The End of Accounting

first posted: 2024-08-02 09:14:28.031064

In 2016, Baruch Lev and Feng Gu publish The End of Accounting. This provocatively titled book is based on a few high level estimations of the informativeness of financial statements.

The subject of this book is important, too few informed discussions look into the problems with accounting based on actual filings and concrete proposals towards improvement. Too often, the discussion focuses on what is and on material that is easy to teach.

Incentives of management has diverged from that of company shareholders
Auditors are concerned with satisfying regulatory obligations while minimizing litigration risk, informing shareholders is a role that is performed reluctantly
The accounting profession is subject to constant lobbying by management and auditors
Baruch Lev study shows that financial statements used to be relevant
Earnings rely more and more on management estimates, but management is not made accountable of past estimates mistakes.
The accounting regulators has imposed more and more rules than make earnings diverge from cashflow. It is not clear however that these rules were shown to improve forecast.
PPE is accounted at purchase value, whereas much more important investments, such as internal software development, R&D, procedures are expensed.
The author does not touch on the fact that accounts lack articulation for most companies, to the point that accounting professors share a small set of companies whose account articulate well to teach accounting. This points to a lack of accountability of management and regulators.
Alternative approaches such as CFROI were proposed, but again, too much internal investment is expensed as SG&A, making the approach difficult.
The author is reports focusing on sector specific KPI that are found to be relevant. While it might be useful, this does not help with cross sector allocation. The accounting treatment of intangible should be modified to be show to improve the predictability of statements.
Management should make their estimates clear, and publish their track record on the accuracy of past estimates.

The main obstacle, as noted by Baruch Lev, is alignment of incentives between the principals and their many agents.

Ossification of Financial Statements

Double entry accounting was first introduced in the renaissance, but it use was generalized and standardized in the 19th century. Yet, the author notes that financial reporting have ossified from 1900 to today.

As the authors explain when looking at US Steel financial statement in 1902 and 2012:

The main components of this report, the balance sheet and the income (profit or loss) statement for the previous year follow very similar rule. The main innovation was the introduction of the cashflow statement, management comments and comprehensive footnotes.

Earnings have become less relevant than Cashflow

The author claims that long/short portfolio of the 5 companies with the highest earnings against the 5 with the lowest has a 25% return. They then claim that if cashflow is used instead of earning, the returns are much higher.

Cash flow is a much simpler metric, more straightforward and easier to compute than earnings.

Earnings Had Its Days of Glory Figure 2.2 traces the annual gains over the past quarter century from following each of the earnings and cash flow investment strategies. That is, investing

ahead of the annual financial report release in the shares of the five companies with the highest earnings (or cash flows) in their industry, and selling short the five companies with the lowest earnings (cash flows) in the industry. As in Figure 2.1, this analysis is performed over all major US industries.8

Note that in the first nine years of the past quarter century, 1989−1997, earnings easily dominated cash flows in yielding investment returns. In some years (e.g., 1991 and 1995), the earnings strategy yielded more than twice the cash flow strategy. This was apparently the raisón d'etŕe for analysts' complex spreadsheets designed to predict earnings. Things, however, changed around the turn of the twenty-first century—and Part II of the book will pinpoint the reasons—as the edge of the earnings over the cash-flow strategy narrowed significantly,

A sketchy argument: insufficient evidence?

A 5 company portfolio is notoriously idiosyncratic, and there is scant discussion that the author checked different subportfolio sizes, or measures of earnings or cashflow.

Loss of Predictive Power of Financial Statements

The relevance of reported earnings and book value to investors has faded:

We see this decline for earnings:

as well as for book value:

Generalising to other earnings quantity does not improve the quality of the market value preduction:

Analyst Estimates and Non Financial SEC filings

We see below that in the last 25 years, the analyst estimates and the non financial filings relevance to investors has increased:

Here are a few comments by the author:

The two most prevalent uses of reported corporate earnings: generating earnings forecasts to guide investment decisions and assessing corporate performance by earnings surprises (relative to analysts' consensus estimates). In both cases, we found that the relevance of reported earnings to investors faded.

In generating investment returns, earnings are surpassed by cash flows, and in assessing corporate performance, investors' feeble reaction to earnings surprises casts serious doubt on the economic meaning of reported earnings.

Stock prices thus reflect the aggregate reaction of investors to the information conveyed to them. Relating stock prices to financial information is, therefore, a straightforward way of assessing the relevance of this information to investors.

The steady decline of the graph, from over 90 percent in the 1950s (see horizontal scale, spanning the past 60 years) to around 50 percent currently—a fall of almost half—tells vividly the story of the decline in the relevance of corporate financial information to investors.

What Figure 3.1 tells us is that in the 1950s, 1960s, and even 1970s, the key financial report variables, earnings and book values, were critical to investors' valuation of companies, whereas the usefulness or relevance of these two variables to investors has diminished considerably since then.

To predict future corporate performance. The quality of the financial information underlying financial analysts' work product, as reflected by their uncertainty about future corporate performance.

