On the Analysis of Firms’ Cash Flows
James A. Ohlson
and
Jagadison K. Aier
School of Accountancy
W. P. Carey School of Business
Arizona State University
September 2006
Preliminary draft; not to be quoted
On the Analysis of Firms’ Cash Flows
Abstract
This paper revisits the why's and how's of cash flows analysis in an equity valuation
context. We argue that the high incidence of non-recurring items and, more generally, the
ambiguity inherent in accruals force analysts to pay attention to cash flows. However,
GAAP’s statement of cash flows embeds several drawbacks. In particular, it lacks a clear
“bottom-line” directly comparable to net income per GAAP. We therefore propose a
model of Modified Cash Accounting (MCA) which presents an income statement without
accruals. This approach differs from GAAP’s in major regards, including the concept of
cash and how one distinguishes operating from financial activities. The paper discusses
how one motivates, prepares, interprets and uses such a statement to evaluate a firm’s
quality of earnings and estimate its core (operating) earnings. The paper thus provides a
practical financial statement analysis tool.
1
On the Analysis of Firms’ Cash Flows
I. Introduction and Summary
A major reason equity analysts evaluate firms’ cash flows stems from analysts'
perceptions that GAAP’s income statement builds in intrinsic problems. At least two
issues concern analysts: The material incidence of non-recurring items (such as
restructuring charges) and a general suspicion that accruals reflects “arcane” accounting
principles, or even earnings management, no less than economic realities.1 These
characteristics of GAAP cause ambiguities when analysts' try to identify a core earnings
starting point for purposes of forecasting subsequent periods' EPS. They also give rise to
issues falling under the heading of “quality of earnings”. There should be no surprise that
analysts’ thus turn their attention to the statement of cash flows to better understand the
income statement.2
A shift of the analysis to GAAP’s statement of cash flows precipitates a new set
of problems. The concept of cash, for example, is arguably too narrow. It can exclude
what would seem to be relevant transactions (like property acquisitions in exchange for
shares issued rather than cash). The classification between operating versus financing
activities can also be problematic (like the treatment of changes in accounts payable).
Most important, however, the cash flow (CF) statement does not supply a bottom-line
which can logically be compared to the bottom line in the income statement. Textbooks,
and practice, do not spell out how one conceptualizes the analysis of cash flows in a
systematic fashion. There is a general notion that “cash due to operations” is central, but
In this paper, we use the term “non-recurring items” in a generic manner that corresponds to special items
per Compustat. However for the purposes of our analysis there is no need for a specific definition.
2
Analysts are increasingly issuing cash flow forecasts for firms with large accruals, volatile earnings and
heterogeneous accounting choices (DeFond and Hung, 2003). A growing number of firms are also
voluntarily providing management cash flow forecasts (Wasley and Wu, 2006).
1
2
what to do next seems to be a big question. It follows that there is great uncertainty as to
how an analyst is supposed to use the data in the statement of cash flows to enhance her
understanding of the income statement.
This paper revisits the problem of how a firm's cash flows can be analyzed. We
propose a mode of analysis which we will refer to as Modified Cash Accounting (MCA,
for short). MCA's structural underpinnings work like the “regular” financial reporting
model, and it relies on debits equal credits to articulate the financial statements. The
critical ingredient pertains to the assets/liabilities recognized: They include only cash and
other items that can be viewed as approximate cash equivalents. Hence the word
“modified” refers to an awkward expression which suggests that there can be
assts/liabilities other than cash because these are sufficiently similar to cash. The balance
sheet foundation serves as a means to an end: The preparation of an income statement
without accruals.
Any implementation of MCA depends, of course, on which assets/liabilities
qualify as approximate cash equivalents. We discuss this issue extensively, and tie it to
existing disclosures in 10-Ks and Qs to make the point that the proposed MCA can be
applied as a practical financial tool. Though the initial step deals with GAAP’s balance
sheet, there will be a need for other kinds of data to derive MCA's income statement.
Some of these may be found in the statement of cash flows as well as the statement of
changes in owners' equity. That said, we underscore that MCA should not be viewed as a
mere “reshuffling” of lines in GAAP’s CF statement.
Our proposed MCA income statement has not only a bottom-line -comprehensive earnings on a cash and approximate cash equivalent basis -- but it also
3
identifies various line items and sub-totals. Consistent with practice and financial
statement analysis text-books (e.g. Penman 2006), our MCA income statement
dichotomizes between dollar amounts due to operating as opposed to financial activities.
A key sub-total identifies income due to operations on a cash (and approximate cash
equivalent) basis. More generally, our MCA income statement can be juxtaposed and
compared to a GAAP-based statement of income flows. Differentials in the line items and
sub-totals will pick up effects due to accruals. This possibility set the stage for an analysis
of the firm's quality of earnings and, in a similar vein, it should provide a richer
understanding of a firm's quality of earnings and the estimation of core earnings.
