Description
Jazz
Technologies (JAZ) is a busted going-public transaction for (what is currently)
a single wafer-level semiconductor fab facility located in Newport
Beach, CA. The fab is focused primarily on specialty
analog and mixed-signal products. The
entity has a decent and high profile management team, several readily-identifiable
catalysts and, I believe, an asymmetric risk-reward profile that makes it an interesting
potential investment for anybody who can look thru the next couple of quarters. The stock sells at a material discount to its
public comparables, at roughly 1x and 11x trailing revs and EBIT, respectively
(vs. comp range of 2-5x revs and 15-30x EBIT).
In addition, given likely changes in the corporate profile over the next
couple of years the upside from an intrinsic value perspective could be
impressive, with potential peak earnings power implying a 4x P/E multiple off
of the current share price, and mid-cycle earnings power implying an 8x PE.
Steve308
posted a couple of very nicely timed write-ups on SPAC warrants several months
ago. JAZ started life as a SPAC as well,
and a little over a month ago received shareholder approval to purchase Jazz
Semiconductor. Jazz Semiconductor was
spun out of Conexant in 2002 in conjunction with a concurrent investment by
Carlyle Group (the Jazz Semiconductor shareholder group prior to purchase by
JAZ included Carlyle, Conexant, and RF Micro).
Approval for SPAC deals has become markedly more challenging in 2007, as
the backlog of deals grows and shareholders have become more discerning in
terms of their willingness to approve transactions. In JAZ’s case, a number of last minute
gyrations took place in order to get the deal pushed through, including a
giveback by the management team of 35% of its carried interest, an incremental
$7.3mm investment by the SPAC’s officers and directors, and a $10mm investment
by Conexant. The latter two investments
were used to fund the repurchase of shares from shareholders who were expected
to vote against the proposed transaction.
In addition, the deal’s underwriters leaned heavily on a number of
equity buyers and “traditional” SPAC investors to purchase equity prior to the
vote and subsequently vote for approval.
Underwriters of SPAC deals have a material incentive to help get the
deals approved, as a portion of their underwriting fees are contingent on
approval. Traditional SPAC holders will
sometimes step in and help support a deal in this manner, so that they can
continue to get priority allocations when the deals first go public. In JAZ’s case, the efforts paid off and the
deal was approved with the narrowest of possible margins. In the days and weeks since approval,
however, the stock has dropped precipitously (in what has become recently a
fairly common pattern) as the “unnatural” holders who got jammed with equity
pre-deal in order to help ensure approval began dumping their shares. As the selling depressed the price, other
marginal holders jumped on board, stop losses likely got hit, and the stock
dropped by roughly 25% to its current trading level of $3.90/shr. At this point, the shares seem to have
stabilized and the shareholder base turnover is likely largely complete.
As
mentioned, the company currently operates a single fab facility in Newport
Beach, CA, manufacturing
specialty analog and mixed specialty products.
Mixed analog/digital products are often used in communications-oriented
applications (typical end markets include cell phone, WiFi, wireline, modems,
GPS, and Bluetooth apps) where a nexus is required between electromagnetic
analog signals (i.e. wireless spectrum communications) and traditional digital
ICs. These products occupy a strongly
growing subsegment of the semiconductor industry, with estimates of growth
rates roughly double those of the overall semiconductor industry (18% in 2006
vs. 9% for overall semi). The products are
typically characterized by less need for extreme miniaturization, but correspondingly
more specialized design requirements.
The niche has historically not been large enough to merit focus by
industry heavyweights whom are typically focused on much higher volume leading-edge
CMOS circuits. The product requirements
allow the manufacturers to use older (i.e. 2nd and 3rd
generation) production equipment, which can often be purchased at attractive
prices after having been fully depreciated by the original owners.
