11(d). Integral and Discrete Approach
Integral Versus Discrete Approach to Interim Reporting
In 1973, the Accounting Principles Board, in Opinion No. 28, mandated the integral approach to reporting quarterly earnings. Under that opinion the accounting period is defined as being one year and quarterly periods are to be regarded as "integral" segments of the annual period. The effect of that standard is to allow the allocation of period costs across interim periods on the basis of benefits received, time elapsed, or other even more arbitrary bases. The result has been allocation that in actuality has ranged from smoothing into outright manipulation of quarterly earnings. Expenses, such as advertising, research, maintenance, income taxes, and so on, are reported discretely year by year. Yet, on a quarterly basis, they are unabashedly smoothed and often in ways that appear dubious. [AIMR/FAPC92, p. 38]
In its early years, the FASB placed on its agenda a reconsideration of APBO 28. AIMR's Financial Accounting Policy Committee submitted comments strongly supporting the discrete method. But the project was removed from the FASB's agenda and APBO 28 continues to prevail. We believe that financial analysis is best served by financial reporting that reports transactions as and when they occur. If there is smoothing to be done, it is the province of analysts to do it. If there are financial reporting anomalies that are attributable to seasonality, it is far better to report and explain them than to conceal them with undocumented smoothing. Thus, we recommend changing interim reporting from the integral to the discrete method. [AIMR/FAPC92, p. 38]
[Context] Meeting of the Investor Discussion Group on March 17, 1993. Part of the meeting was devoted to the topic of auditor involvement. During the discussion on the scope of auditing, an investor referred to the integral versus discrete approach to interim reporting.
Committee/Staff/Observer
Should there be any auditor involvement with quarterly reports or once a year is enough? [Also included in 17(b) and 17(d)] [TI 3/17, p. 12]
Participant I-16
I think the quarterly issue has to be deferred towards the later discussion of what you want quarterly reports to be, in terms of whether they should be like annual reports, that is, discrete periods, or whether they should be an integral part of the whole year (thus, with some smoothing). You need to decide that before deciding whether they should be audited or not. [Also included in 17(b) and 17(d)] [TI 3/17, p. 12]
[Context] Meeting of the Investor Discussion Group on March 17, 1993. Part of the meeting was devoted to the topic of interim reporting.
Committee/Staff/Observer
Question 11, also on interim reporting, describes two views: the discrete view and the integral view. Accounting standards have adopted the integral view. Our question is: are you satisfied with interim reporting based on the integral view? If not, do you agree with its critics that accruals and estimates at the end of an interim period should be determined by the principles that apply to annual periods, and transactions should be reported in the interim periods in which they occur, and any smoothing to be done should be left to analysts? [TI 3/17, p. 42]
Participant I-16
I have some sympathy for the discrete view. It makes things more explicit; management would have to explain why something may look strange or unusually volatile. Under the integral system, you can cover up things for a long time using estimates. I would want to give a lot of thought to moving to the discrete view. If a company does something that makes sense and that's going to make the short-term numbers look worse, it will force the company to explain that to analysts rather than pretending that it didn't happen. [TI 3/17, p. 42-43]
Participant I-12
Most of us operate on the discrete view but I think there is an argument for the integral view. For example, in retailing you have seasonal factors, and in the financial services industry there is the issue of bonus compensation; you may get one terrific quarter where the bonus compensation that will be paid at the end of the year is going to be substantially higher than the company originally thought. So they'll raise each of the following quarters' estimates of bonuses; that tends to smooth a bulge. While we look at discrete periods, I think there's also an element of the integral view that makes some sense. [TI 3/17, p. 43]
Participant I-7
It depends on the industry. What happens if you're a military supplier and you're building a submarine that takes 3 years to deliver using the percentage of completion method. Or worse, if you are a supplier in the nuclear industry that may take 7 or 8 years to provide hardware for a site. In those cases, I have a problem taking one or the other view. [TI 3/17, p. 43]