The outline of the paper is as follows. To motivate the nature of the demand for
an analysis of cash flows, Section II discusses problems with GAAP earnings. In
particular, the effects of non-recurring items and accruals more generally, work against
finding a measure of a firm's core earnings. Section III develops the idea of MCA. We
focus on the relevant implementation aspects of the model which begin with approximate
cash equivalent assets/liabilities on the balance sheet. We then proceed to make necessary
adjustments for shareholder related transactions followed by an elaboration on the
specific line items in MCA. Of particular importance will be our argument that nonrecurring items in MCA-based income statement pose less of a problem as compared to a
regular GAAP-based income statement. Section IV discusses how firm growth influences
a MCA-based income statement. We first argue that for a firm in a steady state, earnings
on a MCA basis should approximate core accrual net income. We then adjust for the
growth component in order to arrive at a cash earnings measure that will be comparable
to core earnings. We also discuss how the cash earnings measure is a useful starting point
4
in equity valuation and quality of earnings analysis. Section V provides concluding
remarks.
II. Why GAAP’s Earnings lead to a Cash Flows Analysis
As a first step, equity analysts usually evaluate the income statement to estimate
the firm's core earnings, or earnings on a recurring basis. The task is crucial because
analysts need to forecast subsequent earnings; thus it makes sense to use a measure of
recurring earnings as a point of departure. That said, how to deal with non-recurring
items is no easy matter for a variety of reasons. The very definition of a non-recurring
item, for example, is beset with problems because many reasonable non-recurring items,
like gains/losses on property dispositions, can occur with considerable frequency. It is
also well-known that non-recurring items can be part of selling, general and
administration (SG&A) expenses and identified only through a careful reading of the
footnotes. Yet another point pertains to the possibility of the firm having deliberately
managed the non-recurring part of earnings within a period. Analysts clearly face a
daunting problem when they address the question: So what do we have in terms of nonrecurring items in the current period?
An analysis of the current income statement does not suffice to assess the
implications of non-recurring items. Past non-recurring items have a feed-forward effect
on subsequent recurring earnings. Write-offs and the “big bath” illustrate the phenomena
since these kinds of charges reduce subsequent expenses like depreciation and cost of
goods sold. Attempts to adjust the current income statement for past non-recurring
charges run into the vexing problem of assessing the extent to which they were
5
“excessive” and thereby provide direct expense-relief for the current period.
Excessiveness is not easy to pin down, but it is likely to occur if there are no cash flows
involved and the accounting is conservative. Few analysts seem to tackle this insidious
problem head on by estimating an actual dollar amount of current period expense-relief.
Instead they tend to incorporate the problem into the realm of quality of earnings
assessment. That is, past non-recurring charges lowers the quality of current earnings
which makes it more difficult for the firm to grow its earnings. Using non-quantifiable
judgments, the analyst then proceeds to forecast earnings for subsequent periods. Thus
the complexity of the task to a substantial degree depends on the incidence of current and
past non-recurring items.3’ 4
Non-recurring items can in many cases be thought of as a symptom of how
accruals more generally can be misleading and even deliberately manipulated so that
posted earnings meet objectives. Accounting research has over the years provided
extensive evidence to that effect.5, 6 Analysts and money managers have without doubt
paid attention to parts of this research. The so-called accruals anomaly is widely known
outside academia, though, of course, even if this research had not existed analysts would
Elliott and Hanna (1996) note an increasing trend of “recurring non-recurring” write-offs. Moehrle (2002)
and Burgstahler et al. (2002) provide evidence of inter-temporal shifting of expenses. Firms may also
attempt to improve core earnings and/or meet analyst forecasts by classifying regular expenses like cost of
goods sold and SG&A as special items (McVay, 2006).
4
Analysts are adept at identifying persistent non-recurring items and incorporate them in their measure of
street earnings (Gu and Chen, 2004).
5
Sloan (1996) documents the “accrual anomaly” by showing that the accrual component of earnings is less
persistent than the cash flow component of earnings. Subsequent studies find that the accrual anomaly is
distinct from the post earnings announcement drift (Hribar and Collins, 2000), analyst forecast revisions
(Barth and Hutton, 2004), value-glamour anomaly (Desai et al., 2004) and is mainly attributable to
discretionary accruals (Xie, 2001). Fairfield et al. (2003) suggest that the low persistence of the accrual
component is part of a more general growth related effect while Richardson et al (2005) attribute the
anomaly to temporary accounting distortions.
6
Research on earnings management finds accrual manipulation to be the primary tool used by firms to i)
show positive profits, ii) avoid earnings decreases and iii) meet or beat analysts’ forecasts (Hayn, 1995;
Burgstahler and Dichev, 1998; Degeorge et al., 1999).
3
6
still be aware of the arcane and fragile nature of accruals. The question has always been
what to do about it. Given the accounting equation earnings = cash flows + net periodic
accruals, it goes almost without saying that analysts must consider the firm's cash flows.
