United States. Congress. Senate. Committee on the.

The Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) online

. (page 49 of 140)
Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 49 of 140)
Font size
QR-code for this ebook

ware development.

The future contribution of a healthy computer peripheral equip-
ment industry could liave been even more dramatic. The ability to
enter the industry with a single product line raised the possibility
of expanding that line to include additional peripheral products, and,
eventually, a central processing unit and a total general purpose
computer system.

But that potential will never be realized because the capital neces-
sary is simply not available today even if a company demonstrates
that it has a superior product, capable management, and current

The capital barrier to entry combined with IBM's practice of mar-
keting a total bundled computer system family gives IBM a high
degree of control over the general purpose computer market, if not
absolute control.

The ability of IBM to shift profitability from peripheral subsystems
to central processing units and back again, depending upon the com-
petitive realities of the market at the time, makes it almost impossible
to compete on any level but on a total systems level, wliere the capital
barrier to entry has proven to be almost prohibitive.

Profital)ility can also be shifted among the various peripheral sub-
systems, between the central processing unit and the operating system,
and from one end of the computer systems family to another, if effec-
tive competition enters the market in any of those areas.

IBM has the added advantage over potential competition of con-
trolling interface standards and new medial standards in the industry ;
of being able to deny field service support ; of controlling the develop-
ment of the operating system ; and of possessing the only truly accu-
rate data in the industry about market size and product profitability.

IBM's ability to orchestrate relative profitability and its many other
advantages over competition would not be so important if the industry
giant did not fully utilize all of those facilities in its competitive battle
with the plug compatible peripheral subsystems manufacturers.

Profitability was shifted; interface information remains a closely
guarded secret at IBM ; media standards are changed, but the new de
facto standard is not immediately released to the industry ; field serv-
ice support is occasionally denied to IBM systems with non-IBM
equipment attached; clianges in IBM's opeiating systems have se-
verely disrupted a competitor's business; and the lack of accurate
industry data has hampered competition.

The effect of IBM's competitive response to the initial success of
subsystems manufacturers has been to warn "Wall Street that competi-


tion on any level except on a total integrated systems level had no
chance of success.

And, the warning was not a subtle hint. Major financial institutions
have suffered, or stand to suffer, large financial losses because of the
experience, including Bank of America because of its heavy commit-
ment to Memorex; Continental Illinois National Bank, Telex's lead
bank; and Metropolitan Life Insurance Co., because of its equity in-
vestment in Telex.

In the context of the disappointing performance of IBM's general
purpose systems competitors and the harsh lesson of the peripheral
subsystems manufacturers, it is not surprising that the flow of new
capital into the computer industry has all but stopped.

My contacts with major financial institutions today indicate that
they have no interest in investing in any peripheral subsystems manu-
facturers that are competing with IBM in the end-user market, and
they have limited interest in investing in companies that are competing
with IBM on a general purpose systems level.

Since capital is so critical to success in the mdustry it can only
mean that competition in the industry will decline even further in
the future.

Should the computer industry be restructured ? The answer to that
question seems simple. A competitive computer industry is vital to our
national defense and to our continued economic and technological

And today, the industry is not only highly concentrated, but the
present structure of the industry suggest that there will be even less
competition in tlie future.

IBM would have us believe that there is no general purpose com-
puter systems market, that there is only a total data processing market
and that IBM's share is relatively small.

IBM's definition of the market is similar to suggesting that there
is no domestic automobile market ; that there is only a transportation
market and that trains, buses, planes, taxicabs, and roller skates
should all be included in the market.

IBM's broad definition of the data processing market is similar to
my transportation industry analogy, but even that comparison does
not fully describe the nature of the industry.

Data processing is much more basic and, because it is so basic, control
of the largest market within the broadly defined data processing in-
dustry also grants effective control over technology in many areas that
are critical to the future of this country.

But isn't teclinological progress best served by having one giant
company control new product development, especially when that com-
pany spends in excess of $700 million annually for research and
development ?

