SecProbe.io

CORRESP
1
filename1.htm

corresp

EDGAR CORRESPONDENCE

December 16, 2005

Ms. Kathleen Collins

Accounting Branch Chief

United States Securities and Exchange Commission

100 F Street, NE

Washington, D.C. 20549

Re: Cray Inc. – File No. 000-26820

Dear Ms. Collins:

     You originally wrote to us by a letter dated August 18, 2005, with comments regarding
certain of our Exchange Act filings. We responded by letter dated September 12, 2005. On
November 8, 2005, by facsimile, we received further comments and questions from you. This
letter responds to your comment letter dated November 7, 2005. Our numbered responses
correspond to the numbering of your comments and questions.

1.  Question:

In response to our Prior Comment No. 4 you indicate that forecasts for 2004 were profitable
based on continuing sales of the Cray X1 system and successful introduction of new products and
that your forecasts, along with the profitability achieved in the previous two fiscal years,
was your basis for reducing your valuation allowance by $58.5 million in the fourth
quarter of fiscal 2003. Provide us with your analysis that supports the reversal of the
full valuation allowance in the fourth quarter of fiscal 2003 and your apparent determination
that it was more likely than not you would generate taxable income to fully realize
your deferred tax assets. Also in your response tell us how you considered
paragraphs 17 through 25 of SFAS 109 in determining that the valuation allowance
should be reversed in the fourth quarter of 2003.

Response:

In early 2004, after the completion of the 2003 year, as noted in our September 12, 2005,
response, we had enjoyed eight consecutive profitable quarters and for the first time in our history,
had cumulative pretax
income over the prior three-year period, excluding unusual items and goodwill amortization. As
we announced on January 29, 2004, revenue for the fourth quarter
of 2003 was a record $67.2 million, an increase of 71 percent from the
prior year, and resulted in total revenue of $237 million for the full year,
a 53% increase from the prior year. That revenue growth was accompanied by the
achievement of significant additional objectives in 2003 and supported

1 of 13

management’s determination that it would generate
sufficient net income to realize the deferred tax asset. During 2003, we
eliminated our bank debt, increased
our market share, and delivered the two most powerful supercomputer systems
we had ever installed.

Based on our success in 2003,
our 2004 outlook for continued sales of the Cray X1, our Cascade and Red Storm development projects, our
contracted maintenance services, and our plans to introduce a major upgrade in the
second half of 2004 to the Cray X1, the Cray X1E, as well as a new product based
on the “Red Storm” supercomputer being developed for Sandia National Laboratories,
in a January 29, 2004 press release, we announced our 2004 plans:

“The outlook for 2004 and 2005 for
Cray is promising and we are confident in our long-term growth prospects as well as our ability to execute. We’ve set a
high bar for ourselves in 2004, planning to grow revenue to about $300 million while
achieving operating profit in the range of eight to twelve percent of revenue. Our plan
depends primarily on the timely and successful introduction of two new products, the
Cray X1E and commercialized Red Storm systems, both of which we plan to
launch in the second half of 2004.

Our outlook also relies on
maintaining momentum with our current customer base, including the
U.S. government. While the fiscal 2004 U.S. federal budget process
was recently completed, it remains difficult to forecast the outcome of individual
budget allocations and the resulting impact on Cray.”

Two other factors played a role
in our confidence toward 2004 and beyond. First, in mid-2003, we had
hired a leading high-performance computer sales
executive, Peter J. Ungaro, formerly at IBM, to head our sales program as
Vice President responsible for sales and marketing. Second,
although Mr. Ungaro was still putting his team together at the end of 2003,
he was in the process of hiring an experienced former employee of IBM to lead the
European sales effort, and was making changes throughout the sales team that we believed
would increase our sales momentum, particularly in the international sector.

Thus in early 2004, we had enjoyed two
years of profitable operations and we believed that we had put the period of
rebuilding Cray (from what we acquired in 2000 from Silicon Graphics Inc.) behind us.

We looked to future taxable income,
exclusive of reversing temporary differences and carry-forwards, and
were not basing our forecasts on tax-planning
strategies (SFAS 109, paragraphs 21 and 22). It is also important to
note that we did not reverse our entire valuation allowance; because of concerns about
the profitability of our overseas subsidiaries, we left slightly over
$6,000,000 of valuation allowance on our books pertaining to the operations of
our overseas subsidiaries.

