Earnings Data Is Wrong & How You Can Make More Money

Unusual gains/losses distorted earnings per share by an average of $1.16/share (22%) for firms in the S&P 500 in 2018. The popular measures of “core” earnings – i.e. street earnings[1], Compustat[2] or non-GAAP metrics – do not properly account for unusual items and are subject to significant bias according to new research from Harvard Business School (HBS) and MIT Sloan.

Why is earnings data wrong?

In “Core Earnings: New Data and Evidence,” Ethan Rouen and Charles Wang from HBS and Eric So from MIT Sloan show that traditional data providers and analysts are missing or mis-categorizing a very material and growing amount of unusual gains/losses.

“…many of the …{unusual gains/losses} collected by New Constructs do not appear to be easily identifiable in Compustat…” – page 13

The authors show the market is inefficiently assessing earnings because too few people read financial footnotes.

Earnings Data Is Getting Worse

The average per share impact of unusual items has increased 170% from $0.43/share in 1998 to $1.16 at the end of 2018. Figure 1 plots the total impact of unusual items or “Total Adjustments”.

Figure 1: Total Adjustments Per Share for the S&P 500

Sources: New Constructs, LLC and company filings

Excludes Berkshire Hathaway (BRK.A), which represents a significant outlier in all years

Total Adjustments captures the after-tax value of all the unusual items that we collect from the income statement and the footnotes. Though our numbers bear a close resemblance, the HBS & MIT paper is circumspect about exactly which of our adjustments are used to provide the superior measure of “core earnings” featured in the paper. To bring as much value as possible to investors, we share details on all of our adjustments.

For a full description of the adjustments used here and in the HBS paper, please see the final section of this report and the Appendix.

How You Can Make More Money

“Trading strategies that exploit {adjustments provided by New Constructs} produce abnormal returns of 7-to-10% per year.” – Abstract, 4th sentence

The paper presents a long/short strategy that holds the stocks with the most understated EPS and shorts the stocks with the most overstated earnings. Positions are opened in the month each 10-K is filed and held until the next 10-K is filed, or about a year.

This simple, low turnover strategy produced abnormal returns of 7-to-10% a year. These abnormal returns show that the market misses important data in the footnotes and that investors who adjust for unusual items can make more money.

Here’s a Long Idea Based on our Proprietary Data

Our research shows AbbVie (ABBV) is a stock investors should look to buy ahead of its Q3 earnings report on November 1. ABBV had some of the largest expense adjustments, or most understated earnings, of all companies in the S&P 500 in 2018. Its largest non-recurring expense was a $5.1 billion asset impairment. ABBV also disclosed the following non-recurring, unusual expenses in the footnotes:

  • $350 million in charitable contributions recorded in selling, general & administrative expenses – page 32
  • $338 million in income tax expense related to the Tax Cuts and Jobs Act – page 79
  • $70 million in restructuring charges recorded in cost of products sold, R&D, and SG&A – page 61

In total, we identified $3.91/share in net non-recurring expenses that artificially lowered reported earnings. $2.81/share of these adjustments were not disclosed on the income statement.

Without our adjustments, ABBV’s reported EPS show a modest increase, from $3.30/share in 2017 to $3.66/share in 2018. However, after removing non-recurring items, we find that ABBV’s core earnings increased significantly from $4.73/share in 2017 to $7.57/share in 2018.

ABBV’s understated earnings make it more likely to beat expectations when it reports 3rd quarter earnings.

Also, ABBV’s valuation is cheap so there’s less risk in the stock. The combination of understated earnings and a cheap valuation earns the stock our Attractive rating.

More Stocks to Buy: Most Hidden Expenses and Understated Earnings

Figure 2 shows the firms in the S&P 500 with the largest total after-tax expense adjustments in 2018. These items artificially decrease reported earnings and must be removed to calculate a firm’s true recurring profitability.

Figure 2: Top 5 Net Hidden Expenses in 2018 – $ value

Sources: New Constructs, LLC and company filings

Excludes Berkshire Hathaway (BRK.A), which represents a significant outlier

While our data and models are updated daily as new 10-K and 10-Qs are filed, we’re only showing data through 2018 because we do not give away too much valuable data for free.

Defining Total Adjustments in This Analysis

In this report, Total Adjustments includes

  • + Acquisition and Merger Expenses, net
  • + Foreign Currency Expenses, net
  • + Legal Regulatory and Insurance Expenses, net
  • + Company Defined Other Expenses, net
  • + Other Non-Recurring Expenses, net
  • + Other Real Estate Owned Expenses, net
  • + Other Financing Expenses, net
  • + Derivative Related Expenses, net
  • + Pension Related Expenses, net
  • + All Restructuring Expense, net
  • + Minority Interest Expense, net
  • + Preferred Dividends, net

See Figure I in the Appendix for the definition of core earnings and total adjustments used in the HBS & MIT Sloan paper.

We further breakdown each of the adjustment categories above in Figure II of the Appendix. This level of granularity allows investors to get a clear picture of all adjustments required to calculate a firm’s true earnings.

Appendix 2 provides details on the most complete data set we offer clients here.