Most analysts still probably regard the evaluation of earnings as more important than
cash flows, but at a minimum, analysts expect the reported cash flows to bear on the
firm's (i) quality of earnings and (ii) the possibility of deliberate distortions in the income
statement.
When the analyst shifts her attention from the income statement to the statement
of cash flows another set of problems emerge, none of which have a generally agreedupon solutions. First, GAAP conceptualizes cash in very narrow terms: Cash in a literal
sense must have been exchanged for an effect to take place in the statement. The issuance
of shares in exchange for another firm's equity, for example, has no effect on the
statement. Common sense economics would suggest that there is an investing as well as
financing transaction as if cash were involved in two distinct transactions. Second, there
is confusion as to the appropriateness of the definitions of the three organizing activities,
namely, (current) operating, investing, and financing. Examples abound, the treatment of
interest expense (operating) and marketable securities (investing) being perhaps the most
obvious ones. But subtler questions can also be raised. Why not treat changes in accounts
payable as a financing activity? The same questions can be raised with respect to account
receivables if these are of high quality and pretty much could be sold for face value, the
proceeds of which can then be used to pay off bank borrowings. These kinds of questions
certainly show that analysts have their work cut out if they wish to recast GAAP’s
7
statement of cash flows in what they consider a preferable framework. Neither practice
nor textbooks provide much of a direction in this regard.
Quite aside from the points raised in the above paragraph, a much broader
problem lurks in the background. How can we arrive at a bottom-line in a schedule that
spells out a firm's cash flows? Analysts and textbooks generally agree that cash provided
by operations serves as a natural starting point. But what to do next introduces
considerable ambiguity because there is no guiding principle or concept as to what needs
to be accomplished. “Where should the analysis of cash flows begin and end?” is the
question that must be answered in addition to the obvious fact that the notion of cash
must be operationalized. In this paper we argue that the question can be answered if one
simply keeps in mind that all financial statements can be prepared on cash (however
defined) basis. Thus one can derive a schedule of cash flows directly comparable to a
regular accrual and cash based income statement by simply changing the underlying
recognition principle. The paper proposes such an approach, and it can be thought of as
falling under the heading of Modified Cash Accounting. In describing our proposed
implementation we show that many issues raised with respect to GAAP’s statement of
cash flows have an answer.
III. A Model of Modified Cash Accounting (MCA)
MCA, in general, conceptualizes an accounting in which only cash-based
transactions show up in the financial statements. It is an integrated framework in the
sense that it depends only on the usual three basic statements (balance sheets, income
statements, and statements of changes in owners' equity), with the clean surplus relation
8
ensuring debits = credits. The word “modified” refers to the idea that the model does not
have to consider “cash” in a strict, narrow sense; it can be extended to treat a variety of
assets and liabilities “as if” they were cash, e.g., marketable securities, bank borrowings
and accounts payable. We will discuss this practical aspect of defining approximate cash
assets/liabilities later in this section.7 At any rate, via the clean surplus relation, the
calculation of the grand bottom line for the income statement is now straightforward
given appropriate allowances for dividends and other transactions with the (common)
shareholders. Thus one obtains “Comprehensive Earnings on a Cash and Approximate
Cash Equivalent Basis”, to use a long and awkward phrase. We will simply refer to this
grand bottom line as Comprehensive Cash Earnings, or CCE.
Our proposed model of MCA goes beyond the CCE number because the model
provides various line items and subtotals in the income statement (the consecutive
balance sheets are mere means to an end, of course, so these do not classify items into
groups). Specifically, the income statement dichotomizes between operating vs. financial
activities comprehensively. Hence, the net financial expense added to CCE determines
the (net) Cash Earnings provided by operations. There will also be a “top” line
corresponding to sales, and operating expenses will be split into “current benefits” and
“long term benefits”.
The remainder of this section spells out the implementation of our proposed MCA
model. Most of this discussion bears on the judgments required as to what constitutes
7
MCA differs from so-called fair value accounting (FVA) in that the latter assigns fair values to a broader
set of assets (like inventories) and liabilities. From a user-oriented perspective, one can perhaps also argue
that FVA underscores the information in the balance sheet, whereas in contrast MCA relies on the
consecutive balance sheets as means to an end, namely, the development of an income statement. For
example, no significance should be attached to the possibility that a firm's net book value might well be
negative for MCA although on a GAAP basis the firm is perfectly solvent and profitable. In our MCA, of
interest is only the change in the cash based book value.
9
approximate cash equivalent assets/liabilities. We invoke a perspective focusing on how
one extracts the relevant items and data from GAAP (10-Q and K) reports. This approach
serves the useful purpose of aiding practical quality of earnings analysis. But it goes
almost without saying that the model could also be implemented from the ground up via
transaction by transaction book keeping procedures.
We begin by enumerating the steps required to implement our MCA model.
