Opening Up the Black Box of Accounting

Describe the ‘accrual anomaly’ that you have studied, and its effects.

The accrual anomaly relates to the negative association between accounting accruals (the non-cash component of a company’s earnings) and future stock returns. This anomaly has been recognized by researchers and practitioners alike for over 20 years, and was first highlighted in 1996 by Richard Sloan, who is now a professor at Berkeley. Prof. Sloan found that firms with high accruals (i.e. high earnings but lower cash flows) have weaker performance in future years, while firms with low accruals (low earnings but higher cash flows) have stronger performance in future years. This is predictable, and you can earn returns based on it.

This was one of the most persistent stock market inefficiencies ever identified. It is referred to as an ‘anomaly’ because it is something so basic that it shouldn’t predict future returns, and yet it seems to — or at least it did, for quite a long time.

Why does this happen?

There are two explanations. The first is a mispricing-based explanation, which states that investors don’t understand that earnings consist of a cash flow component and an accruals component. The cash flow component of earnings tends to persist, while the accruals component tends to reverse, because that’s the way the accounting process works: you can call something a ‘receivable’ in a given period and then collect it in the following period, or if you have a lot depreciation in one period, you will have less in another period, and so on. Because investors don’t understand this ‘differential persistence,’ they consistently get misled. The second explanation is that risk is not being measured properly, and the higher returns are the result of taking additional risk.

In your most recent research, you found that the accrual anomaly might have disappeared; please explain.

The accruals based strategy consistently produced returns, but in the past decade, it anomaly has apparently disappeared, and in my research, I link this with another interesting phenomenon: the emergence of analysts issuing cash flow forecasts.

For a long time, analysts — like investors — were getting misled by accruals. My research looked at the emergence of analysts issuing cash flow forecasts in addition to their earnings forecasts, suggesting that over time, they began to pay attention to accruals. My question was this: as analysts pay more attention to accruals and thereby provide better forecasts to the markets, will there be a decline in the accrual anomaly? And that is exactly what I found.

Basically, as analysts started forecasting earnings and cash flows, they were less likely to be misled by ‘differential persistence’. As analysts are such important conduits of information to the capital markets, the markets were also less likely to be misled. Simply put, analysts started doing a better job of forecasting as they were paying attention to accruals, and as a result, the anomaly has all but disappeared.

In some sense, my findings support the mispricing explanation mentioned earlier, by showing that in the past, information was poor and investors made decisions very mechanically, and that’s why they were misled. These days,

as information has gotten better (in part thanks to analysts’ efforts) investors are less likely to be misled.

This is not the only explanation for the decline of the accrual anomaly in the last ten years. Another explanation is that institutional investors are actively investing in accruals based strategies, which has also ‘arbitraged away’ some of the returns.

You have also devised a way to grade growth stocks. Please describe it.

In much of my research, I focus on how people make decisions by using information mechanically, without thinking about what is behind the information. The fact is, a large part of any growth valuation is hype. A company might have had fantastic recent performance — so investors naively belief that it’s going to persist forever. The dot.com boom is an extreme example of what I refer to as ‘naïve extrapolation’.

In 2005, I came up with an index called the GSCORE, to separate out the overly-hyped firms from the solid-growth ones. The GSCORE looks at three kinds of signals. The first is earnings and cash flows, focusing on firms that earn more and generate more cash flows than peer firms. The next set of signals is the variance of profitability and sales growth. I argue that if the variance is lower, you are more likely to have consistent, stable performance going forward. Your current strong performance was probably not just the effect of luck — it was probably something that you were genuinely capable of.

The third set of signals is related to a firm’s accounting practices, and how their approach might be hurting current profitability, but might actually help future profitability. Simply put, if firms are undertaking certain kinds of activities, they are more likely to have future profitability and live up to their valuation. The three cues the GSCORE looks at are R&D investments, advertising and capital expenditures. I have found that, even if the firm’s current performance isn’t great, if it is investing heavily in these areas, it is creating value for the future.

What firms would you recommend, based on your approach?

First, an update: my GSCORE paper came out in 2005, and more recently, I’ve been working with my Rotman colleague Kevin Li to refine it further, by combining it with other methods for fundamental analysis. Basically, we try to combine financial statement-based strategies (like the GSCORE strategy) with formal valuation methods that measure ‘intrinsic value’. Basically, our new approach tries to focus on high GSCORE firms that are trading below their intrinsic value to invest in — i.e. clearly undervalued firms. Conversely, our strategy recommends shorting low GSCORE firms that are trading above their intrinsic value.

I don’t have access to 2014 financials yet, but I can tell you what forecast I would have made in early 2014, based on 2013 data. One firm that scored very highly on both my valuation screen and my ratio analysis screen is Jetblue Airlines, and since June of 2014, it has earned a 50 per cent return — significantly higher than the S&P’S four per cent over the same time frame. Other firms that scored highly are Brocade Systems, which went on to earn 23 per cent, and Hewlett-packard, which earned 11.7 per cent.

