NYU accounting and finance professor Baruch Lev is one of the most incisive contrarian critics of current accounting practices. In his recent book, The End of Accounting (written with Feng Gu) and his blog, Lev argues that current accounting methods have become hopelessly out of step with how value is created in the modern economy, and that an accumulation of new accounting regulations have only made things worse. Professor Lev backs up his critique with reams of market data and regression analyses to demonstrate how flawed accounting measurements have caused earnings and book value to become nearly meaningless to investors and now create very serious managerial biases and errors in how capital is allocated.
MarcumBP’s Drew Bernstein met with him to understand his views on where the principles of accounting went awry and how the structure of accounting might be reformed.
Drew Bernstein:
So I understand your book The End of Accounting has just now been translated into Chinese. How did that come about?
Baruch Lev:
That's actually the fourth language it has been translated into. It has already come out in Japanese, Spanish and Korean and a couple more translations are underway.
DB:
In your book you argue that the value and the quality of information that's provided by traditional accounting has declined precipitously over the last several decades. How did you go about determining this to be true?
BL:
I don't make the argument. I provide evidence. There are seven chapters in the book that show the deterioration and almost complete loss of value of financial reports. You see a huge decrease in the correlation between earnings and book value, on the one hand, and market value. That correlation used to be very high in the 1960s and 1970s, and now it's low.
We also tested immediate investors’ reactions to financial reports during the three days around the release of quarterly earnings reports. Again you see a decrease in impact of the reports. Over the same period there has been an increase in the impact of alternative information sources like analyst reports and non-financial SEC reports, such as 8-K filings.
We also test the utility of reported earnings to predict future earnings, which is the major use of earnings by financial analysts. This also declined significantly. No matter how you look at it, financial reports have almost completely lost their value.
I'm making a presentation in an hour to the faculty here at NYU Stern School. And we look at the latest data on this issue. This comes from a paper by Feng Gu and myself that is the most downloaded paper in finance right now. It looks at the potential returns from accurately predicting earnings. We look at the gains that you can make if you predict all the firms that will either exactly meet or beat analyst consensus. And you can see that the gains from doing that have declined precipitously. The abnormal returns are now very close to zero. So even if you are the most successful brilliant stock analyst, with the very best spreadsheets, and you can predict earnings perfectly, you're not going to get a significant reward from it.
So these are not arguments — these are facts.
DB:
You also make the remark that the structure of accounting has been "frozen in time for over a century." Can you just explain what you mean by that?
BL:
The rules are not frozen in time. There are lots of new standards by both the FASB (Financial Accounting Standards Board) in the United States and the IASB (International Accounting Standards Board). I am talking about the structure of the report. I compare a financial report of U.S. Steel from 110 years ago to today. You see that the same structure, the same balance sheet, the same income statement was released to investors 110 years ago is basically released today with essentially the same type of information.
This struck me as ludicrous. Can you think of an information system that didn't change over 110 years? Same old balance sheet and old income statement, that's the best way to inform investors?
DB:
If you watch or read financial news, you still get the impression that earnings is what "moves the markets." Yet you have the data that shows this is largely not true. Why do you think the media and the markets are locked in this mindset?
BL:
That is changing. If you look at this recent article in the Wall Street Journal, for example, they say "The Forgotten Earnings Season: Wall Street's favorite valuation measure has fallen at a speed usually only seen in a crisis.” Finally they realize that earnings don't move markets anymore.
DB:
So you think sophisticated journalists are catching up to you.
BL:
Everyone is catching up to it.
There is a quotation here from Oakmark, a large investment house, and they said "now economic value is not closely tied to book value, and the income statement no longer provides a reliable indication of the value a company created in a particular year." There was just a week ago an article in Barron’s "Earnings, Shmernings: When The Value of Numbers Decline." Meaning they don't mean much.
DB:
What are the things that changed, either in the real economy or in the proliferation of accounting rules, that caused the information content of accounting data to decline?
Intangible assets "are clearly investments in the long-term and create assets. But due to the folly of accounting regulators, they are expensed on the income statement."
BL:
Two big things. The first is the huge increase in intangible investments, particularly in the United States. U.S. companies now invest more than $2.2 trillion a year in R&D, software, brands, and information technology.
All these are clearly investments in the long-term and create assets. But due to the folly of accounting regulators, they are expensed on the income statement. So with more than $2 trillion weighing on the income statement, the earnings number is massively distorted.
The second big thing is that over the past 25 years, the regulators at FASB and IASB created rule after rule that have increased the use of management estimates and projections. For example, fair value accounting that requires marking assets to markets. Now if these assets are traded in liquid markets, like stocks and bonds, that's good. But most of the assets are unique to the companies. They are not traded. So you mark to models layered with multiple subjective estimates and assumptions. As Warren Buffet remarked, marking to market quickly degenerates into “mark-to-myth” in those cases.
So many companies fudge these numbers, use them to manipulate earnings, and they decrease significantly the quality of information.