Committee/Staff/Observer
In your example, I don't think the discrete method would negate the use of the percentage of completion method for that contract. [TI 3/17, p. 43]
Participant I-5
To some extent, when you have specific events causing longer term accruals, you end up treating years as part of an integral longer period. Quarters should obviously then also be part of a longer period. [TI 3/17, p. 44]
Committee/Staff/Observer
But we clearly have discrete years. You don't make an assumption at the end of the year that it is part of an integral era. The question is: do you cut it like that for a quarter? [TI 3/17, p. 44]
Participant I-5
What would be one instance of when you make an estimate for a quarter that you wouldn't make for a year? [TI 3/17, p. 44]
Committee/Staff/Observer
Advertising expenses, for example, where a company concentrates advertising in the second quarter and they argue that it benefits the whole year, so they spread the advertising costs ratably over 4 quarters. [TI 3/17, p. 44]
Participant I-16
I would like to know that a company is engaged in a major step-up in its advertising in the second quarter and not just find out that they're going to be up 5% for the year. If they're going to be up 50% in the second quarter, most likely it's some kind of new program and I'd like to ask them questions about it. Again, with the discrete viewpoint, you force more disclosures to be made. [TI 3/17, p. 44]
Committee/Staff/Observer
So you would expense it, but disclose it. [TI 3/17, p. 44]
Participant I-16
Yes, management will disclose things that is in management's interest to disclose. If something is disadvantageous by using the discrete method, they will disclose information about it. If you use the integral method and spread it out, there is no reason to disclose information about it. I would like to maximize the chance that it will be disclosed and be a subject of discussion. [TI 3/17, p. 44-45]
Committee/Staff/Observer
But if you could be sure that it would be disclosed, it wouldn't make any difference what the numbers were? [TI 3/17, p. 45]
Participant I-16
Yes. [TI 3/17, p. 45]
Committee/Staff/Observer
Let me give you another example: a major repair that happens in the second quarter and was not budgeted. Management says that it's clearly repair, not a capital expense. Management decides to spread the effect out over the next two quarters. That's a better example of the difference between the discrete and integral approaches. Should that repair be recognized as an expense fully in the second quarter or should it be spread to the second, third, and fourth quarter? [TI 3/17, p. 45]
Participant I-5
If the benefit you received from having made that repair is longer than for one year, then the repair should be capitalized. [TI 3/17, p. 45]
Participant I-16
The repair just restored you where you were; it doesn't increase value. [TI 3/17, p. 45]
Participant I-12
The big difference between the two is cash versus accrual accounting. If we had quarterly cash flow statements, a lot of these factors would be captured; for example, the compensation example, the advertising example, the repair example. We would know that the particular event happened in a specific quarter but we would also have statements that would allow us to do the trend line analysis that is essential to our work. [Also included in 5(c) and 11(c)] [TI 3/17, p. 45]
Participant I-16
The problem here is that we're ignoring one thing; we're assuming that everything else remains the same other than the accounting. If you change the accounting, you change people's behavior. If you can't spread the cost of that repair, you buy insurance to spread the cost. This thing would be the greatest thing for the insurance industry. The reason why I said that in theory I like the discrete approach but that I'm not sure, is because that management may react to it in a way that I'm not sure I want them to. There might be other things that might have real costs; the example of buying insurance to protect against things for which they really should not pay insurance premiums for, might be an unintended cost that the shareholders would have to bear. Another example: it might make a lot of sense to do that TV advertising in the second quarter; spot TV advertising time may have come down but the company may decide not to do that because they can't spread the cost over the full year. So they might pay more to buy spot advertising time in the third quarter. That's the kind of things I worry about; there may be some real consequences here as companies try to evade the consequences to a change to the discrete methodology. [TI 3/17, p. 46]
Committee/Staff/Observer
Having said all that, you're now on the FASB and you have one vote; discrete or integral? [TI 3/17, p. 46]
Participant I-16
