Not so good news for SEI holders: This article showed up in todays Barron's
June 14, 1999
Data Mining
Seitel boosts profits by writing down costs slowly
By Bill Alpert
The timing of an echo from a seismic signal shows surveyors where there's
oil underground. An accounting entry's timing shows where the money is. This
law of fiscal physics determines how investors get a fix on reported profits --
particularly in the energy business, where bookkeeping estimates can be as
important as supply and demand, land availability and oil strikes.
With petroleum prices having jumped to $18 a barrel from their $12 depths,
oil-service stocks have revived, and one of the quickest rebounds has been in
shares of Seitel. The Big Board-listed stock has bounced to 16 from 9,
because Seitel is a pure play on seismic data for oil exploration.
Some might say that the Houston-based firm is
also a play on seismic accounting. Seitel has
grown annual sales impressively over the past
decade, from a few million dollars to $145 million last year. High margins on
those sales pumped out a gusher of cash. And Seitel shares soared from less
than a dollar to as high as 37 in 1994, fueled by earnings reports whose
consistency beat the performance of such rivals as Veritas DGC.
But the nifty quarterly earnings reported by seismic firms like Seitel are largely
determined by accounting estimates of non-cash items. And at Seitel, those
non-cash numbers come from management sales forecasts that peer as far as
seven-to-20 years into the future. Seitel's 25 cents per share in operating
earnings reported for the latest March quarter, in fact, reflected some
extraordinary non-cash maneuvers.
What's more, Seitel takes the cash
gushing from its lucrative sales and
pipes it right back into acquisitions of
more data for its library. All the
company's book value -- and then
some -- is tied up in the largest seismic
databank in North America. Looking to
book value or earnings, therefore, Seitel
investors are betting that the firm
continues its notable success, at least
enough to justify its aggressive
accounting explorations.
The upfront nature of Seitel's data spending is no different from that of peers
such as Schlumberger, Baker Hughes and Petroleum Geo Services. Seitel
commissions seismic surveys, but also buys existing data from others. The firm
recoups those upfront costs, and any profit, when it resells the data for cash,
or for a share in resulting oil strikes.
As at film and television production companies, the rate at which these upfront
cash outlays show up as profit-denting expenses is determined by
management's sales forecast. Expenses, in theory, should be deducted in
tandem with corresponding revenues. But data from most seismic shoots can
be sold to more than one client, just as Bloomberg, say, sells its stock-price
data to many clients. So seismic managers forecast a survey's ultimate sales,
then write down costs over the forecast period.
The length of time over which these expenses are recognized is crucial to the
company's bottom line. A longer sales forecast allows a longer writedown of
upfront costs, which puts less pressure on near-term reported profits. The
inherent subjectivity in this forecasting has made some investors wary of
shares like Petroleum Geo Services (commonly called PGS) and Seitel. These
investors recall the 1980s debacle at Cannon Group, which carried costs of
its B-grade films as assets on its balance sheet long after those films stopped
earning revenues.
The analogies to show biz aren't farfetched. Seitel was founded by TV
syndication veteran Herbert Pearlman, who realized that the results of a
seismic shoot had the same recurring sales potential as a sitcom shoot. A
hard-driving sales force, under Paul A. Frame, pushed Seitel into the big time.
Since 1992, Frame has been Seitel's chief executive.
What's unusual about Seitel is that it writes down its data costs far more
slowly than do other seismic firms. While PGS and Veritas DGC write down
their 3-D seismic shoots in four years or less, Seitel takes up to seven years to
write down 90% of the cost of a shoot, and up to 20 years for the last 10%.
Seitel's more leisurely process, of course, takes a smaller bite from near-term
earnings.
Seitel's warranted in its longer writedowns, says Chief Financial Officer Debra
D. Valice, because Seitel is different from all other seismic vendors.
"We are without peer," Valice says cheerfully. Unlike the other publicly-held
seismic data vendors, Seitel owns only data. Other firms have fleets of
expensive survey ships and crews that must keep busy conducting speculative
surveys, Valice notes, which results in their accumulation of unwanted data.
That is why Veritas DGC and others have taken one-time writeoffs of
unsalable data, while Seitel hasn't.
"It's incorrect to compare us to those companies," says Valice. "They don't
have the same kind of data that we have. What we create generates much
more revenue." Another reason Seitel gets more sales out of its data, she
adds, is a sales force that is far better than anyone else's. The selling staff
meets twice a day with Seitel geophysicists, to discuss where the driller clients
want to prospect and what kind of data they want. As a consequence, Seitel
says that at the end of its writedown period, total sales on a typical data
project have equaled 200% of costs. In comparison, PGS averages 180% of
costs in its data sales.