1. Identify cash and approximate cash equivalent assets/liabilities to derive the net
worth on a cash (and approximate cash equivalent) basis, for the starting and ending
balance sheets.
2. Derive Comprehensive Cash Earnings, CCE, by taking the difference in the net
worth on a cash basis and then adding the period's dividend distributions net of
capital contributions.
3. Identify all financial revenues/expenses on a cash equivalent basis, net of taxes.
These will primarily relate to marketable securities and interest bearing financial
obligations.
4. Derive net operating earnings due to operations by adding the net financial expense
to CCE.
5. Identify all the accounts that reflect transactions with customers. These include
sales revenues, accounts receivables, allowance for uncollectibles, and (potentially)
deferred revenues.
6. Derive the top line in the MCA income statement -- cash and their equivalents
received from customers -- by adjusting net revenues per GAAP for changes in
10
accounts receivable ( if deemed to be a non-cash equivalent ), allowance for
uncollectibles, and deferred revenues.
7. Derive total sales sustaining expenses, SSE, on a cash basis by deducting cash
operating earnings from the top line (cash received from customers).
8. Identify SSE with long term benefits, such as capital expenditures (net of
dispositions) and acquisitions of companies.
9. Derive the part of SSE that benefits only the current period by deducting those SSE
with long term benefits from total SSE.
The Cash Earnings Statement, CES, can thus be formatted as below in Table I.
< TABLE I ABOUT HERE >
It is apparent that the above CES differs in much respect from GAAP’s statement
of cash flows. Most important, the concept of “cash” is entirely different since out
approach is much broader by allowing for approximate cash equivalents and, further, our
MCA also picks up on “as if” transactions like the issuance of shares in exchange for an
acquisition. Our notion of operating versus financial activities also differs from that of
GAAP. And we maintain a sharper distinction between the effects of current versus noncurrent operating cash (and cash equivalent) flows as compared to GAAP’s way of
looking at it. On a more basic level, however, our approach focuses on deriving an
income statement, with its related bottom line, which has eliminated all accruals. It can
be juxtaposed and compared directly to not only past CES but also the income statement
per GAAP. The question “How much of GAAP’s earnings is due to accruals?” now has a
11
numeric answer. We discuss comparison issues in the next section to get a practical
handle on the assessment of a firm’s quality of GAAP earnings.
The remainder of this section deals with judgments necessary to implement our
MCA. The steps above capture the general framework and its broad objective, but, as is
always the case for any practical accounting model, there will be more than a few devils
in the details. In particular we must consider the practical meaning of approximate cash
equivalents, of course, as well as the distinction between financial line items and the two
kinds of operating line items. And the accounting for transactions with owners also raises
a few issues.
Turning our attention first to the starting point, the balance sheet and its
recognized cash and approximate equivalents, the following Table II, lists the natural
candidates. We stress that these are potential assets/liabilities, and thus they may be
pruned because of facts, judgments, and practical considerations. Materiality comes into
play too; here we stress that in this regard the correct dimension concerns the dollar
change over the beginning and ending date. Further, there is always the possibility that
the list can be expanded though, in our view, this seems unlikely for most industrial
firms.
< TABLE II ABOUT HERE >
On the asset side the first two items are unambiguous, as long as the marketable
securities are reasonably liquid and marked to market. These conditions should generally
be in place. We further note that marketable securities could be long term as well as short
12
term. The distinction between available-for-sale and other categories of securities owned
should not be of any significance either.
Accounts receivable qualify only if they are judged to be of high quality. This
condition should generally be met if the allowance account has a small balance relative to
the gross accounts receivable, such as less than 4% say. In other words, if the firm
wanted to sell the receivables at approximate net value, that should not be deemed a
problem if they are to qualify.
Finance receivables are conceptually about the same as accounts receivables. The
relevant information, again, is a question of the history of collections and the use of
allowances.
Equity method investments pose a non-trivial problem unless they can be marked
to market in the MCA through use of footnote information in the GAAP disclosures.
Otherwise it seems more reasonable to disqualify such items. In general, the problem
should be mitigated due to immateriality.
Pension assets, if liquid and properly disclosed, naturally qualify. There are no
compelling reasons that this judgment should depend on the treatment of related pension
obligations. For purposes of income measurement, as opposed to assessments regarding
what may happen in case of bankruptcy and reorganization, marketable securities for
funding pension obligations are a perfect substitute for regular marketable securities.
Hence their balances add.
Turning our attention to the liabilities, the interest bearing debt items should
obviously qualify. This takes care of loans, bonds payable, and lease payables (unless the
latter for some reasons have dubious accounting). Preferred stock should be included
13
whenever the accounting is from the perspective of the common shareholders (which is
of course the case in a context of equity valuation).