Predictive Power of Earnings and ROE

The author notes that earning prediction if based on prior earnnings and ROE are on the rise:

According to him:

Three major reasons for accounting's relevance lost: The inexplicable accounting treatment of intangible assets—the dominant creators of corporate value.

physical and financial investments, unable to create substantial value by themselves, are fully recognized as assets on corporate balance sheets
whereas investments in internally generated intangibles, such as patents, brands, or knowhow immediately expensed;
accounting isn't about facts anymore, but rather about managers' subjective judgment, estimates, and projections. employee pensions and stock options expenses are based on multiple estimates, the prevalence of estimates in accounting is constantly increasing,

Unrecorded events increasingly affect corporate value. Increasingly, however, nontransactional business events affect corporate value too: success or failure of drugs or software products under development in feasibility tests, competitors' new products or strategic moves disrupting incumbents, environmental mishaps, the signing of new contracts or contract cancellations, strategic moves by the firm (restructurings, launches of new products or services), or new regulations affecting the company.

Consider Boeing and competitor Lockheed Martin, which follow different innovation strategies: Boeing develops most of its technology internally—its 2012 R&D-to-sales ratio was 4 percent and its goodwill-to-assets ratio, reflecting the intensity of technology acquisitions, was 5.7 percent.
Lockheed Martin relies mainly on technology acquisitions. Its 2012, R&D-to-sales ratio was only 1 percent, but its goodwill-to-assets ratio was 26.8 percent, almost five-times Boeing's.

For growth companies, for example, acquiring intangibles, rather than developing them internally, will raise reported earnings and asset values (avoiding the expensing of intangibles).

Better for managers—but not for investors, of course—to immediately delete such assets from the balance sheet (and investors' memory) by expensing them, than being held responsible for years for their creation or acquisition.

Research refutes this accounting rule by documenting a statistically significant correlation between R&D and stock prices or subsequent sales growth. For example,

There is no single magic number, no “bottom line,” or even triple bottom line (profit, people, planet).4

Carol Corrado, Charles Hulten, Daniel Sichel, Intangible Capital and Economic Growth, working Paper 11948 (Cambridge, MA: National Bureau of Economic Research, 2006).
Bamford and David Ernst, “Managing an Alliance Portfolio,” The McKinsey Quarterly, 3 (2002):
Littleton, “Value and Price in Accounting,” The Accounting Review, 4 (1929):
Baruch Lev, Stephen Ryan, and Min Wu, “Rewriting Earnings History,”
The Fund Theory of Accounting and Its Implications for Financial Reports Chicago

Analysts Uncertainty Increased

While reported earnings are less informative than analyst forecase, the perplexity of analysts has increased:

Note that the analysts as a group have an interest in publishing research that are within the consensus window. Having a wider analyst opinion window might point to a situation where analysts are freer than before to express their opinion and broaden the consensus. According to the author, the following improvements should be made in the company statements:

Usefulness attribute no. 1: Inform investors about the strategic resources (assets) of the enterprise, their characteristics, value, and related attributes (such as number of patents in the company's portfolio, patents supporting products/services, number of patents licensed out, patent quality, protection mechanisms against infringement, etc.).6
Usefulness attribute no. 2: Inform investors with specificity about the investments (expenditures) made in the process of building the enterprise's strategic assets (customer acquisition costs for telecom and internet companies, for example). We challenge anyone to explain the logic underlying the current practice of separately reporting in the income statement innocuous expenses, such as interest, but not, say, the generally larger and much more consequential information systems expenditures. Even the currently reported R&D expense isn't particularly meaningful without breakdowns to, for example, the “R”—research, mainly aimed at developing new technologies—and the “D”—development,
Usefulness attribute no. 3: Articulate the major risks to the company's strategic assets from infringement by competitors, disruptions by new technologies, and regulatory moves, as well as the measures taken by management to mitigate these risks.
Usefulness attribute no. 4: Outline the specific deployment (uses) of the firms' strategic assets—the strategies to extract value from the assets.
Usefulness attribute no. 5: Quantify and report the consequences—value creation—of managers' activities in creating, preserving, and deploying strategic assets.

A report on the creation, deployment, and value added by strategic resources is industry specific. According to the author, the improvement can not be rule based, it needs to be pushed by management.

The capitalization of certain intangible investments, just proposed, will undoubtedly improve the quality of reported earnings, but much more is needed to provide relevant information on intangibles to investors.

Even the most rudimentary managerial estimates, like those for uncollectible receivables, or warranties provisions, cannot be verified after the fact, because no systematic information is available in financial reports comparing specific estimates with the subsequent facts.27 This is an invitation to careless and even manipulative estimation.

The endless chase of the perfect accounting rule is well demonstrated by the accounting for leases. For decades, accounting regulators strove to clearly define lease arrangements that, in substance, resemble purchasing the leased equipment with a loan from the lessor—termed capital lease by accountants—to be recorded by the lessee as an asset and a liability. But as soon as new lease accounting rules were formulated, parties to lease arrangements changed the terms of the lease to avoid the recording of assets and liabilities on the balance sheet,

Current accounting and financial reporting have deteriorated into a compliance activity, where managers and auditors check the very detailed regulatory boxes.

Even for R&D, if investors “capitalize” it by adding it back to earnings, past R&Ds should be amortized. But, by how much? The short of it is this: Investors cannot capitalize intangibles on their own.