ken@models.UUCP (02/02/84)
The question of whether "old AT&T" stock is worth more or less than the
sum of its parts is not at all silly, as someone recently suggested; a lot
of Wall Street professionals have been making a lot of money arbitraging
things like this. In the first few weeks that when issued trading of
the pieces of AT&T started, one trader for a big house (Morgan Stanley?)
was quoted as saying they were "buying old and selling new until our fingers
were bloody."
It is true that if the sum of the prices of the pieces were way out of line
with old AT&T, then arbitrageurs would push them back closer together. On the
other hand, it's a difficult arbitrage to execute; you have to buy and sell
nine things more or less simultaneously. If you can guess which way the
difference between the whole and the sum of the parts will drift before being
pushed back, you can make some money. In other words, the question of why they
should ever get out of line in the first place deserves consideration.
I don't understand the comment that "investors as a group hold the same number
of shares in each piece... Just remember that for every seller there is a
buyer." It sounds like a fractured version of a zero-sum game, which investing
in AT&T definitely isn't. Dividends and taxes are two obvious spoilers.
If utility functions were linear, then this whole issue wouldn't exist. However,
primarily because of taxes and different levels of risk tolerance, utility
functions are quite nonlinear. I find the argument that there is more choice
quite compelling, because different investors will choose the things they have
higher utility for. On the other hand, the disutility of holding a lot of little
pieces may be a factor. In any case, there'll probably be some kind of drift
away from equilibrium before arbitrage can catch it; if you can call it, good
luck. You don't have much time, though; the game stops February 24.
The issue of different utility functions comes up in a different way with a
thing called Americus Trust. It works like this: you take a share of old AT&T
and give it to Americus Trust; they give you back a certificate called a unit.
The unit is perforated down the middle; it separates into two parts called a
PRIME and a SCORE. The trust holds the AT&T shares (and the pieces, when they
become pieces) for you for five years. If, in five years, the sum of the prices
of the pieces is $75 or less, the PRIME is worth that sum and the SCORE is
worth nothing. If the sum is $75 or more, the PRIME is worth $75 and the SCORE
is worth the difference between the sum and $75. The PRIME gets the dividends
and the voting rights. (I've simplified a little; brokers have more in depth
literature about the exact way this works.)
You can tear a unit down the middle and sell off either piece on the New
York Stock Exchange (these things are in the newspaper under "AmTrUn,"
"AmTrPr," and "AmTrSc"). Recently the PRIMEs closed at $59 1/2, the SCOREs
at $7 1/4, and old AT&T stock at $65 3/4.
Generally the PRIMEs will be preferred by investors with low risk tolerance, and
probably a low tax bracket; the SCOREs (which are like warrants) by the higher
risk players. This system allows you to select whether you're more interested
in speculating in AT&T (SCORE) or in getting a relatively safe return (PRIME).
The interesting thing is that the sum of the prices of the PRIME and the SCORE
has consistently been well above the price of old AT&T since these vehicles
started trading in October, probably because people have different utilities
and this offers more choice in a different way than the breakup does.
Ken Winston
{philabs, cmcl2!rocky2}!cubsvax!models!ken