I would want to hear the other side of the argument from somebody from the company; does it really create a problem for them? To me, it's a slight benefit to go to the discrete method, but it's not something I would insist on. [TI 3/17, p. 46]
Participant I-11
I lean toward the integral method. I'm also concerned about unanticipated consequences. I think about what happened in the wake of the postretirement benefits standard, where there have been some tremendous unanticipated consequences in terms of companies solving the problem by doing away with postretirement benefits. I don't think the FASB intended or anticipated that, but in reality it happened. The discrete method theoretically sounds better; I'm afraid though that none of us have thought through closely enough what it would mean. [TI 3/17, p. 46-47]
Participant I-12
Like [participant I-11], I would tend to lean toward the integral approach if I could get quarterly cash flow statements. [Also included in 5(c) and 11(c)] [TI 3/17, p. 47]
Committee/Staff/Observer
Assuming you couldn't? [TI 3/17, p. 47]
Participant I-12
There is a big issue with the unintended consequences and it concerns me. [TI 3/17, p. 47]
Participant I-5
I would be leaning toward the integral approach with greater disclosure, particularly if you can build the disclosure through the cash flow statement. But if you take away the cash flow statement, then I switch and go with the discrete method. [Also included in 5(c) and 11(c)] [TI 3/17, p. 47]
[Context] Meeting of the Creditor Discussion Group on March 11, 1993. Part of the meeting was devoted to the topic of interim reporting.
Committee/Staff/Observer
Question 12 can be characterized as integral versus discrete and basically ask the question that we dealt with in smoothing perhaps on a micro scale and says in effect you have a choice when you're doing interim financial statements to anticipate what the year as a whole is going to look like and set accounting accruals and conventions in anticipation of the year or you can account for things pretty much as they happen without anticipating reversals of losses and other kinds of things that balance out a year. Some people have characterized much of accounting currently under an integral notion, although I'm not sure that that's altogether applicable. The question we ask you is whether or not integral versus discrete, that is, smoothing across a year versus having everything essentially accounted for as it occurs makes a difference to you and, if so, which one do you choose? [TC 3/11, p. 50]
Participant C-4
We would choose the integral. We don't really look at rolling twelve month numbers. We set up our credit lines on a twelve month basis based on year end results and then we consistently measure to the year end results. So the integral view is obviously a lot more important to us in that type of analysis. [Also included in 11(b)] [TC 3/11, p. 50]
Participant C-11
I certainly want to know about the seasonality or cyclicality of important revenue and expense items but I am aware of certain things where the particular cost is based on an annual performance or estimate and you cannot necessarily know anything different from quarter to quarter unless the business really changes materially. And the ones that I've thought about over the years in that context are things like management bonus awards, which are based on an annual payment. Also the complex tax area where for many companies it's complicated enough to figure out taxes and having a rough estimate of understanding for the year and if that had to be thought about quarterly, it would produce a lot of stress. So my answer is there may be areas where it is not worthwhile to readjust everything quarter to quarter based on that quarter's earnings and results. [TC 3/11, p. 51]
Committee/Staff/Observer
With respect to the approach that is currently being used which does emphasize integral, does that cause a problem for you now? [TC 3/11, p. 51]
Participant C-11
In my mind, you're going back to the smoothing question and I think there are limits to what you should smooth. [TC 3/11, p. 51]
Committee/Staff/Observer
But in terms of interim reporting of management bonuses, taxes, things like that, the way we're doing it now, do I hear you telling me it is acceptable from an analyst's point of view? [TC 3/11, p. 51]
Participant C-11
Yes. [TC 3/11, p. 51]
Participant C-7
We'd probably adopt an integral view at the expense level but maybe focus more discretely on the revenue side. [TC 3/11, p. 51]
Participant C-6
In my experience, the interim information that we receive when we receive it is really a truer picture of what's happening with the company than the year end statement. Because obviously many privately-held companies do year end tax plan and obviously they suppress profits. Whereby during the interim periods that's not done and you will see a truer picture of what the company is actually doing. So when we do get interim statements, and we try very hard to get them, I think it gives more of a real instinctive look at what's going on. [TC 3/11, p. 52]