DB:
And yet FASB, the SEC, and PCAOB have invested an enormous amount of energy in trying to improve the quality of financial reporting and audit process. So did these reform efforts have unintended consequences?
BL:
Much of this effort is misguided. If you consider Pfizer, the largest pharmaceutical company, it invests $7-9 billion every year in R&D, recorded as expenses. Expense in accounting is something which doesn't provide any future benefits, like salary or rent. If indeed the $8 billion of R&D investment doesn't provide any future benefits, you should immediately fire the CEO of Pfizer! He's just wasting investor's money — which of course is ludicrous. It's a completely misguided effort of accounting regulators.
DB:
You also point to the increasing use of nonrecurring or one-time items in decreasing the value of accounting information. Those write-offs usually are backed out of analyst's earnings models, so why are they so detrimental?
BL:
Some are backed out and some are not. When they are specifically identified in the income statement, like special items land restructuring costs, it's easy to back them out. But usually components of these one-time items are in SG&A expenses or are in cost of sales, so you cannot back those out. Also studies have shown that companies manipulate these things. They have regular expenses that they classify as special items, so investors think that these are one-time items. Lots of manipulation goes on.
DB:
Can you explain why the current approach to booking intangibles is so ineffective? What would an alternative approach to booking these expenses would look like?
BL:
I cannot explain why the accounting standards setters have disregarded reality for 30 or 40 years. This is not for me to explain. This is for experts in other fields, perhaps psychology, to explain why a group of people disregard reality. But the fact is that they do.
It's easy to capitalize these intangible investments. So, if you buy a house, this is an asset. If you buy a patent, which is often much more valuable than a house, then it is also an asset. But if you internally generate that same patent, that's an expense. This kind of asymmetry doesn't make any economic sense.
Several years ago I was invited to the FASB to talk with them. I was there for a whole day, and they said to me, "Well, if you buy a patent it's an arms-length transaction, you have a determined price." I said, "If you develop a patent, 60% to 70% of the cost of research and development is the salaries of scientists. Do you think that this is not a market price? Don't you pay the market price for top scientific talent? So what is the difference?" It's exactly the same thing.
They waved their hands and didn't have an answer. The solution to this is very simple, but they don't do it.
Whenever I say capitalize intangibles as assets, people respond, "But we don't know the values of these intangibles." I don't say that you have to value intangibles at current market values. Intangibles are not traded in markets — they're unique to companies. All I'm saying is, consider the expense as an investment, the same thing you do with a house.
DB:
Another statement from the book that stuck with me, is: "Accounting isn't about facts anymore." You talk about the huge growth in the use of estimates. But given the increasing financial complexity of many businesses, is it possible to accurately record financial results without recourse to estimates?
BL:
There has been a huge increase in the use of estimates. Looking at the S&P 500, we analyzed their financial reports and found the average number of estimates per financial report rose from 30 in 1995 to 150 in 2013. It's mind boggling! My position is that there are estimates based on good data, such as the bad debt reserve, where you have solid data on the percentage of customers that will not pay. These are solid estimates.
And then there are estimates that are just guesses, which don't make any sense. One example is pension expense, which for labor-intensive companies is a large expense. This expense is based on five or six different estimates that are co-mingled together, including managers predictions of the rate of return on the funds they set aside to pay retirees over the next five to seven years.
You cannot predict the performance of capital markets tomorrow! And they are being asked to predict five to seven years? If they could do that they would be richer than Jeff Bezos! So it is just an invitation to manipulation, to huge errors. I don't reject all estimates, but those which are just “guesstimates” with a G, have no place in accounting.
DB:
You also talk about the increasing correlation between non-financial disclosures and stock returns. The most obvious example is when a biotech company has a failed drug trial, that will wipe out shareholder value without any impact on the reported financials. But isn’t it inherently outside of the ability of accounting systems to capture that sort of information?
BL:
I agree with you. It's probably outside accounting systems. I don't consider the accounting system as all-powerful to capture everything. The reason why we talk about this in the book is that this is one of the major reasons for the detachment of stock prices from accounting information. Stock prices will react to failure or success of clinical tests, but it's not in the accounting information. But pharmaceutical companies, in the United States at least, are obliged to report failure and success of clinical tests on a website.
DB:
What are the managerial consequences of working within this flawed accounting structure?
BL:
Very serious. A year ago, I was invited to present the book at the Harvard Business School. I largely focused on the impact on investors. A professor there said, "You are talking about the deterioration of the quality of earnings. I agree with you. But those same earnings are used by managers for decisions whether to expand or contract production, spend on R&D, make mergers and acquisitions, invest in brands, pricing decisions, compensation — critical decisions with major implications for the business."
I looked at him, and I said, "Wow! You are perfectly right!” I have no doubt that the same poor quality earnings are also used internally with dire consequences. It's very difficult to measure the adverse consequences of those decisions. You don't have any data on wrong decisions by managers. But we do have overwhelming data showing that most corporate acquisitions destroy value over the long-term. So most acquisitions are really bad decisions — bad decisions likely based on flawed accounting information and flawed incentives!