Unfortunately, the present structure of the computer industry re-
tards technological progress. IBM's marketing strategy is to iiitro-
duce new products when competitive pressure within the industry de-
mands advanced products.

In the absence of competition the rental-base nature of the industry
makes it more attractive for IBM to keep installed products out in
the field for as long as economically possible.

Consequently, IB]M seldom takes the lead in new technological de-
velopments in the industry, except where there is no existing base
of product.


Although the answer to the question of whether the industry should
l3e restructured seems simple, the question of how the industry should
be restructured is extremely complex, in part because it has been
structured to resemble a monolithic industry.

My remarks have focused primarily on the rental-base nature of
the computer market and on IBM's equipment financing business be-
cause they are key elements in IBM's dominance of the general purpose
computer market; and they must be addressed in any restructurmg

Any future restructuring plan of the industry should also recog-
nize that IBM's data processing operations are actually the sum of a
number of businesses, including the development, production, and
marketing of central processing units, operating systems, applica-
tions programs, and numerous peripheral subsystems; and it is also
in the business of providing a number of services such as field mainte-
nance, education and systems engineering, as well as lease financing.

Any one of those lAisinesses could potentially be established as
separate and independent competitive entities, significantly increasing
overall competition within the industry.

Thank you.

[]Mr. Collins' prepared statement appears as exhibit 1 at the end of
his oral testimony.]

Senator Hart. Thank you very much, sir, for an excellent state-
ment. I want to ask just one question to underscore a point. The staff
of the committee, I'm sure, will have many questions.

As I understand from the thrust of your statement, you are sug-
gesting that mitil a company reaches a certain size, the more success
it has. the greater will be its losses just because of the leasing nature
of the business.

]SIr. Collins. Yes, Senator ; absolutely so.

Senator Hart. What is the size necessary, in your judgment, for this
trend of the more you ship the gi'eater your lossage to change ?

Mr. CoLLixs. In terms of the break-even level, if we're addressing
the market on a total computer systems level, that break-even point has
Leen estimated at about 10 percent of the total market.

It's probably slightly less, but we don't have these numbers from the
industry to really come up with exactly what that number is.

Now, that, Senator, is break-even level on a total systems level. The
point where losses reach a peak is probably somewhere at the mid-
point of that cycle; probably about 5 years, although, again, it's a
rough estimate.

If it takes 10 years to break even, then it's probably in the 5th year
that losses peak out.

And as a percent, I would guess it would be about 5 percent of the

Now, Senator, could I qualify that a bit? That's on a total com-
puter systems level. Another point I was trying to make is that if a
company can enter on a peripheral subsystems level, which is a much
smaller 'market, that break-even level is reached much more quickly — •
even though it is a rental-base market — perhaps in 3 or 4 years.

So, the ease of entry, if you can address the market on a peripheral
subsystems level, is considerably greater than it is on a total bundled
svstems level.


Senator Hart. Mr. Nasli ?
■ Mr. Nash. AAHi}^ is it, Mv. Collins, that the strncture of the industry
seems to be essentially one of a total systems basis ?

Mr. Collins. Mr. Nash, there's a number of reasons for that. Num-
ber one, the industry itself is an outgrowth of the tabulating machine
market of the 194()'s and 1950's and tabulating machines were also
marketed on a rental basis and on a total systems basis.

Number two, I should think it would seem obvious. It creates a major
barrier to entry into the industry which creates a very comfortable
environment for IBJM.

Perhaps a third reason might be mentioned. In the initial growth of
the industry the central processing unit, itself, was a very large portion
of the total system.

That is less true today and will be even less true in the future.

Mr. Nash. If I understand your response to Senator Hart a moment
ago — and correct me if I am wrong — you made the point that in your
judgment if the industry were not structured on a total systems basis,
then the barriers to entry would be less because the capital requirements
would be less '^

Mr. CoLLixs. Absolutely.

