Biannual Financial Reports Would Be More Accurate

August 17, 2018

Fuqua professors study misrepresentation of earnings

Campbell Harvey found biannual financial reports would be more accurate

The norm for public financial reporting in the U.S. is quarterly, but the president has directed the Securities & Exchange Commission to investigate the implications of changing to semi-annual reporting.

Professors Campbell Harvey and John Graham, both of Duke University’s Fuqua School of Business, and Columbia University’s Shiva Rajgopal, have a series of research papers showing how quarterly reporting leads firms to intentionally misrepresent earnings and even destroy long-term value in order to meet quarterly targets.

Harvey discusses his findings in this Fuqua Q&A.

Why are quarterly earnings reports a problem?

It is well known that firms are severely punished if they do not hit their quarterly earnings targets. Managers go to extraordinary lengths to meet those goals, even willingly taking actions that destroy firm value. Our research showed that 78 percent of CFOs would willingly destroy long-term value to smooth their quarterly earnings.

This seems contradictory, but it is the essence of short-termism. You can generate a short-term profit at great cost in the long term. For example, our research showed that 55 percent of CFOs would knowingly sacrifice value by delaying a valuable new project if it would help them hit the current quarterly target. And 80 percent admitted to cutting discretionary spending, such as research & development and maintenance, even though they fully realize this destroys value in the long run.

There are many levers managers can use, such as delaying hiring, slashing advertising – all things they would not necessarily do unless they feel they will be short of their earnings target. It is a classic example of short-term gain and a lot of long-term pain.

It gets worse. We have a more recent research paper on earnings quality with three main findings. First, 20 percent of firms intentionally misrepresent earnings. Second, the size of the misrepresentation is very large, on average 10 percent. Third, the misrepresentation is not always overstatement of earnings. One third of the misrepresentation is the low-balling of earnings. If you are going to exceed the target, you misrepresent downwards and fill up the cookie jar.

What reporting frequency preference have you seen among fund managers?

Large pension plans strongly prefer less frequent reporting. A global conference of senior pension fund officers I attended in Paris in 2014 had at least $4 trillion of assets in the room. During my presentation, I did a poll of these officers to get their view on the frequency of reporting. The results were striking. Eighty six percent preferred semiannual or annual reporting. They understood that managers are engaged in shenanigans to hit the target, and these actions destroy value. Even if the value destruction is small – even 0.25 percent per year – that makes a huge difference for a pensioner with a 40-year horizon.

Wouldn’t there be the same incentives to manipulate six month earnings?

Yes, but the amount of manipulation will be less. When you report over a longer horizon, you naturally smooth out some of the noise. For example, suppose you are used to getting an order in the last week of a quarter. For some technical reason, the order shows up a few days late – after the quarter. All of a sudden you have a shortfall and you need to reach into your bag of tricks to misrepresent the earnings in order to hit your target. This would happen less often with semi-annual reporting. The noise is even further reduced for annual reporting.

What is the downside of less frequent reporting?

In theory, it means less information for investors. But the information that we are getting at a quarterly frequency now is not very high quality, because we have shown that earnings are being misrepresented.

In the age of electronic data delivery, why don’t we do the opposite – report more frequently?

It might seem paradoxical but more frequent reporting also solves some of these problems. The quarterly and annual reporting is a legacy choice that was determined a century ago when accounting was done with pencil and paper. Today, with modern blockchain technology, we have the ability to provide real time financial statements to investors. Given that orders are lumpy, the real time data is very noisy. There are no daily earnings “targets”. Hence, there is no incentive for managers to manipulate the real time earnings. However, in my opinion, we will move to less frequent reporting well before taking a leap in the opposite direction to real time reporting.

Contact Info

For more information contact our media relations team at media-relations@fuqua.duke.edu.