Participant C-13
I would think those who like rolling twelve months wouldn't like the integral method because it severely distorts your twelve month rolling. I prefer to look at slicing the borders as [participant C-5] put it. So the integral doesn't bother me particularly. But if you rely a lot on twelve month rolling, the integral is going to throw you off. [TC 3/11, p. 52]
Participant C-10
Reading the two descriptions you give here, I would go for the discrete. I'd rather have things stand alone more on their own. [TC 3/11, p. 52]
Participant C-1
I think [participant C-7]'s view is really right on, discrete in terms of revenue and expenses related specifically to that revenue, more integral with relationship to longer term expenses. [TC 3/11, p. 52]
Committee/Staff/Observer
So it's safe for me to say that those of you who use rolling twelve months are not disturbed by current interim use of integral? With the exception that there's no fourth quarter and that's something you do want? [Also included in 11(b)] [TC 3/11, p. 52]
Participant C-1
Yes, definitely. [Also included in 11(b)] [TC 3/11, p. 52]
Participant C-13
Definitely. [Also included in 11(b)] [TC 3/11, p. 53]
[Context] Responses to the postmeeting questionnaire to the March 17, 1993 Investor Discussion Group meeting.
QUESTION 15
Interim financial statements are less reliable than annual financial statements, and the numbers tend to fluctuate more, because they depend more on estimates, allocations, and judgments than annual statements. In general, the shorter the period on which statements report, the greater the volatility and the greater the questions that are raised about their reliability.
Accounting standards have attempted to reduce the effects of short periods on interim reporting by making each period an integral part of the annual period, and deferrals, accruals, and estimates at the end of each interim period are affected by judgments made at the interim date about earnings for the annual period. As a result, an expense that falls wholly within an annual period (no accrual or deferral at year-end) may be allocated to interim periods based on estimated time, sales volume, productive activity, or some other basis. Supporters of that integral view argue that it allows companies the necessary latitude to best reflect the earnings during the interim period and to reduce unnecessary and unhelpful volatility.
However, some investors and creditors, and some accountants, argue that the integral view in accounting standards is a source of serious problems with interim financial statements because it results in allocations that have ranged from smoothing to outright manipulation of quarterly earnings that undermines analysts' ability to detect real changes on a timely basis. They argue that financial analysis is best served by financial statements that report transactions as and when they occur and recommend a switch from the integral view to the discrete view, in which earnings for each interim period is determined in the same way as earnings for an annual accounting period--deferrals, accruals, and estimates at the end of an interim period are determined by the same principles and judgments that apply to annual periods. They ask that economic anomalies that are attributable to seasonality be disclosed and discussed rather than concealed by undocumented smoothing.
Are you satisfied with interim reporting based on the integral view?
YES 2 NO 3
Participant I-9: In some industries, toys for instance.
Or do you agree with its critics that:
Accruals, deferrals, and estimates at the end of an interim period should be determined by the principles that apply to annual periods, and revenues and expenses should not be allocated between interim periods if they would not properly be allocated between annual periods?
YES 3 NO 2
Transactions should be reported in the interim periods in which they occur, and any smoothing to be done should be left to analysts?
YES 3 NO 2
Comments
Participant I-16: I favor discrete reporting, but have some concerns that managements unable to smooth earnings trends by allocations will alter the timing of transactions for reporting, rather than economic reasons.
Participant I-11: I'm not comfortable with my answers here. I've dealt with companies where the integral view creates problems and I'd like a way of coping with those problems. I'm not sure, though, that the discrete view would do more than swap one set of problems for another. At the very least it would probably would lead to an unsettling transition period for analysts.
Participant I-7: Yes, if all assorted information to handle this is issued.
[PMQI 3/17, p. 27-28]