Mr. Nash. In terms of inherent economies of scale, in your judg-
ment, must the industry continue to be structured the way it is or
would it be feasible for it to be structured on a nonsystems basis?

Mr. Collins. I believe it would be Aery feasible to structure it on a
nonsystems basis. In terms of economies of scale, in each of the individ-
ual areas — in central processing units, operating s^'stems, in each of the
various peripheral subsystems, in applications software — there are
different economies of scale.

The largest economy of scale is probably in the operating system,,
because once that's developed then the cost of producing the next
copy is simply the cost of punching new cards.

So there are tremendous economies of scale in operating systems,^
Then again, on a peripheral subsystems level, there are not very
large economies of scale to mamifacturing and for electronic products,
such as central processing units and controllers, there are not large-
economies of scale to manufacturing.

Consequently, there are various economies of scale to the industry,
if it's segmented.

Ml'. Nash. You have nnderscored the point that because of the
structure, the largest single barrier to entry is the massive capital

You attribute the capital requirement to primarily the leasing nature
of the business. Is that I'ight ?

Mr. Collins. Yes, Mr. Nash ; plus IBM's practices make that capital
requirement far more significant than, in fact, it should be.

I'd also like to go back a little bit and suggest that there is another
principal barrier — ^that systems lock — which encompass a lot of those
things that previous testimony has addressed — the A^ery high level of
support that IB]\I giA'es to its customers plus the fact that IB^SI cus-
tomers are in a sense locked into IBM software, such that they cannot
easily change manufacturers.

I've taken all of those things and put thein together and called that
a "systems lock" so that I could focus on that other barrier to entry
■which I see operating every day — ^and that's the capital barrier.


Mr. Xash. Could you enlighten us as to why the nature of the com-
puter industry business seems to be one of leasine; rather than outright

Mr, Collins. It seems like that comes back to a previous question.
There are many reasons for IBM shipping equipment on a rental
basis or on a lease basis ; many very positive reasons.

First it allows equipment to be shipped to customers that could not
normally afford the equipment. If it becomes a question of buying a
million dollar computer or renting it at so much per month, the million
dollar buy decision is a very difficult ca):)ital decision for most firms,
while rental on a month-to-month basis is an operating decision.

There are other reasons why customers rent computer equipment.
Another reason is that they're afraid to get locked into current tech-
nology; that technology has been moving along fast enough in the
industry and customers are reluctant to commit to a piece of equip-
ment by purchasing because of the risk of having that equipment
obsoleted in some meaningful time frame.

And that, of course, is the type of reasoning on the part of the users
that IBM very much encourages.

I think that major reason for a rental industry is obvious from the
structure of the industry.

It provides a major barrier to entry into the industry. And I can
go back and recite chapter and veree. Control Data, initially, success-
fully entered the industry by cojicentrating on very large scientific
computers — a market where equipment is sold outright.

Scientific Data Systems entered the market by concentrating on
medium-scale scientific systems, a market where equipment is sold

I've left out the whole minicomputer industry from my analysis.
I've focused on general purpose systems manufacturers.

The minicomputer industry has grown very significantly, and it's
very healthy today because it's an outright sale industry. Every time
we "run into the problem of addressing a rental base market, such as
the general purpose systems market, competition has had extreme
difficulty entering the industry.

]Mr. Nash. Oii the other hand, you would agree, would you not, that
it would be detrimental to the user to require computers to be sold
rather than providing them on rental or lease terms ?

Mr. Collins. No; not necessarily. And the question does pose a
problem, because we run into the old economist ploy of other things
remaining equal.

It's hard to hold all other things equal. There would be tremendous
advantages to requiring the manufacturer to sell computer equipment
outright'^ to the customer and to separating that manufacturing
marketing function from the financial function.

It Avouid significantly accelerate technological progress in the indus-
try. It would lower a major barrier to entry. It also would have some
disadvantages, as you suggest. It would make it more difficult to put
computer equipment in the hands of those customers who are reluctant
to make the purchase decision.

