The Accruals Framework – Earnings vs Accruals

Earnings are more persistent than cash flows, on average, but the cash flow component of earnings is more persistent than the accrual component.

Dechow and Schrand, Earnings Quality

The Distinction between Earnings and Accruals

In considering the accruals system, two things stand out:

  1. It makes sense:  A cash purchase of land should create an asset not an expense.  A sales contract has value, even if cash has not been received;
  2. It is subjective:   How do asset values change over time?  What is the current value of a sales contract that has not been paid?

In reconciling these two somewhat contradictory statements, we need to distinguish between:

  • Earnings and Cash Flow; and
  • Cash Earnings and Accrual Earnings.

Earnings vs. Cash Flow – Which is better?

Standard market dogma is that cash is king.  However, this isn’t necessarily the case.  Relative to cash flows, earnings are:

  • More persistent;
  • Less volatile;
  • Have a higher association with contemporaneous share price returns; and
  • Are probably a better starting point for valuations.

Earnings Quality, Decho and Schrand, 2004

Cash Flow Earnings vs. Accrual Earnings

It turns out the the distinction is not so much between earnings or cash flow as it is between those earnings comprised of cash versus those comprised of accruals. Sloan highlighted that cash earnings are more persistent than accrual earnings, but the persistence of both earnings and cash flows varies with the level of accruals.

Earnings are more persistent than cash flows at each level of accruals, as shown by the gap between these two lines.  However, where accruals are large, the persistence of earnings diminishes.  i.e. earnings not backed by cash are not as reliable.

Deschow and Ge (2003)

Peristence of Earnings vs Cash Flow


We can summarise this contradiction between cash and accruals as follows

In a situation like this where the company has heavy capex accruals, earnings are a better estimator of underlying trends than short term cash flows.

  • The period of 20012/13 was a particularly extreme example of where earnings outperformed cash flows.

However, if we have one set of earnings that is consistently backed by cash flows, and we are comparing it to another set of earnings which is largely made up of accruals, we would have much greater confidence in the cash flow earnings.

Therefore, because accruals are a better estimator of performance than cash, but this estimation is imperfect, we need to add a third layer of analysis.

We need to distinguish between the good part of the accruals estimation and the bad part.