It is important to note that this strategy may not work for individual stocks, because there is too much firm-specific risk. It will work much better in a portfolio setting. My approach actually suggested a couple of firms in the oil industry, and clearly, that sector has gone through a very bad period. So, one company, Helmerich & Payne, went on to post -43 per cent returns. On the whole, the strategy works very well, but the way to implement it is to look at maybe 30 stocks and focus on building a portfolio of those stocks. You have to do your own math and then apply these approaches to ‘screen in’ and ‘screen out’ firms. But I don’t recommend using the GSCORE in isolation — and certainly, don’t use it for just one or two stocks.

What is your advice for investors who want to add a margin of safety when searching for growth stocks?

Say a firm has a very high GSCORE. As indicated by my current research, that isn’t enough to recommend it anymore. It is best to use multiple lenses — to combine formal valuation models with indexes, so that you have two, three or even four screens to compare when you’re looking at a stock, and you can focus on firms that score positively on every screen.

For example, you might use the GSCORE and a simple ratio like the PEG ratio, which is a stock’s price-to-earnings ratio divided by the growth rate of its earnings for a specified time period. Think about it this way: if three different people tell you that somebody is very good at what they do, that has more credibility than if one person says so, and the same applies to stocks.

In an attempt to beat the market, some investors buy high-scoring stocks and sell low-scoring stocks short. Do you recommend ‘short-selling’?

Not for individual investors; it’s more for institutional investors and people with deep pockets, because when you go long in a short stock, the maximum you can lose is your investment. So, let’s say you spend $100 to buy some shares of a company. If the stock price goes down to 0, your $100 is the maximum you can lose. But if you short a company that is trading at $100, and the stock price rises to $1,000, your loss is $900.00. So, your losses are technically infinite, and hence shorting is very risky.

If you’re going to do this as part of a portfolio — say, ‘buy these 30 stocks and short these other 30 stocks’ — that might be one way to do it. A second approach is, you can replicate a short-like strategy through options. So, essentially, buy put options, so you have the opportunity to benefit from a decline in stock price. But your losses are not going to be infinite, which is the case with short-selling. For individual investors, I recommend focusing on the long side.

In other research, you investigated whether managers manipulate earnings to inflate their own compensation. What did you find?

I am one of many researchers who have looked into this phenomenon, and it happens all the time. One of my papers on this topic looked at option exercises. The way management compensation is structured, in many cases, managers receive options every year. So, at any given point of time, an individual manager will have a lot of options — some from last year, some from two years ago, etc. Typically, these options have a vesting period of, say, three or four years, and they are in play for another five or six years. This gives managers opportunities to decide, ‘You know what, I’m going to exercise a bunch of options next year; and then I’m not going to exercise any for the next few years’.

Of course, the way accounting works is, you can move stuff around across periods — for instance, by deferring revenues or pre-loading expenses. So these leaders can essentially ‘make’ the year that they plan to exercise their options a really good year. We looked at firms where the CEO appeared to have abnormally large option exercises, and we found a pattern whereby that year was a fabulous year for their firm. You might argue that it was just market timing, or that the person knew that there were strong microeconomic or industry-specific factors that year. But we investigated further, and that wasn’t the case: there was very strong evidence of “earnings management” to make the previous years look bad, or the year leading up to the option exercise look really good. I’m not suggesting that doing this is illegal, but it certainly is manipulative. The fact is, there is a lot of flexibility in accounting. But some of that flexibility is being used in a manipulative fashion.

Your work underscores the importance of checking the balance sheet before investing in a company. What should investors be looking for?

I wrote an op-ed in Forbes during the financial crisis, where I argued that one of the reasons for the crisis was that people weren’t looking at balance sheets. The fact is, it’s the only way to understand the risk and the quality of the assets a firm has.

One key thing to look for is whether a firm is manipulating its earnings by ‘capitalizing expenses’. Sometimes, expenses are not shown as such — they’re shown as assets. For example, if you have a lot of R&D underway, you can sometimes refer to that as an asset rather than an expense. Or, if you have a lot of branding expenses, you can refer to them as a ‘brand asset’. If you see a lot of these capitalized expenses, you will realize that this is the reason why the firm is showing higher earnings. Understanding what the assets actually are, and whether they deserve to be there, is a critical part of financial statement analysis; but not many investors do it.

It’s very important for investors — and executives and managers — to have basic financial literacy, and to understand the basics of Accounting. People shouldn’t view Accounting as a black box, and think of accountants as a bunch of bean counters. In reality, the numbers they present have real implications. There are stories behind the numbers.

There is a lot of flexibility in accounting; but some of it is being used in a manipulative fashion.