DB:
And I suppose it also has an impact on the way that investors allocate capital among investing choices?
BL:
Yes, I use a recent example from Target. On May 22nd of this year, Target published their financial reports for the first quarter, which were basically good but showed a decrease in operating margin from 7.1% to 6.2%. This really alarmed investors, and there was a sell-off of the stock. But within a couple of weeks the price recovered as investors understood that the operating margin decreased because Target invested in their online platform, which was key to their longer term strategy. So investors who sold on flawed earnings data lost money for no good reason.
"These reports don't reveal much. If within a few months after seemingly good report, the company spirals down, then that tells you that these reports didn't say much."
DB:
In the book, you analyze earnings calls and find that analysts and institutional investors have various ways that they gather information about these intangible assets and the drivers of future cash flows. So are the flaws in the accounting system fatal? Or do investors already have workarounds?
BL:
In some cases they have alternative information sources. But these are not complete information or perfect alternatives to financial reports. I'll just give you one example, General Electric. General Electric’s stock has declined significantly in the past two years.
If you look at the financial report for 2016, the last report that Jeff Immelt presented, it appeared that GE was profitable. The financial analysts were generally happy, and they gave compliments to management on the earnings call. Then just a few months later, the whole thing changed. So they don't have any authoritative information. They rely on these reports. These reports don't reveal much. If within a few months after seemingly good report, the company spirals down, then that tells you that these reports didn't say much.
DB:
You marshal a lot of compelling evidence about these flaws in the current accounting system. What are your suggestions for rebuilding this century-old system in a way that would be more valuable to investors and managers?
BL:
I have two kinds of suggestions. I describe them as a small turn of the dial, and a big turn of the dial.
A small turn of the dial would deal with the intangibles and consider them as investments. Secondly, delete or abandon estimates that are not based on good data, such as market prices or verifiable historical data.
The big turn of the dial would be to create reporting on the strategic assets of the company. We have six or seven chapters in the book about it, showing examples of how this would work in various industries. An operating system, if it is a successful one, is a strategic asset. Patents, brands, content, wireless spectrum, a loyal customer base — all of are strategic assets. We give examples for four industries, very specific examples on how to report on strategic assets.
To come up with this we did a significant amount of research, listening to earnings calls and other documents to see what analysts and investors really consider valuable information. For example, when are your patents going to expire? What is your customer churn rate? Based on this, we created the template of a report that, if it were provided by companies systematically, would improve the situation significantly.
DB:
In the book, you mentioned that both management and auditors have certain incentives to stick with the old system. As you've been talking about these ideas, what's been the reaction among management, auditors, investors, or other stakeholders you'd need to make major changes?
BL:
Let me start from the positive, and end with the negative. There has been a huge positive reaction from investors. I am frequently invited to large hedge funds to talk about these ideas and help them develop systems that will include the new information and new assets. Due the collapse of traditional “value investing,” everyone is interested. There is a constant move from managed funds to index funds. So they are actively looking to improve their performance.
As for managers, I have been invited to several forums for CFOs and CEOs. But when it comes to really providing this detailed disclosure on intangible investments and strategic assets, I see much less enthusiasm. Managers agree with me that current accounting is of limited value. There was a survey in which 92% of CFO's said that financial reporting today is just a compliance exercise. It’s not an attempt to inform investors. It’s an attempt to comply with complex regulations from the FASB and IASB.
But most managers, it seems, are happy enough and they know how to work within the system. It’s costly — but it’s not costly to them. It’s costly to their shareholders.
I didn't see any significant will to change from auditors. I worked with one of the Big Four. I put a lot of time and effort to help them develop a new paradigm for reporting. And then it died. One of the partners told me that there was complaint that this effort doesn't provide any billable hours. They should not be improving accounting. We should get paid to test the financial reports.
DB:
So, lets imagine you were placed in charge of modernizing accounting and financial reporting systems. What would be your first year agenda? How would you build consensus that reforms would result in better outcomes for investors?
BL:
I am no expert in building consensus. If I worked to build consensus, I wouldn't have done any of this research!
I would focus on fixing the broken system. Accounting properly for investments in intangibles. Stripping out the “guesstimates.” Developing systems to report on and document strategic assets of the firm. I am not calling for more regulation. If FASB took on such a project it would take them 25 years and the result would be another thousand pages of the regulation.
I would like to see a bottom-up approach where industry associations come up with a reporting template for the industry. As a regulator, I would encourage these things. I would bring together managers, auditors, investors, and academics to develop these new reports. I would focus on measures that matter — rather than measures that are largely irrelevant. That's what I would do.
DB:
That's fantastic. Thank you so much for your time.
BL:
Thank you as well. And now I must go. I have a very urgent presentation to give!