Micron booked further inventory write-downs in every one of the
next six fiscal quarters. Was the devaluation of Micron’s inventory a
nonrecurring event, or had it become a chronic condition? Reason-
able minds can differ on this particular case, but one thing is clear:
The intelligent investor must always be on guard for “nonrecurring”
costs that, like the Energizer bunny, just keep on going.^5
THE PENSION DIMENSION
In 2001, SBC Communications, Inc., which owns interests in Cingular
Wireless, PacTel, and Southern New England Telephone, earned $7.2
billion in net income—a stellar performance in a bad year for the
overextended telecom industry. But that gain didn’t come only from
SBC’s business. Fully $1.4 billion of it—13% of the company’s net
income—came from SBC’s pension plan.
Because SBC had more money in the pension plan than it esti-
mated was necessary to pay its employees’ future benefits, the com-
pany got to treat the difference as current income. One simple reason
for that surplus: In 2001, SBC raised the rate of return it expected to
earn on the pension plan’s investments from 8.5% to 9.5%—lowering
the amount of money it needed to set aside today.
SBC explained its rosy new expectations by noting that “for each
of the three years ended 2001, our actual 10-year return on invest-
ments exceeded 10%.” In other words, our past returns have been
high, so let’s assume that our future returns will be too. But that not
only flunked the most rudimentary tests of logic, it flew in the face of
the fact that interest rates were falling to near-record lows, depressing
the future returns on the bond portion of a pension portfolio.
The same year, in fact, Warren Buffett’s Berkshire Hathaway low-
eredthe expected rate of return on its pension assets from 8.3% to
6.5%. Was SBC being realistic in assuming that its pension-fund man-
agers could significantly outperform the world’s greatest investor?
Probably not: In 2001, Berkshire Hathaway’s pension fund gained
9.8%, but SBC’s pension fund lost 6.9%.^6Commentary on Chapter 12 327(^5) I am grateful to Howard Schilit and Mark Hamel of the Center for Financial
Research and Analysis for providing this example.
(^6) Returns are approximated by dividing the total net value of plan assets at
the beginning of the year by “actual return on plan assets.”