I’ve been thinking a lot about how somebody could commit fraud in retail. In my opinion, accounting fraud in retail is a bad idea. Fraudsters cannot both (A) avoid jail and (B) inflate earnings by a meaningful amount or otherwise significantly mislead investors. Smart fraudsters gravitate towards mining, oil & gas, pharma, etc.
While fraud is a bad idea in retail, some fraudsters are crazy enough to commit fraud even if there is significant risk of jail time. I think some people are biologically wired to steal and cheat even if it doesn’t make sense.
For investors, detecting accounting fraud in a retailer may be extremely difficult. I have not figured out reliable methods of independently verifying a retailer’s numbers.
Sam Antar describes various frauds that the Antar family engaged in while running the Crazy Eddie electronics chain at White Collar Fraud .com. The skimming activities are described in the section on Crazy Eddie’s early years. (For the record I don’t trust Sam Antar. I don’t really believe in reformed criminals like Antar’s friend Barry Minkow. I don’t believe Antar’s “it takes one to know one” schtick.)
Before Crazy Eddie IPOed, insiders were skimming cash from the business and paying some employees under the table. This allowed the electronics chain to save money on taxes and to pay less payroll tax. Some employees paid less income tax (and social security) if they did not report all of their income. On cash sales, the chain would keep the sales tax instead of remitting it to the government.
Leading up to the IPO, the Antar family skimmed less and less cash from the business. This created earnings growth on Crazy Eddie’s official set of books. I believe that this fraud would have been extremely difficult for investors to detect, especially because the underlying business was actually profitable and growing.
Crazy Eddie’s auditors did notice that something strange was happening (see the Look, More Profits! section on the White Collar Fraud website):
Both external audit firms Penn and Horowitz in 1980-83, and Main Hurdman in 1984, noticed that many employees who were previously paid what seemed to be extremely low wages (considering their positions and responsibilities) had received raises in multiples of 3 to as many as 20 times their previously-reported salaries. The auditors did not know that Crazy Eddie management was “grossing up” their employees’ total check compensation to make up for the loss of “off the books” compensation.
The gullible auditors accepted our silly explanation that our employees had sacrificed many years working at below average wages for the opportunity to be part of what they hoped might become a growing public company.
Inflating same store sales
One way the Antars inflated same store sales was by injecting their own money into the company. While there were ways that auditors could have detected this scheme, I think it would have been difficult for investors to detect it.
Another way the Antars inflated same store sales was be inappropriately attributing revenues to the stores from sales of merchandise to resellers, retailers, and wholesalers. None of this resold inventory actually moved through Crazy Eddie retail stores.
Because auditors normally only count a small portion of a retailer’s inventory, it is possible for fraudsters to game the auditors’ inventory checks. Crazy Eddie employees “helped” the auditors by climbing on top of heavy boxes and shouting out inaccurate numbers. Fake inventory inflates reported earnings.
One way for investors to detect fake inventory is to look at the relationship between inventory and cost of goods sold (e.g. inventory turns, days sales inventory). Fake inventory will cause the ratio between inventory and COGS to go up. The problem for investors is that there may be legitimate reasons as to why a retailer is failing to move inventory. Looking at suspicious increases in inventory may create a lot of false positives. That being said, it is worth looking at how quickly any retailer is moving its inventory. It is likely a sign of bad management or (less likely) egregious fraud.
Another type of fraud that Crazy Eddie employed was understating the amounts owed to vendors. One vendor helped Crazy Eddie commit fraud by shipping inventory first and invoicing the retailer after the quarter ended. Auditors did not apply sufficient cut-off testing to detect this type of fraud.
Crazy Eddie also created fake documents claiming that Crazy Eddie was owed more manufacturer rebates that it was actually entitled to. Auditors did not verify the actual amounts with the manufacturers.
One way for investors to detect this type of fraud is to look at the relationship between accounts payable and inventory. Unfortunately, this test may generate a lot of false positives.
Capitalizing software development costs
Personally, I think that all software development costs should be expensed. To help investors understand a company, accounting rules could force all companies to report their estimated software development costs. They should be forced to provide some information that would help investors form their own opinion about the value of a company’s software.
In reality, GAAP accounting rules generally give retailers the option of capitalizing software development costs. The rules create lots of fees for accountants, don’t actually help investors understand a business, and create opportunities for abuse. I suspect that the people who determine the rules intentionally create areas that can be abused so that they can land lucrative consulting gigs helping companies abuse the rules or to help them navigate complex rules. Anyways, the rules are what they are.
