Sustainable (Persistent) Earnings

Sustainable (persistent) earnings are earnings that are expected to recur in the future. Non-recurring earnings are not sustainable and thus are low-quality earnings. More persistent earnings are more practical inputs for equity valuation models involving earnings forecasts. One way to measure earnings persistence is to use a regression model such as:

$$\text{Earnings}_{t+1}=\alpha+\beta_{1} \text{Earnings}_{t}+\epsilon$$

A higher coefficient (\(\beta_1\)) indicates more persistent earnings relative to a low coefficient.

Earnings are composed of a cash component and an accruals component. The accruals component is discussed in details in the following section:

Accruals Component

The accrual basis of accounting requires recognizing revenues when they are earned and not when they are received in cash. Similarly, and recognition of expenses occur when incurred and not when paid.

Decomposing earnings into its two major components, cash and accruals, further enhances the quality of earnings as input for forecasting future earnings. The accruals component is less persistent than the cash component. In the following regression model,

$$ \beta_1 > \beta_2\ $$

Where \(\beta1\) is the coefficient of cash flow, whereas \(\beta_2\) is the coefficient of accruals implying that the cash flow component of earnings is more persistent than the accruals component:

$$\text{Earnings}_{t+1}=\alpha+\beta_{1} \text{Cash flow}_{t}+\beta_2 \text{Accruals}_{t}+\epsilon$$

It is crucial to note that non-discretionary accruals occur as part of normal transactions in the period. In contrast, discretionary accruals result from non-normal transactions or non-normal accounting choices and are sometimes used to distort earnings.

One approach to identifying abnormal accruals is first to model companies’ normal accruals and then to determine outliers. Discretionary and non-discretionary accruals are separated by modeling (using regression) total accruals as a function of a set of factors that usually give rise to normal accruals. The factors may include the growth of credit sales and the amount of depreciable assets. The residuals from such a model would be an indicator of discretionary accruals.

Finally, a stronger signal of questionable earnings quality is when a company reports positive net income but negative operating cash flows. The following example demonstrates a hypothetical company’s fraudulent reporting:

$$ \textbf{Example of Fraudulent Reporting Showing Positive Net Income but Negative Cash from Operating Activities} $$

$$\small{\begin{array}{l|r|r} \textbf{Three months ended 31 March (\$ millions)} & \textbf{2018} & \textbf{2017}\\ \hline\text{Net income} & 178 & 91\\ \hline\text{Net cash used in operating activities} & -711 & -704\\  \end{array}}$$

The quarterly data for the above company shows a positive net income but negative cash from operating activities in quarters that were shown to have been misreported.

$$\small{\begin{array}{l|r|r|r} \textbf{Year ended 31 December (\$ millions)} & \textbf{2017} & \textbf{2016} & \textbf{2015}\\ \hline\text{Net income} & 732 & 646 & 703\\ \hline\text{Net cash used in operating activities} & 4,532 & 981 & 1,640\\  \end{array}}$$

From the above company’s annual information, the net cash flow from operating activities is more than double that of net income in 2015, almost a half higher than the net income in 2016 and almost six times net income in 2017. An analyst can, therefore, raise a significant red flag about the earnings quality of this company. A company’s considerable accruals may reflect potential manipulation and, consequently, low-quality earnings. However, it is not automatically the case these companies will have such a profile. The investors should, therefore, explore and understand the existence of such differences.

In conclusion, albeit accrual measures can serve as indicators of earnings quality, they cannot be used exclusively or applied mechanically. Comparing cash-basis measures such as cash provided by operating activities, with net income, may provide a false sense of confidence about the net income. Net income is calculated using subjective estimates, such as the expected life of long-term assets, which can be easily distorted. It suggests that an analyst should also consider investing activities in examining the quality of earnings.

Question

An analyst estimates the following regression model to examine a company’s earnings persistence.

$$\text{Earnings}_{t+1}=\alpha+\beta_{1} \text{Cash flow}_{t}+\beta_2 \text{Accruals}_{t}+\epsilon$$

Solely based on the regression model, earnings persistence for the company is most likely to be highest if:

A. \(\beta_1\) is greater than \(\beta_2\).

B. \(\beta_2\) is greater than \(\beta_1\).

C. \(\beta_1\) is less than 0.

Solution

The correct answer is A.

Research shows that the cash component is more persistent than the accruals component when earnings are broken down into the two components.

A higher beta coefficient (\(\beta_1\)) on the cash flow variable than the beta coefficient (\(\beta_2\)) on the accruals variable shows that the cash flow component of earnings is more persistent than the accruals component. This result provides evidence of earnings persistence.

B and C are incorrect. If the beta coefficient (\(\beta_2\)) on the accruals variable is higher than (\(\beta_1\)) on the cash flow variable, it shows that the cash flow component is less persistent than the accruals component. This indicates low earnings persistence.

Reading 15: Evaluating Quality of Financial Reports 

LOS 15 (e) Describe the concept of sustainable (persistent) earnings.