One analyst who has highlighted Seitel's exceptional writedown rules is Mark
Roberts, research director at Off Wall Street Consulting Group, in
Cambridge, Massachusetts. Since February, when Seitel shares traded at 12,
Off Wall Street has urged clients to sell the stock.
Peering at footnotes in Seitel's 10-K reports for the past two years, Roberts
says the company seems to have made its writedowns even longer. At
year-end 1997, Seitel had a net investment of $180.9 million in data and
expected to write down 23% in 1998 and 20% in 1999. That would have
been $41.6 million for the '98 writedown and $36.2 million for '99.
But come the following 10-K, for yearend 1998, Seitel had revised its
writeoff plans downward to 13% for 1999 and 15% for 2000. That meant
that instead of the originally planned $36.2 million writedown for '99, the new
plan for a 13% writedown amounted to just $34.2 million -- even though
Seitel's net data at the end of '98 had jumped $82 million, to $262.9 million.
This revision, notes Roberts, could boost the earnings that Seitel reports
during 1999. Indeed, Seitel's accounting lifts a $3 million writedown burden
from next year, too. The new schedule calls for a total of 28% in data
writedowns in the first two years, compared to an average of 41% for the
previous three reported schedules. The new schedule also leaves 30% of the
data bank to be written off in the sixth year, compared to an average of just
23% in the previous schedules.
Auditors at Arthur Andersen have always signed off on Seitel's accounting,
says Russell J. Hoffman, Seitel's investor-relations chief, who formerly
covered oil stocks at Bear Stearns. With just pride, Hoffman points to Seitel's
growing cash flow from operations-the traditional measure that adjusts
earnings for the effects of cashless writedowns. Operating cash flow boomed
from $24 million to $97 million over the five years ended 1998.
But that standard measure traditionally includes cash sunk into inventory; a
data business like Seitel has no inventory in the conventional sense. Because
such firms resell the same information, they instead call their data bank a
capital investment, which is not deducted from cash flow from operations.
Seitel spends hugely on data-but that spending isn't reflected in the operating
cash-flow measure that Hoffman prefers. Last year, in fact, Seitel poured
$139 million into new data, adjusting for writedowns. That's more than its
$126 million in data sales and far more than its $97 million in operating cash
flow.
Hoffman (along with others at Seitel) actually refers to the databank as
"inventory." In any case, a more conservative look at Seitel's performance is
provided by "free cash flow," which includes cash spending necessary to keep
a business humming, whether for inventory or maintenance of capital assets.
Over the long haul, investors might expect a business to throw off free cash,
even after such spending. But counting Seitel's huge outlays for data, it turns
out that the company's annual free cash flow has averaged a negative $3
million in the past five calendar years. In the March "99 quarter, moreover,
free cash flow was more than $33 million in the red.
Financial chief Valise says Seitel must invest in new data. "We're a library
company," she says. "We can stop adding data at any time and generate
significant boatloads of cash." But Seitel surely cannot stop all data spending
without its inventory -- its library -- becoming stale.
The March '99 quarter shows how Seitel's accounting works to the benefit of
its reported earnings. In the quarter, Seitel's purchase of a Canadian databank
from Amoco comprised a large piece of the quarter's $63 million in data
purchases. The cash expenditure didn't dent Seitel's profits, because Seitel
calls it an investment. But a $4 million noncash swap of Seitel data to Amoco
-- which Seitel's 10-Q says was part of the Amoco data exchange -- was
counted as revenues on Seitel's income statement. This noncash sale certainly
helped Seitel achieve the quarter's $38 million in reported revenue and $12
million in operating profit. Amoco wanted some Seitel data, says
investor-relations exec Hoffman, so the accounting was correct.
That accounting also helped take the sting out of Seitel's March recognition of
an $8 million impairment in its ownership of a survey business. Seitel took
pains, however, to tell investors that this writedown was "non-cash and
non-recurring."
Seitel's accounting calculations aren't academic, because management
bonuses are calculated from reported sales and earnings. Executives like
CEO Frame also have been substantial sellers of the stock. The company
registered nearly 350,000 shares held by its brass last month, during which
Frame sold 160,000 shares. He also filed sale notices for 103,000 shares in
June.
Frame and his crew deserve rewards for expanding the business steadily
through several dips in the oil cycle. They say they are comfortable with Wall
Street forecasts of $1.20 in earnings this year. Seitel seems likely to get those
earnings the recognition they deserve. |