There is some subjectivity as to what costs constitute software development. Corporate overhead and IT costs are arguably needed to support software development. The subjectivity creates room for abuse. Unscrupulous management teams can try to find excuses to capitalize costs as “software development”. Capitalizing costs rather than expensing them inflates reported earnings equal to the amount of costs capitalized. I am not aware of any individuals who have gone to jail for being egregiously aggressive in this area.
Investors can potentially spot these shenanigans by looking at the amount of capitalized software on a company’s balance sheets.
Capitalizing other costs into PP&E
Any company could engage in Worldcom-style accounting by inappropriately capitalizing costs into property, plant, and equipment.
To look for red flags, investors can look at a company’s statement of cash flows. Investors can add back capital expenditures and subtract depreciation and amortization associated with PP&E. However, this adjusted earnings number will lag actual earnings if a company is rapidly expanding and actually investing capital in new stores and warehouses.
Capitalizing costs into other assets
A variation on capitalizing costs into PP&E is to capitalize costs into other asset types. For example, a retailer could capitalize its advertising costs (e.g. TV ads, catalogs) into “other assets”.
Estimated useful lives
A company can lower its reported depreciation and amortization expenses by making more aggressive assumptions about the useful lives of its assets. For example, it can claim that its software has a useful life longer than the typical 3-5 years.
Subprime lending involves a high degree of uncertainty. It is difficult to predict exactly how many subprime loans will be paid back and what the losses on unpaid loans will be. Uncertainty generally creates room for aggressive accounting and fraud. One obvious method of inflating earnings is to underestimate loan losses.
Another method is to hide bad loans and delinquencies by changing the definition of what a delinquency is. Usually the 10-K will have some clues as to whether or not a company is using a bizarre definition of what a delinquency is. The annual report will usually mention any re-aging, loan extensions, refinancings, etc. that turn delinquent loans into “current” performing loans.
Derivatives also involve a high degree of uncertainty and are prone to abuse. See my previous post “Reading financial statements: options”.
Fraudsters can take various steps to hide their fraud. There may be stable relationships between revenues or profits and the following items:
- Warehouse space: it would make sense for a retailer to grow its supply chain infrastructure alongside its store count increases. Retailers often report the number of warehouses, their locations, and their square footage.
- Office space.
- Employee count.
- Contingent rents. Many retailers pay a percentage of sales to their landlord(s). If revenues increase, contingent rents should probably increase too.
- Some companies release additional information such as the size of their customer database and web traffic statistics.
- Sales returns.
- Gift card sales and estimated breakage.
- Sales tax paid.
- State and federal taxes paid.
Fraudsters can take steps to make sure their numbers make sense by overpaying taxes, overpaying contingent rents, etc. etc. For example, the Crazy Eddie fraud intentionally overpaid taxes because a low tax rate would have raised suspicions from investors.
Do insiders have integrity? Have they ever tried to mislead shareholders?
If management is willing to mislead investors on little things, it is more likely that they are also trying to mislead investors on important matters (which may not necessarily be fraud). This is the same idea as the “there’s never just one cockroach in the kitchen” theory. See my post on Avoiding bullsh**ers.
Management teams that have stolen in the past are more likely to steal from investors in the future (or to commit fraud). In my opinion, staying away from shady management teams is an excellent filter for avoiding fraud and other shenanigans. It is probably the best way that investors can avoid various undesirable outcomes (accounting fraud, management employment agencies, management stealing from shareholders, insiders running away with the cash and the company, etc. etc.).
Store count (and square footage)
Hot retailers with high profitability and same store sales growth generally close very few stores. For any given retailer, some stores will be less successful than others due to a bad location or other reasons (e.g. bad store managers). Even the best retailers often have bad locations that they will close or relocate. However, the best retailers in a given niche will generally open far more stores than it closes. If the store economics are rapidly getting better and better, it won’t make economic sense to shrink the store count.
See my post on Google Trends. Unfortunately, Google Trends may not be accurate about inflection points in revenue. Google Trends also lacks precision.
Compete.com, alexa.com, Quantcast, Comscore, and other companies provide independent web traffic data. Like Google Trends, they also have problems. Alexa has problems with demographics since there are certain types of users who install the Alexa toolbar (webmasters and people who install spyware).
The big picture
I think that retail fraud (the egregious type where insiders make up the numbers) is incredibly difficult for investors to detect. Looking at the Crazy Eddie annual reports available at White Collar Fraud (scroll down to “Selected Crazy Eddie Public Filings”), I’m not sure I would have spotted the fraud without 20/20 hindsight.
Perhaps the most time-efficient way of avoiding fraud is to simply look for the honest management teams. It’s worth looking at the financial statements and accounting notes for signs of aggressive accounting.
On the short side, the retail sector is not a great place to look for frauds.
Escaping detection: why auditors do not find fraud – An excellent post from Tracy Coenen’s blog.