2 of 13

Our 2003 pre-tax profit was $31.2 million in the
U.S. and a negative $10.2 million internationally, although we believed that our
overseas subsidiaries, under our transfer pricing approach, should become at least
minimally profitable with increased sales revenue and reduced service expenses.

In
your inquiry dated November 7, 2005, you asked us to comment on
how we considered SFAS 109 paragraphs 17 through 25 in
determining that the valuation allowance should be reversed in the fourth quarter of 2003:

In evaluating paragraph 17:

    •

    We measured deferred tax assets and liabilities for each entity in each
jurisdiction in which it operated at the applicable tax rate for that jurisdiction.

    •

    We evaluated the likelihood that some portion or all of a deferred tax asset would
not be realized for each entity and relative to the tax laws applicable in that
jurisdiction. In doing this, as mentioned above, we did not believe that it was more
likely than not that most of our deferred tax assets from our foreign subsidiaries
would be realized. Therefore, we did not reverse the valuation allowance relative to
those deferred tax assets.

In evaluating paragraph 18:

    •

    In calculating deferred tax assets and liabilities we used, and continue to use,
the tax rates expected to apply to taxable income in the periods in which the deferred
tax liability or asset will be settled or realized.

    •

    In 2003, there were no enacted future change in tax rates in any jurisdiction
affecting our calculations under SFAS 109, so the rates we used were the current rates
in effect in 2003.

In evaluating paragraph 19:

    •

    We did have an alternative minimum tax liability in 2003 and carried a deferred tax
asset for the related credit carry-forward.

    •

    Similarly, other jurisdictions in which our subsidiaries operate also
have alternative tax systems that may give rise to current tax expense even though a
subsidiary may have a loss for income tax purposes.

In evaluating paragraph 20:

    •

    As described below regarding our evaluation under paragraphs 23 and 24, all
available evidence, both positive and negative, was evaluated in determining whether
it was more likely than not that our deferred tax assets would be
realized.

3 of 13

    •

    At the end of 2003, we no longer had the negative evidence of “cumulative losses in recent years”
as we moved into a position of cumulative U.S. tax profit when
looking at the current year and two prior years. We used this three-year period as a
measure of “cumulative income/loss in recent years”, which is the period suggested by
paragraph 100 of the Basis for Conclusions section of SFAS 109, and was deemed
appropriate by management with advice from our auditors as the appropriate time period to consider, as it
provided heavily-weighted objective evidence as to our ability to utilize our deferred
tax assets in the near future.

    •

    The expectations for 2004 were very good. We planned for revenue of $300 million
and we expected to be profitable. Approximately $100 million in product and project
sales backlog existed at the end of 2003 and there were substantial sales transactions
being negotiated, including $75 million, in the aggregate, from Oak Ridge National
Laboratories, Pittsburgh Supercomputing Center and Korea Meteorological
Administration, all of which were later booked, shipped and accepted.

    •

    Based on internal financial models at that time, our projected taxable income would
allow us to use all of our net operating loss carry-forwards within three years. As
described in more detail below with respect to paragraph 24, we expected to utilize
our net operating losses even under more conservative assumptions.

    •

    While losses incurred in earlier years represent negative evidence, they were
largely attributable to the costs implicit in emerging from being a development stage
company.

    •

    As mentioned before, our revenue is derived from low volume, high priced
items with a long sales cycle. The timing of a few sales transactions could impact
revenue and profitability significantly in either direction.

In evaluating paragraph 21:

    •

    We estimated future taxable income in the U.S., exclusive of reversing temporary
differences and carry-forwards. Additionally, the reversal of some temporary
differences would give rise to an additional U.S. tax liability.

In evaluating paragraph 22:

    •

    We considered, but did not take advantage of, any tax planning strategies to
specifically prevent loss carry-forwards from expiring, as such strategies would not
have changed our conclusions on our valuation allowance judgment.

4 of 13

    •

    Additionally, while not considered heavily, there were some tax planning strategies
that, given our expectations of profitability, would have benefited our worldwide tax
position. For instance, a transfer/sale of certain intellectual property to a
subsidiary in a country with lower tax rates could accelerate the utilization of U.S.
loss carry-forwards and could provide for an overall lower effective tax rate.

Negative evidence considered per SFAS 109, paragraph 23 included:

    •

    Consistent losses prior to 2002, while we were emerging from the development stage,
created the large tax loss carry-forward included in deferred tax assets.

    •

    A risk that certain departments and agencies of the U.S. federal government would
spend less on supercomputers in the future.