Current, but not deferred, income taxes payable can be viewed as a relatively
clear-cut liability with few accounting problems. It is a “hard” liability that must be paid
in the near future since it pertains to taxes actually owed to various governmental
jurisdictions. Disclosures may often perhaps be less than satisfactory for this item, but
that is a different matter (10-Qs seem to be particularly problematic).
Accounts payable would also seem to be essentially straightforward and involve
few estimation issues. A possible exception occurs if the payables have to net out
material estimates or product-returns to suppliers.
Interest payables obviously qualify. Such items are rarely material, however.
Pension obligations, and similar benefits, require some tricky judgments not easily
resolved. One argument is simply to dismiss them because the underlying estimation
procedures are inherently too subjective and arbitrary. The accounting for projected
benefit obligation has somewhat less than a stellar reputation. On the other hand, one may
argue that such is not the case for the vested projected benefit obligation, which is
available in footnotes at least on an annual basis.
Having completed our discussion about practical balance sheet issues, the income
statement comes next. Cash earnings must be sliced into its components.
(Comprehensive) net cash earnings is taken as a given, by steps 1 and 2, and two SSE
items, 7 and 9, derive as “plugs”. Hence we need to discuss the financial expense, the top
line sales, and the operating expenditures with long term benefits.
14
Starting with the net financial expense, this item includes those interest revenues
(and dividends) and expenses due to the recognized underlying assets/ liabilities, i.e.,
marketable securities and interest bearing debt. These pose no conceptual problems. The
same can be said for the realized and unrealized gain/ losses related to the underlying
securities. Further note that the occurrence and magnitude of these gains/losses have no
effect on the cash earnings due to operations. Hence the fact that bonds payable have not
been marked to market does not affect the operating earnings; an “error” in the bottom
line is offset by an identical “error” in the financial item.
So far the financial item has been straightforward. However, the recognition of a
pension obligation, if done, raises a measurement problem. It can of course be put to zero
with some reason since one can think of the underlying economic “cost” as an add on to
pension payments made in cash, i.e., they are effectively operating in nature. But that
may seem illogical unless one then also classifies the revenues due to pension assets as
operating too. Perhaps the most sensible approach is to impute an expense due to the
pension obligations. This captures the broad spirit of our MCA: One can think of all
recognized assets/liabilities as defining the financial assets /liabilities. Taking this
thinking one step further one can then also impute income/expenses to other items like
accounts payable and cash though they have no direct effect on the financial income.
The financial item, whether positive or negative, should be shown net of taxes to
maintain a clean distinction between operating and financial activities. Thus the question
of intra-period tax allocation arises, and one naturally applies the statutory rate to
financial activities. Given current tax laws, which gives few breaks on “regular” financial
activities, this approach would seem to be sensible. The fact that the tax-return may differ
15
due to, say, unrealized gains/losses does not suggest a refinement in the measurement of
taxes. It is important that one keeps in mind that MCA is a way of measuring income
given certain asset/liability recognition rules. There is therefore no requirement per se
that the line items in the income statement must describe the cash flows in a narrow and
precise sense.
To determine the top line, cash (and their approximate) equivalents received from
customers, GAAP’s total revenue item provides the starting point. Thereafter one adds
any increase in deferred revenues. Finally, whether or not one deducts the increase in the
accounts receivables (net of allowances) depends on the status of the accounts receivables
in the balance sheet: If they have been excluded, then of course the adjustment is
necessary. No other measurement problems would seem to arise in the context of revenue
recognition, with the possible exception of GAAP’s revenues being affected by customer
return of products.
After having determined cash operating earnings and the top line, the line sales
sustaining expenditures, SSE, falls into place as a plug. The terminology and approach
suggests that a firm's operating expenditures exist for purposes of generating sales for
future periods as well as the current one. And some of these expenditures may even be
due to sales made in the past. Thus SSE ends up as an amalgam of all sorts of
expenditures that can be vastly different, like the final payment on an operating lease that
has expired, acquisition of a subsidiary, purchases of inventories, legal fees, payments to
employees in a restructuring situation, environmental cleanups etc., etc. There can also be
inflows of cash (or their approximate equivalents) due to transactions other than those
16
related to customers such as sale of property. These inflows are of course netted against
the outflows.
Given the rich set of transactions that fall into SSE it is hard to say how much
should be viewed as SG&A or any other specific category encountered in a GAAP based
income statement. Nonetheless, SSE is an interesting line insofar that it is crucial element
of a firm's periodic performance which is measured relatively uncontaminated by the
ambiguities inherent in the use of accruals for cost of goods sold, depreciation,
impairments etc. The trade-off in terms of useful information in a regular income
statement vs. an MCA statement becomes apparent. The accrual approach permits a much
richer set of interpretations as to what has occurred, but only at the cost of depending on
inherently ambiguous numbers. MCA rests on “harder” data and facts, but much of the
“story”, which the accruals purportedly would have captured, has now been left out.
