the impacts of unrealized gains and losses in the communications that accompany their earnings releases and financial statements.