July 24, 2000

Manager's Journal

CreativeAccounting.com

By Andy Kessler, a partner in Velocity Capital Management LLC, based in Palo Alto, Calif.

A company that loses money isn't necessarily a bad investment. I'm more than happy to pay up today if a company's business model promises long-term success. The long-term model drives lofty valuations, and each quarter's earnings release gives subtle clues as to whether we are on target or not.

But some companies, eager to keep their stock prices up, fatten their quarterly reports, so that what looks like a tasty sandwich is actually stuffed with baloney. This used to mean extended receivables, mispriced inventory, shipping bricks. But the new economy requires new tricks. Here are five types of deception that -- while perfectly legal -- fall between chicanery and tomfoolery:

• Empty COGs. You probably think the term marketing expenses refers to TV ads and billboards. Think again. Many Internet firms classify as "marketing" what ought to be considered cost of goods, making it appear as if they earn more on each sale than they actually do. CDNow is famous for sending $10 off your next purchase a week before the end of the quarter. Whatever you buy counts as a sale, the $10 off as marketing expenses.

Similarly, when you buy a book at Amazon.com, there are fulfillment costs with each order. To quote a recent Amazon 10-K filing: "Fulfillment costs included in marketing and sales expenses represent those costs incurred in operating and staffing distribution and customer service centers, including costs attributable to receiving, inspecting and warehousing inventories; picking, packaging and preparing customers' orders for shipment; and responding to inquiries from customers."

Sounds like cost of goods to me. These are substantial numbers; in the first quarter of 2000, $99.5 million of Amazon's $140 million marketing expenses were for fulfillment. Gross profit was $128.1 million. If these fulfillment expenses had been considered cost of goods, gross profit would have been $28.6 million. Gross margin for the quarter would have been just 5%, vs. the reported 22.3%.

• Log rolling. Since many companies raised tons of money in their initial public offerings, the temptation to turn the cash on the balance sheet into sales is overwhelming. Sure, you can sell below cost, but that takes a long time. Why not give someone else the money and have him turn around and buy something from you?

In January Amazon announced a deal with drugstore.com, exchanging $30 million in borrowed money from Amazon's balance sheet for drugstore.com stock at just over $28 a share. (Drugstore.com closed Friday at $6.50.) Tied to that deal was a sponsorship arrangement whereby drugstore.com would pay Amazon $105 million over three years for a health-and-beauty tab on Amazon's home page.

Similarly, Microsoft invested $250 million in Healtheon/WebMD, and guaranteed the company another $150 million in revenue -- in part to subsidize thousands of physicians' $29.95 payments for Internet access. And who provides that access? Why, Microsoft's MSN. Voilà, cash on the balance sheet into sales. Doesn't sound healthy to me.

• Tweaking knobs. Sales are sales, right? Well, sort of. Every CFO loves to play a little with the reported numbers, adjusting knobs to smooth things out and move things around. One I call the Fram oil filter knob: Take revenue now or take revenue later? It is a sure sign of imminent trouble. Close a customer for a three-year deal to use your software or service, then report all that revenue now. This is the practice that got data-mining company MicroStrategy in trouble this past March. The Securities and Exchange Commission had issued a memo in December clarifying its revenue-reporting rules for software companies, and the company had to restate earnings for 1997, 1998 and 1999. The company lost 90% of its market value.

Other companies that have lumpy businesses from big one-time increases in sales -- such as the launch of a new product -- save a lot of those sales for future quarters by deliberately putting them into deferred revenue, where perhaps they don't belong. This helps cover errors and missed quarters later. Microsoft is still unwinding deferred sales of Windows 98 to cover Windows 2000's delay and ho-hum reception. Investors are now discounting these deferred sales.

• Inflated revenues. Priceline.com matches wannabe fliers with "distressed" (empty) airline seats. To investors, it reports the full value of the airline ticket as revenue, almost as if Priceline were flying the plane. The company says it owns the ticket -- which it does, but only for a few nanoseconds -- and claims it is actually liable, as opposed to merely acting as a commission-collecting agent. So will William Shatner beam me to New York the next time my flight gets canceled? In effect, Priceline overstates its revenue on tickets by perhaps six or seven times.

DoubleClick is more or less an old-line advertising agency in a new medium. But unlike traditional advertising agencies, it reports as revenue its total advertising billings instead of the commission it keeps, which averages about 25%. Its explanation? The company says it operates a network that accepts ads, rather than places ads on a network. These semantics increase its ad-sales figures nearly fourfold.

In the early 1990s I met the executives of Itochu, a Japanese trading company. They boasted of being one of the world's biggest companies, with more than $100 billion in sales. Turned out they do that in trade, get a small fee on the goods they trade (4.6% in 1999), and thus are really more like a $5 billion sales company. Should we consider the New York Stock Exchange a $100 trillion company because of the value of the shares it trades?

• Play money. Portal companies like StarMedia and iVillage, in an effort to meet Wall Street's growth demands, do 10% to 25% of revenues in barter. A company will exchange, say, $1 million in banner ads on its site with some other dot-com for $1 million in ads on the other site. No money changes hands, just banner ads that have zero cost. Both companies get $1 million in extra revenue and record $1 million in marketing expenses.

And how do they decide how much to barter? Why not trade $10 million worth of ads and really beat Wall Street estimates? Better yet, do a barter deal with Itochu and report $10 billion in revenues.

I'm just getting started. Add paying employees with options, funny user counts, merger accounting, capitalizing the costs of acquiring subscribers. My accounting has a first name, it's O-S-C-A-R . . .

For information about subscribing, go to http://wsj.com