To enhance the usefulness of the MCA statement we suggest that it helps if one
identifies those SSE with “long term benefits” and then let the remainder represent those
SSE with “current” benefits.( The elaborate term “SSE with current and past benefits” is
perhaps more descriptive). In this spirit one can measure SSE with long term benefits by
adding up capital expenditures net of dispositions, acquisitions of firms, R&D and
advertising & brand maintenance expenditures. Such numbers are readily extracted from
firms’ reports, of course.
A few comments about the statement of changes in owners’ equity may be
appropriate since the step 2 computation relies on a measurement of dividends net of
capital contributions. In general, it should conform to that of GAAP as long as the
transactions only involve cash dividends, treasury stock transactions and the issuance of
17
shares. An exception occurs for changes in minority interest which should not be present
because the accounting focus is from the viewpoint of the common shareholders. Also,
any preferred stock dividends and changes in the balance of preferred stock is part of
financing activities, not equity. A more subtle issue emerges because MCA requires that
approximate cash equivalent estimates should be used for equity transactions. This
measurement attribute rules out “pooling of interest” accounting for acquisitions of
course. But this accounting is no longer permitted in the US anyhow. The accounting for
compensation options, however, cannot be avoided. In our MCA framework they should
be accounted for using exercise-date accounting, in which case the capital contribution is
measured by the market value of shares issued at the exercise date (net of tax-effect like
GAAP). In other words, this part of the MCA accounting is essentially how the tax law
accounting works for options. From this perspective it is indeed straightforward. On the
whole the same can be said for the measurement of the total net dividend adjustment
necessary for a measurement of the (comprehensive) net cash earnings. Table III
summarizes the schedule that must be implemented for our MCA accounting.
< TABLE III ABOUT HERE >
IV. The MCA Statement and the Analysis of Quality of Earnings
This section looks at the uses of the MCA income statement, especially as a
complement to better “understand” the GAAP income statement. Possibilities to do so are
of course innumerable if one includes all sorts of contextual information (like
management turmoil or a recent history of SEC investigations), inside and outside the
18
financial reports. Here we confine the discussion to make a few general points that we
believe are helpful in the assessment of any firm’s performance. Our focus will be on
what is often called the quality of a firm’s earnings. The idea is that the reported earnings
can be misleadingly high (or low for that matter) because the periodic total (net) accrual
is excessive (or deficiently low).The words “misleading” and “excessive” as used here
means that the firm's current earnings are unlikely to be sustainable if the underlying
operating economics of the firm remains roughly the same. We think of this as being an
adequate definition of the concept of quality of earnings. The distortions, if present, may
be intentional and possibly even violate GAAP, or they may be inadvertent and simply
due to the arcane nature of implementing GAAP. As a first cut of financial analysis, this
motivational aspect is of no significance when one tries to assess a firm's quality of
earnings.
The concept of quality of earnings is initially best appreciated if one considers
what happens in a so-called steady state setting, without any net financial item to keep
matters simple. Steady state means no growth and the economics of the firm remains
unchanged from one period to the next. Under such stylized conditions it is well-known
that cash accounting works at least as well as accrual accounting because under cash
accounting and consistently applied accrual accounting, net income is the same for all
periods. Cash accounting works better, of course, to the extent there is some discretion in
the implementation of the accruals across periods. Now there will be inconsistencies
across periods: The average accrual will be zero, but in some periods the accrual will be
positive (negative) while in others reversals must take place and the accrual will then be
negative (positive). Thus positive accruals correspond to low quality earnings, which is
19
an obviously sensible conclusion, give the underlying assumption of the firm being in a
steady state.
Except in an ideal world, firms cannot exist in a steady state due to inherent
uncertainties in the environment if nothing else. That said, the previous discussion
suggests that one may consider an approximate steady state as occurring when the top
line remains unchanged. In the absence of information to the contrary, one should now
expect the (total net) accrual to be zero.8 To make a first cut assessment to that effect one
can thus simply compare the two comprehensive earnings numbers, MCA vs. GAAP.
More refined analysis can thereafter address if the difference in the two earnings numbers
has a reasonable explanation. Note that this mode of analysis is oriented to determine the
quality of GAAP earnings rather than to come up with a measure of the firm's recurring
earnings. That takes additional analysis as we discuss below. Nevertheless, it would seem
to be quite useful to come up with a measure of the net accrual when there is no sales
growth. This approach takes a good stab at sorting out the net effect due to current and
past non-recurring accruals when other kinds of accruals are deemed to be appropriate. In
many circumstances one can reasonably hypothesize that these should approximate zero
when sales are essentially flat.
How do we modify our thinking about a firm’s quality of earnings if it is
apparently growing? If sales are growing by a material percentage, then there is every
reason to expect that the firm must invest in inventories, PP&E etc. GAAP’s earnings
should accordingly, at least in general, exceed the MCA income because the growth in
This observation about a firm’s steady state can be viewed as an application of the “cancelling error”
concept pertaining to two consecutive balance sheets. That is, given a steady state one can reasonably
hypothesize that the “true missing values” in the two balance sheets are roughly the same and, thus, to a
corresponding degree there will be no impact on the measurement of earnings.