JAZ
management includes a number of well-known ex-Apple executives, including CEO
Gil Amelio (ex-CEO of Apple), President and COO Ellen Hancock (ex-EVP/CTO of
Apple), and EVP/CTO Steve Wozniak (co-founder of Apple). Gil and Ellen also served, respectively, as
Chairman/CEO and EVP/COO of National Semiconductor and, in addition, Gil was
the former President of Rockwell Communication Systems (now Conexant – Jazz
Semi’s original owner). This management
team has deep experience in the semiconductor industry and knows the asset that
they have purchased well. While their
track record at prior endeavors has been mixed, I consider them to be an
extremely high caliber team for a company of this size.
Since
closing the acquisition, management has moved aggressively to cut costs at the Newport
Beach facility.
Roughly 15% of salaried employees were let go and $6mm of total costs
were taken out. This more than makes up
for the incremental $3.5mm of “SPAC” costs (i.e. senior mgmt comp + public
company costs) that will arise.
Moreover, mgmt believes that there may be as much as $10mm of
incremental costs that can come out of the existing business over time. The CFO, Paul Pittman, believes that they now
have a cost structure that will allow them to be cash flow breakeven (after
debt service) at the bottom of the cycle, and materially cash generative as the
cycle is more favorable.
Last
year, Jazz Semiconductor did $230mm of revenues and $31mm of EBITDA (adjusted
for non-recurring deal expenses and the like).
With $10mm of maintenance capex at the facility, this yields what I am
calling $21mm of “economic” EBIT. As
mentioned, there are new public company costs on top of the entity, but these
have been more than offset by cost cuts at the operating level. 2006 was a near-term peak year for the
company’s products, so 2007 results are likely to come in. First quarter 2007 results showed a
sequential decline on the top line as well-publicized inventory adjustments in
the channel have temporarily depressed demand.
This adjustment appears to now be passing (see recent TI results), and
growth should resume by the back half of the year.
Capital
structure includes the following:
23.9mm
common shares
$167mm
convert; converts @ $7.33/shr
46.2mm
warrants; $5 strike
$30mm
cash
With
the convert and warrants both underwater currently (28% upside until warrants
are in the money, near double before converts are in the money), I get a basic
EV of $93mm equity (23.9mm shrs @ $3.90/shr) + $137mm net debt (convert less
cash) = $230mm. This yields trailing
multiples of 1.0x revs and 11x “economic” EBIT.
These multiples represent material discounts to public competitors that
trade at 2-5x revs 15-30x EBIT. On a
revenue multiple basis, some discount is arguably warranted as JAZ’s Newport
Beach-based facility puts it at a cost disadvantage to its Asian-based
competitors. At today’s levels, you have
28% upside in the current stock price before you begin to feel dilution from
the warrants, and the stock has to nearly double before the converts come into
the money.
What
makes the investment prospect interesting, however, is not the current company
profile, but what the management team plans to create with it. This high profile management team did not
sign on to run a single fab entity. The
plan is to leverage the company’s IP onto a much broader platform, and grow
revenues and cash flow correspondingly.
Within the year, management expects to close on the purchase of an
Asian-based 2nd generation fab.
This will have a number of important ramifications: 1) the Asian-based
fab will significantly lower the company’s cost structure, giving it an
enhanced ability to prevail in competitive bidding situations, 2) addition of a
2nd fab will materially increase the prospective customer base, as
many customers refuse to do business with single-fab entities due to concerns
over business risk, 3) enhanced ability to leverage the IP and corporate
overhead over a much larger revenue base.
For
rough order of magnitude, management believes that it will spend roughly $180mm
to buy a fully-depreciated Asian-based fab (several are available, and talks
are ongoing) and an additional $25mm to refurbish. Year 1 that fab probably produces $30mm of
EBITDA (6.8x), 3 years out $65mm of EBITDA (3.2x). Management believes that ultimately such a
fab could produce $500mm of revenue and $115mm+ of EBITDA in a fully-loaded
(i.e. peak cycle) year. More
importantly, by adding a second fab, JAZ likely gets a 500bp improvement in
potential margins, either mid-cycle to mid-cycle or peak-to-peak, given the
combination of operating leverage and improved cost position.