But within that environment a whole new leasing industry would
probably evolve that would assume the risk of financing short-term
leases within the industry.


Mr. Chujmbris. Would you yield for a moment, Mr. Nash ? You said
the prospective person would be reluctant. Wouldn't he have an in-
ability as well as reluctance?

I think you mentioned that in respect to if he had to pay cash rather
than go on a rental basis.

Mr. Collins. It's a question of capital commitment to a corporation.
A computer is a piece of capital equipment and can be judged just as
any other piece of capital equipment is judged in the total financing
plans of a corporation — whether it be a new piece of production ma-
chinery — and can be considered in the context of all of those available
means of financing the acquisition of a new piece of capital equip-
ment — new equity offering, new debt offering, et cetera.

Mr. Xash. Mr. Collins, to alleviate the financing burden on a user
would it be feasible to create a large new industry; say a computer
leasing industry or a GMAC-type entity for computer leasing ?

Mr. Collins. Mr. Nash, as a generality, yes ; it is feasible and it is
i-easonable. In fact, through the whole third generation of computer
systems we did have a reasonable leasing industry develop which was,
in part, developed on some very bad equipment life assumptions; but,
nevertheless, an industry did begin to develop.

"\Miether or not we could set up a GMAC-type of operation sounds
feasible ; but again, it could not be done entirely in the current context
of the industry.

IBM has too large a share of the market. IBM has too much control
over technology and OA'er prices in the industiy that if we simply set
up a G^MAC, that GMAC would want to purchase almost all IBM
central processing units ; although, it would purchase many independ-
ent peripheral equipment subsystems.

;Mr. Xash. Does that mean creating an effective competitive atmos-
phere in the mdustry that requires a sort of multifaceted approach?

]Mr, Collins. Absolutely.

]Mr. Xash. You noted in your statement that IBM has a cash posi-
tion of rouglily $3.3 billion.

I was wondering whether it was miique or commonplace for large
corporations to have such a cash position.

!Mr. Collins. It is extremely miusual for corporations to have a cash
position that large. And it's unusual in the context of IBM for it's
cash to be quite that much of total current assets and total corporation

There are probably a number of reasons.

Xo. 1, IBM was restricted to some degree as far as the amount
that the}' could increase dividends over the last year or 2 — which built
up that cash position somewhat.

Also, the sales to lease mix on shipments of some models of System
370 seem to be higher than what IBM expected.

So they got that cash initially in terms of profits and cash flow
rather than deferring it until later years. And I can't help suspect that
IBM doesn't want a large cash position now because of some of the
things that are happening on an antitrust front.

Mr. Xash. What are the competitive ramifications of such a large,
$3.3 billion, cash position ; if there are any ?

Mr. Collins. Well, in terms of capital and cash flow and the ability
to raise external financing, that $3.3 billion is extremely important


along with the cash flow of $3.3 billion in 1973, because it suggests
that IBM could go into a new equipment cycle ; for example, a future
system which could obsolete a great deal of their current lease base of
370's and have the cash to support that capital commitment to the new
rental base.

Whereas, competitors do not have that flexibility. Consequently, it
would prematurely obsolete current systems competitors' equipment at
a time when they are not in a position to commit the capital to build-
ing and expanding a new lease mix.

Mr. Xash. What does IB:M do with the $3.3 billion in cash? Do they
deposit it around the country in different banks, invest in securities, or

Mr. CoLLixs. I believe it is primarily invested in short-term Treasury
bills. There is a portion of that cash in longer term notes that are kept
below the current asset line and called securities for repayment of
long-term debt.

Mr. Nash. I take it their long-term debt position is considerably
less than their cash position. Is that right ?

Mr. Collins. Absolutely ; and that is all foreign debt, if I am not

Mr. Nash. You made the point several times that the capital re-
quired for entering into an increased maiket share in this business is
virtually impossible to acquire, for a variety of reasons.