8
20
the underlying investments ought to show up as a positive total net accrual. This
observation raises the question if one can estimate what the accrual ought to be, so that
one can thereby adjust MCA’s earnings and make it more comparable to GAAP’s
earnings. One way of proceeding is to assume that for each dollar of sales there should be
some underlying supporting investment in (total) net operating assets (or NOA, the often
used textbook acronym). In the jargon of financial statement analysis, the practical
hypothesis is that the firm’s (net operating) asset-turnover ought to be fixed. An order of
magnitude of such a number can be estimated by looking at the firm's recent history of
NOA relative to sales. Hence the “proper” estimate of the dollar amount (total net)
periodic accrual that should be added to MCA’s net earnings equals the assessed
sales/NOA ratio multiplied by the dollar increase in sales. While this approach is hardly
one of a hard science, it holds out the possibility of getting some insights about the
quality of a firm's GAAP measured earnings.9 We suspect that the insights can be quite
dramatic if the firm has a history of using non-recurring charges.10
Rather than assess the firms quality of GAAP earnings, one can try a more
ambitious route and estimate a firm's recurring earnings using the MCA statement as the
main source of data. Changing the objective of the analysis does affect the need for an
adjustment due to growth in sales or growth more generally. The discussion in the
9
Two fundamental principles of value, when combined, imply the imputed accrual concept: Accrual =
growth x Value. First, the present value of cash flows determines value. Second, the capitalization of a
perfect measure of earnings also determines value. The formal argument runs as follows. Let g, c, V, acc, r
denote, respectively, growth, cash flows, value, accrual, and a discount factor. The first principle implies V
= c/(r-g). The second implies V = (acc+c)/r, where acc+c defines perfect earnings. Combining these two
expressions leads to acc = gxV. Note that in this context, V and r are notional accounting constructs as
opposed to market related quantities.
10
Starting from MCA, Ohlson (2006) develops a full-fledged accounting model which includes accruals
and satisfies clean surplus. The basic idea is that SSE with long term benefits can be (i) capitalized each
period and (ii) also for each period, a certain amount is passed on into the income statement as a period
expense (which of course can differ from the debit to the account). The credit to the account is obtained via
a rule that matches an appropriate expense to sales. The scheme exploits concepts inherent in inventory
accounting.
21
previous paragraph is therefore as valid as ever. But the focus on estimating recurring
earnings from the MCA statement raises an additional problem: just like the GAAP
statement, some of the MCA lines can include non-recurring items. A litigation
settlement is one example. Inflows due to dispositions of property are another. However,
non-recurring items in the MCA statement should be nowhere near as material as nonrecurring items in the GAAP statement since the related accruals are pervasive in GAAP.
The case for MCA becomes even stronger if one considers the non-recurring financial
items. Financial non-recurring items in MCA can generally be estimated without much
difficulty. A simple and direct approach can estimate the recurring part by applying some
average “normal” earnings rate to the net financial asset/liability. In sum, one can
estimate recurring earnings by adding three numbers: operating cash earnings, plus an
accrual due to the sales growth, plus an estimate for a recurring financial item. The
resulting bottom line might well give a much better indication of a firm’s ability to create
wealth for the common shareholders as compared to the GAAP statement, especially if
there has been a long and nasty history of non-recurring items like impairments and
restructuring charges.
Capital expenditures, and similar transactions with long term benefits (LTB), have
an immediate negative impact on the current period’s cash earnings. This inherent
consequence of MCA flags another danger of using such a statement to estimate earnings
on a recurring basis. Cash earnings may look high simply because the capital
expenditures (and other LTB expenditures) have been relatively small. To assess whether
such a hypothesis makes sense in a particular case the analyst can compare SSE-LTB in
the current period with those of prior periods. SSE-LTB can also be normalized by the
22
top line and then compared across periods to inform on whether the current SSE-LTB has
aided or depressed the bottom line. Another angle to the same problem works with
assumption that the accrual in question -- the depreciation expense -- is more informative
than the capital expenditure. All of these modes of analysis should allow the analyst to
make some reasonable adjustments to SSE-LTB to better estimate earnings on a recurring
basis. The difficulties and complex judgments involved are not obviously worse than
relying on GAAP’s earnings after having stripped out those line items judged to be nonrecurring.
