Money to be made by forensic accounting; And it can also be a useful tool to hunt down market inefficiency

June 30, 2013 2:00 pm

On Monday: There’s money to be made through forensic accounting

By John Authers

And it can also be a useful tool to hunt down market inefficiency

The words “forensic accounting” conjure up images of detectives on the hunt for corporate malfeasance. By such means were frauds like Enron hunted down. But it is also a way to hunt down something much more prosaic: market inefficiency. And if harnessed correctly, it could prove to be a way to make money. Aggressive accounting and earnings management may be quite legal, but when companies use them, it is a great sign that they are running into trouble, at least relative to their peers and to the expectations for them. Measure aggressive accounting well and there is money to be made. With the second quarter just ended, and a new earnings season about to begin, it is an important point to remember.

Until now, forensic accounting has been the province ofspecialist short-sellers, such as James Chanos, who famously bet against Enron. This involves minute attention to specific companies, and making big bets against them. It is a labour-intensive and risky way to make money – and keep markets and managements honest at the same time.

Now the search is on to use forensic accounting in a more diversified strategy that could compete for institutional funds. The last few years have seen interest move from traditional indexing, based on market capitalisation, to “fundamental” indexing, based on factors such as revenues, profits or dividends. The next step is to look at the quality of those fundamentals, and weight on this basis.

That at least is the plan of the Flag ETF, run by John Del Vecchio. It sells nobody short, but excludes the 100 companies in the S&P 500 that take the worst grades for aggressive accounting using a proprietary metric. The secret sauce for turning measures of earnings strength into one index is confidential. But Mr DelVecchio shares the outlines.

All of the factors in the model are recognisable from basic accounting courses. The most important is revenue recognition. Big discrepancies between revenue recognised and the flows recorded on the cash flow statement imply that a company is effectively borrowing sales from the future in order to make quarterly targets, and this is a big negative factor.

Of course, this is not necessarily fraudulent, and Generally Accepted Accounting Principles (GAAP) accounting allow wiggle room precisely because there are times when it makes sense to book revenue before all the cash is in the bank. Construction companies that rely on a few large long-term contracts, for example, will have very lumpy flows of revenue, and it makes sense to smooth them out. But statistically, when companies have aggressive revenue recognition relative to their own history and to their peers, Mr Del Vecchio contends the odds are strong that they are set to underperform. Being human, their first reaction to trouble is to obfuscate, using the freedom GAAP gives them, rather than come clean.

Moving down the income statement, the model also looks at profit margins, and flags suspicious share improvements. If, for example, a company has written off inventory in one quarter and sold it in the next, this will show up as a 100 per cent profit margin; but will also signal that it is running into trouble.

Other factors include shifts in operating items such as research and development spending and taxes – all potential red flags. Finally, there are shares outstanding. The model looks for how share buybacks have been funded; buybacks funded with internally generated cash flow are positive, those funded with debt are negative. All of these factors grade companies on a scale from A to F, and the worst 100 S&P 500 members are excluded. Avoiding the losers, of whom there are many in capitalism, can be more important than picking winners.

Does it work? It does, if the back-tested results are to be believed. According to Mr Del Vecchio’s data, which is public, the model has outperformed in all but two years since 2000. And those two were 2007 – the top of the credit bubble when banks with what turned out to be very questionable earnings fared well – and the crash year of 2008, when the model fared even worse than the market, but rebounded more the next year.

As for current exclusions, names like United Technologies and Kraft Foods presently receive an F and are excluded. The call on Kraft, up more than 20 per cent for the year and comfortably outperforming both the S&P and other consumer staples in the process, is notably contrarian: but the company gets its F grade for its revenue increases which are barely positive quarter over quarter.

And if strategies like this can prove that they work and become widespread, maybe chief financial officers would save us all a lot of trouble and abandon attempts to obfuscate they run into trouble.