    •

    The risks associated with new product development: whether product development
projects would be completed, would be competitive, would be on time, and would be at a
reasonable cost. At the time, we believed these risks were outweighed by the success
of the Cray X1 in 2003 and the rather straight-forward nature of the product upgrade
to the Cray X1E, and that the Red Storm development program was then substantially on
schedule.

    •

    The risks associated with our competition’s potential impact on our revenues and
margins. Our sales are represented by a small number of sales transactions with high
average selling prices. Delays or winning or losing just a few of these sales
transactions could have a large impact on our operating results.

Considering the criteria for positive evidence (as described in SFAS 109, paragraph 24):

    •

    We believed that our financial position was strong and sufficient to execute a
profitable strategy. We had $74 million in cash and short-term investments and almost
no debt as of December 31, 2003.

    •

    We had strong recent performance and we believed that the losses in 2001 and
earlier were due to emergence from our development stage period in which we had few
new products since our acquisition of the operating assets of the Cray Research business unit
from Silicon Graphics Inc. in
2000. U.S. pre-tax earnings were $5.1 million in 2002 and $31.2 million in 2003.

    •

    Our expectations for 2004 at the time were for increased revenues and continued
profitability. Between existing backlog (about $100 million at December 31, 2003) and
sales transactions being negotiated, as detailed above relating to our evaluation of
paragraph 20, we believed we had reasonable

5 of 13

    visibility toward achieving 2004 goals and planned our incentive compensation plan for employees and executives accordingly.

    •

    We had been successful in selling our Cray X1 line in 2003 ($144 million) and
expected well over $100 million in 2004 revenue from this product line. Additionally,
we expected to drive additional growth through a significant upgrade
to the Cray X1 in
the second half of 2004 (the Cray X1E system), completion of the Red Storm project
with Sandia, and the sale of Red Storm-based products (the Cray XT3 system).

    •

    The first expiration of tax loss carry-forwards was in 2010, or seven years from
the end of 2003. The last year for the expiration of tax loss carry-forwards was in
2022, or 19 years from the end of 2003. Some tax planning strategies might have been
possible to extend the dates further but were not considered in our analysis. Assuming
U.S. pre-tax taxable income remained at 2003 levels ($31.2 million) then full
realization of deferred tax assets would take just 5.4 years ($58.6 million U.S.
deferred tax asset divided by ($31.2 million annual U.S. pre-tax profit X 35% tax
rate)). This would be complete at the end of 2009.

In evaluating paragraph 25:

    •

    We believed, as set out above, that the positive evidence supporting the reversal
of our valuation allowance outweighed the negative evidence and was more objective in
nature than the negative evidence.

Under
SFAS 109, paragraph 24(c), we believed we had moved
beyond the periods of initial loss, had enjoyed a period of
profitability and were entering a period of sustained profitability.
In deciding to substantially reverse the tax valuation allowance, we considered the
issue carefully over time and sought the expert advice and judgment of others. We
first reviewed the possibility of the reversal with our auditors and our Audit
Committee in October 2003. A final decision was made in late January 2004, after
full review of fiscal year 2003 results, our history, management’s plans for 2004 and
subsequent years, and the accounting requirements, with the strong concurrence of our
auditors and approval by our Audit Committee. This is documented in the minutes of the January 27, 2004,
meeting of the Audit Committee. We received unqualified audit
opinions with regard to our financial statements for the years ended
December 31, 2003 and 2004.

Just to allay any possible concern or
misunderstanding, the decision to reverse the tax valuation allowance at the end of 2003 had no impact whatsoever
on management’s incentive compensation for that year, which was based on operating results, not net income,
and in which non-operational adjustments, such as this change, were disregarded by the
Compensation Committee and full Board.

6 of 13

We
believe we exercised the judgment required under SFAS 109,
paragraph 25, and appropriately considered the relative impact of both negative and positive evidence. We
were aware of the risks involved, as disclosed in our public documents. But we had
confidence in our ability to reach, if not exceed, our targets supported by recent success,
and concluded that it was more likely than not that we would recover the full U.S. valuation
on deferred tax assets. Under these circumstances we continue to believe that the reversal
of the valuation allowance was appropriate at the time. We appreciate that this judgment
may be questioned with the full clarity of hindsight. Nevertheless, we made a reasonable
judgment at the time with the information then available to us in accordance with the
appropriate accounting standards.

As a result of a series of events that took place
ov