V. Concluding Remark
We have developed the proposed MCA from the perspective of someone who
wants to apply it using actual financial reports. Depending on the degree of detail and
“accuracy” desired this can generally be done without much time. Hence an analyst has
an additional tool available to better understand the effects of non-recurring items and,
more broadly, the effects of accruals on a firm's posted measures of earnings. We believe
this can be helpful in an equity valuation context as a practical matter. That said, it should
be apparent that the analysis raises a question that bears on a decision accounting
regulators have to face every decade or so: Given that users of financial statements seem
to demand a statement that describes a firm’s cash flows, how can such a statement best
be structured? The answer to such a question depends, of course, on why the users feel
such a statement fills a need. If, in fact, the demand for the statement reflects users’
queasiness with accruals and the difficulty of estimating a firm’s recurring earnings,
especially on an operating basis, then it follows logically that the statement should be
23
organized and conceptualized much like any income statement. The crucial qualification
is that accruals have now been proscribed. Paying heed to this observation would seem to
be much more important than the precise definitions of cash vs. accruals and operating
vs. financial activities.
24
References
Burgstahler, D. and I. Dichev. 1998. Earnings management to avoid earnings decreases
and losses. Journal of Accounting and Economics 24 (1): 99-126.
Burgstahler, D., J. Jiambalvo, and T. Shevlin. 2002. Do stock prices fully reflect the
implications of special items for future earnings. Journal of Accounting Research 40 (3):
585-612.
Collins, D. and P. Hribar. 2000. Earnings-based and accrual-based market anomalies:
One effect or two? Journal of Accounting and Economics 29 (1): 101-123.
DeFond, M. and M. Hung. 2003. An empirical analysis of analysts’ cash flow forecasts.
Journal of Accounting and Economics 35 (1): 73-100.
Degeorge, F., J. Patel, and R. Zeckhauser. 1999. Earnings management to exceed
thresholds. Journal of Business 72: 1-33.
Desai, H., S. Rajgopal, and M. Venkatachalam. 2004. Value-glamour and accruals
mispricing: One anomaly or two? The Accounting Review 79 (2): 135-155.
Elliott, J., and J. D. Hanna. 1996. Repeated accounting write-offs and the information
content of earnings. Journal of Accounting Research 34 (supplement): 117-134.
Fairfield, P., J. Whisenant, and T. Yohn. 2003. Accrued earnings and growth:
Implications for future profitability and market mispricing. The Accounting Review 78
(1): 353-371.
Financial Accounting Standards Board (FASB). 1987. Statement of cash flows.
Statement of Financial Accounting Standards No. 95. Norwalk, CT.
Gu, Z. and T. Chen. 2004. Analysts’ treatment of nonrecurring items in street earnings.
Journal of Accounting and Economics 38 (1-3): 129-170.
Hayn, C. 1995. The information content of losses. Journal of Accounting and Economics
20 (1): 125-153.
McVay, S. 2005. Earnings management using classification shifting: An examination of
core earnings and special items. The Accounting Review 81 (3): 501-531.
Moehrle, S. 2002. Do firms use restructuring charge reversals to meet earnings targets?
The Accounting Review 77 (2): 397-413.
Ohlson, J. 2006. A practical model of earnings measurement. The Accounting Review 81
(1): 271-279.
25
Penman, S. 2006. Financial statement analysis and security valuation. Third edition.
McGraw Hill.
Richardson, R., R. Sloan, M. Soliman, and I. Tuna. 2006. The implications of accounting
distortions and growth for accruals and profitability. The Accounting Review 81 (3): 713743.
Sloan, R. 1996. Do stock prices fully reflect information in accruals and cash flows about
future earnings? The Accounting Review 71 (3): 289-315.
Wasley, C. and J. S. Wu. 2006. Why do managers voluntarily issue cash flow forecasts?
Journal of Accounting Research 44 (2): 389-429.
Xie, H. 2001. The mispricing of abnormal accruals. The Accounting Review 76 (3): 357373.
26
Table I
Cash Earnings Statement (CES)
Steps
Cash Equivalent Collections From Customers
5, 6
SSE
XXX– Current Benefits
Current Operating Cash Earnings
<XXX>
SSE – Long Term Benefits
<XXX>
7, 8
XXX
4
<XXX>
3
XXX
1, 2
Operating Cash Earnings
Net Financial Expense
Comprehensive Cash Earnings
9
XXX
27
Table II
Potentially Qualifying Assets & Liabilities in the Balance Sheet
Assets
Liabilities & Stockholders’ Equity
Cash & Other Equivalents (per GAAP)
Loans and Similar Debt
Marketable Securities
Accounts Payable
Accounts Receivable
Current Income Taxes Payable
Financial Receivable
Interest Payable
Equity Method Investments
Bond Payable
Pension Assets
Lease Payable
Pension/post retirement Liabilities
Preferred Stock
28
Table III
Statement of Changes in Cash Equivalent Stockholders’ Equity
 Cash equivalent Assets
 Cash equivalent Liabilities
XXX
<XXX>
 Common Stockholders’ Equity
Cash dividend
Stock issuances for cash or other assets
Treasury stock purchases
XXX
XXX
<XXX>
XXX
Treasury stock reissues
<XXX>
Employee stock option exercises (net of tax benefits)
<XXX>
Comprehensive Cash Earnings
XXX
29