In your judgment, but for the omnipresence of IBM's market posi-
tion, would debt and equity capital be available to this industry?

Mr. CoLLixs. Oh, absolutely, Mr. Nash. This is the most exciting
industry on the U.S. business scene today. It is a technology-based in-
dustry. It is growing rapidly.

The markets that are out there for the technology of the data proc-
essing industry are tralj exciting, and capital would be committed at
a very high level.

The late 1960"s was certainly an exaggerated period, but it reflected
the enthusiasm of investors to participate in some of the exciting
technological things that would be happening in this industry.

Part of my point is that we have subsequently learned it requires
more than just a good product, and more than just a capable manage-
ment, and more than an effective sales and field service organization to
succeed in this industry.

Mr. Nash. In your statement you said :

IB^M's marketing strategy is to introduce new products when competitive pres-
sure witliin the industry demands advanced products.

In the ahsenee of competition the rental-based nature of the industry makes
it more attractive to IBM to keep installed products out in the field for as long as
economically possible.

Could you elaborate on that insofar as this type of decision relates
to the rental -based nature of the industry ?

Mv. CoLLixs. Certainly.

It may be possible to go all the -way back to the beginning of the in-
dustry, although I am not absolutely sure.

It was Univac that shipped the first commercial computer system,
at a time when Univac had a relatively small portion of the tabulat-
ing machine market and IBM had the major portion of the tabulating
machine market.


IBM seemed to be initially reluctant to go into the computer systems
industry, and perhaps one of the reasons Avas th.e very large lease-base
of IB^M tabulating machine equipment that Avould have been obsolcted
if, indeed, the computer system began to replace tabulating machine
equipment at those data processing installations.

Another illustration is the example of peripheral subsystems, in tape
and in printer, through the decade of the 1960"s. Technological prog-
ress had been very slow, in part because it was in IBM's best interest
to keep that equipment, some of which was fully depreciated equip-
ment, out in the field generatiiAg revenues and profits on an annual
basis, and only obsoleted that equipment when forced to by competi-

I think perhaps an example that is a little bit more cuiTent relates to
an independent peripheral subsystems manufacturer, Storage Tech-
nologv, the independent leader in tape drive subsystems today.

In 1973 Storage Technology announced that they would have a tape
drive with a higher packing density than previous tape drives in the

The previous standard had been 800 and 1,600 bits of data per inch
of tape. IB]\I subsequently announced a new tape drive themselves,
with 6.250 bits per inch of tape.

Nov*', I can't say for sure whether or not IBM would have introduced
that new device if it was not for competition, if it was not for Storage
Technology, but because of the rental-based market there seems to be
little reason for IBM to introduce a higher density tape that allows
their computer users to store as much information on less tape and use
less equipment than he is currently using.

Mi\ Xasii. One defense of the present industry structure, we have
heard, is that users get good service, good quality products, at low

From your perspective I am obtaining the impression that you be-
lieve — and correct me if my impression is wrong — that the industry
rstructure has resulted in utilization of less than current technology,
and its prices are higher than current technology would allow. Is that
a fair understanding?

Mr. Collins. That is a fair understanding of my statement. I am
suggesting that may be more true in the future because a great deal
of technological progress came from the semiconductor industry and
from competitors within the industry that are no longer in the in-
dustry or are on their way out of the industry, so that the competitive
structure in the future looks like it will promote competition far less,
which suggests to me that technological progress will slow down in
the future.

Mr. Xash. In evaluating whether a positive or negative effect on
the stock market or stock market values of computer companies would
result from a restructuring of IBM, do you believe it is appi'opriate
to anah'ze the present market values of the computer companies and
ascribe different value to them which would exist, but for the present
industry structure?

By that I mean: Your table indicates that IBISI has a price per
earnings ratio of roughly 18, but Burroughs and other computer com-

Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 49 of 140)