There
are clearly a lot of variables to consider when trying to value JAZ pro-forma
for a 2nd fab purchase, but here is my stab at both a peak and mid-cycle
scenario:
Peak:
$750mm
revs ($275mm existing fab, $475mm Asian fab)
23% EBITDA margins (per mgmt guidance) =
$173mm EBITDA
$50mm
maint capex
$123mm
economic EBIT
$7.5mm
interest expense
$75mm
fully-taxed net income (35% tax rate)
$0.93/shr
net income (see shr count below)
Funding
for new fab (50% equity, 50% debt):
Uses:
Purchase
price = $205mm (incl incr capex)
Sources:
$15mm
cash on hand
$87.5mm
debt
$102.5mm
equity (issued at current price)
Cap
structure:
23.9mm
shrs (current)
22.8mm
shrs (convertible converts to equity)
26.1mm
shrs (issued to finance 50% of fab @ current price)
7.7mm
warrants (assume treasury method – repurchased @ $6/wrnt)
=
80.5mm shrs
$87.5mm
debt
$15mm
cash
Multiples
@ $3.90/shr (current):
4.2x
net income
2.2x EBITDA
3.1x economic EBIT
Mid-cycle:
$650mm
revs ($240mm existing fab, $410mm Asian fab)
18% EBITDA margins (per mgmt guidance) =
$117mm EBITDA
$50mm
maint capex
$67mm
economic EBIT
$7.5mm
interest expense
$39mm
fully-taxed net income (35% tax rate)
$0.48/shr
net income (see shr count below)
Funding
for new fab (50% equity, 50% debt):
Uses:
Purchase
price = $205mm (incl incr capex)
Sources:
$15mm
cash on hand
$87.5mm
debt
$102.5mm
equity (issued at current price)
Cap
structure:
23.9mm
shrs (current)
22.8mm
shrs (convertible converts)
26.1mm
shrs (issued to finance 50% of fab)
7.7mm
warrants (assume treasury method – repurchased @ $6/wrnt)
=
80.5mm shrs
$87.5mm
debt
$15mm
cash
Multiples
@ $3.90/shr (current):
8.1x
net income
3.3x
EBITDA
5.8x
economic EBIT
So
what is it worth? I believe that it is
easy to justify a 15x multiple off mid-cycle EPS for an industry subsegment
that grows the top line at >10% over the cycle. If this is the case, the stock could be worth
$7.20/shr, or roughly 85% above today’s price.
At this price, the convert is still marginally out-of-the-money and
associated dilution is lower. However, I
have chosen to assume conversion in order to be conservative and, in addition,
because there are so many moving pieces.
You can obviously make your own choices.
Additional
upside could come from better tax planning (tax rates below 35% are likely,
given > half of production will be in Asia), as well
as the associated tax shield resulting from the asset purchase of the 2nd
fab (since I am using “economic” EBIT, based on maintenance capex, I am implicitly
ignoring this tax shield). Moreover, as I
mentioned, the company can currently shield roughly $50mm of pre-tax income, which
I have not included in my valuation (present value of probably $0.40/shr
assuming it all accrues to current existing equity outstanding). Finally, management has been extremely
aggressive recently in purchasing warrants as the prices have come in, thereby
materially lessening future potential dilution to the extent the operating plan
works as hoped.
Currently,
it is only covered by Think Equity, the underwriter, but I expect it to garner
incremental sell-side interest over the next couple of quarters.
Risks:
Some
remaining customer concentration (Conexant), current downturn in semi cycle is
more prolonged than expected, warrant dilution, overpayment for new fab
Catalyst
Turnover in shareholder base abates, sell-side coverage launch, end of semi inventory correction (corresponding reacceleration of top line), cost cutting measures take hold, purchase of new fab, customer adds