Avago (AVGO) and Broadcom (BRCM) …and Intel/Altera

Two days ago the rumor hit Wall Street that chipmaker and serial acquirer AVGO had found its newest target, BRCM.  Yesterday the offer was announced:  cash and AVGO stock, in approximately 45/55 proportions, totaling $37 billion.

my thoughts

When customers in a given industry group become bigger and more powerful, the natural response among suppliers is to do the same.  This is part of what is going on here.  More than that, AVGO appears to seek out companies whose technological virtuosity far outstrips their management skills.  So it gains not only the marketing benefit of size but also the rewards of improving the profitability of firms whose main virtue has been their intellectual property.

What’s striking about this deal is that in revenue terms AVGO is more than doubling its size.  Although I have no intention of selling the AVGO shares I own, experience says that acquirers often bite off more than they can chew when they make the jump from small acquisitions to super-size ones like this.

One of AVGO’s rumored other targets had been Xilinx (XLNX), the junior partner with Altera (ALTR) in the field programmable gate array duopoly.  I had thought that ALTR would feel more favorably disposed to overtures being made by Intel (INTC), given the possibility that AVGO would buy XLNX and turn the firm into a much more aggressive competitor.  That threat is now gone.  INTC must now rely on pressure on ALTR management from its major shareholders (shareholders are, after all, legally the owners of ALTR and the employers of management) to return to the negotiating table.

As a practical matter, managements have a lot of autonomy, despite the fact that we the shareholders are, technically speaking, the bosses.  Wall Street seems to believe that ALTR is holding out for a higher price from INTC.  While that may be the rhetoric being used, I think the real issue is more basic.  Who would want to go from being the master of all he surveys as the top dog (and treated as a demigod) at a major publicly traded company to being a near-invisible division head in a conglomerate?

an Intel (INTC) – Altera (ALTR) deal re-emerging?

Market gossip is that ALTR recently refused a friendly offer from INTC at $53 a share.

Speculation resurfaced yesterday with rumors that talks have started up again.

The catalyst seems to be the fact that serial acquirer Avago (AVGO–I own shares) appears to be considering a bid for ALTR’s rival Xilinx (XLNX).

AVGO seems to have a knack for finding firms that have excellent technology but which, for one reason or another, find it difficult to achieve consistent profit growth.  AVGo buys them, reorganizes them and puts the profit machine into high gear.

In this case, the sub-industry involved is the sleepy world of field programmable gate arrays (FPGAs), dominated by the cozy duopoly of ALTR and little brother XLNX.  AS the name suggests, FPGAs are chips whose program structure is not hard-wired (those are application-specific integrated circuits–ASICs).  So they can be reprogrammed, upgraded, debugged…even after they’ve been put into machines that are now in use.  This allows manufacturers to get, say, cutting-edge telecom equipment into customers’ hands very quickly.  The drawback is cost.

The AVGO move suggests the FPGA arena is about to become considerably more competitive.   AVGO/XLNX would be four times the size of ALTR, implying easier access to capital and the ability to offer a much wider variety of products to customers than ALTR.  This suggests ALTR realizes the status won’t be quo for much longer and it needs to be part of a bigger entity in order to compete.

To my mind, the big winner in all this would be INTC.

Intel (INTC) and Altera (ALTR): the numbers

Let’s look at ALTR before word leaked to Wall Street that INTC was considering buying the firm.

the basics

ALTR was trading at about $35 a share, with earnings of, say, $1.75 a share in prospect for 2015   …in other words at about a 20x multiple.  The long-term growth rate of eps is probably in the low teens.   The market cap was $10.5 billion or so.

ALTR is one of two firms that together dominate the highly specialized market for programmable logic devices–a relatively stable, by technology standards at any rate, area.

20x for 10%-12% earnings growth doesn’t sent me running to the computer to place a buy order.

where the value is

small stuff

ALTR had $1.6 billion in net cash on the balance sheet at the end of 2014–that after spending $655 million buying back stock last year.

Yearly SG&A is running at about $300 million.  Let’s say INTC could eliminate half of this by substituting its own corporate infrastructure.  That would be enough to boost eps by 25%, so we’re looking at a 16 multiple on current earnings, which would be more reasonable.

the big attraction–intellectual property

The main source of value for INTC is the company’s accumulated knowledge, experience and computer code for creating and operating PLDs. How do we measure that?

The simplest, and only straightforward, thing to do is to add up R&D expenditures over, say, the past decade and see what that totals.  This will be an understatement, of the value of ALTR’s intellectual property because:

–there will always be some R&D related expenditure elsewhere on the income statement,

–duplicating the firm’s accumulated knowledge means spending in today’s and tomorrow’s dollars–not yesterday’s.  The former is always more expensive, and

–we won’t capture stock based compensation.


1. For ALTR, the 10-year total R&D  is $3 billion.  Arbitrarily add $500 million for stock based compensation.  Add in the net cash.  We get a total of $5 billion in “asset value.”  That doesn’t stack up well with ALTR’s pre-leak market cap.

2. Another approach.  Current R&D expenditure is running over $400 million a year.  Let’s say it would take ten years of spending at the current rate to duplicate ALTR’s intellectual property.  That gets us to $6 billion in “asset value.”

3. Let’s consider the future earnings stream (this is arguably just dart throwing).  Ignoring SG&A synergies, and with 300 million shares outstanding, $1.75 a share in eps translates into net income of $525 million.  Let’s say earnings in eight years are double that, or $1.05 billion.  If earnings progress in a linear fashion (another incredible simplification–but, hey, this is what securities analysts do), then the total earnings over the next eight years will be just over $6 billion.  (Why eight years?  My experience in analyzing corporate behavior in takeovers is that eight years is the outer limit of future earnings that companies are willing to pay for in an acquisition.)


Okay, we’ve got one figure, #1, that’s too low and another, #3, that’s too high- (and a giant leap of faith).  Let’s add them together!

They total $11 billion.

Ta da!

That gets us to around the market cap of ALTR before the leak.  To be clear, I’m not willing to defend to the death anything I’ve written so far.  But Wall Street had to be tacitly thinking something like this for the price of ALTR to be at $35.

the leak   …and a dilemma

Look back at #2 above.  For INTC, the alternative to acquiring ALTR is doing #2.  This would be expensive.  More important, it would be time-consuming–time that INTC probably doesn’t have.  And there’s the risk that its effort wouldn’t be successful.

Therefore, the value of ALTR is higher to INTC than to you or me.  INTC is probably also figuring that it can expand the ALTR business dramatically over the coming years by stuffing every one of its servers full of ALTR chips.  Therefore, $10 billion price leaves room for the acquisition to be accretive to earnings in a few years.  Also, SG&A synergies.

On the other hand, any of us with a loose $10+ billion will probably find a lot of things we’d rather do than plunk it all down to buy ALTR.  For us, $35 a share is a pretty rich price.

this brings us to the leak…

Both sides can make up numbers as well as I can.

Both know there are no other suitors.

Both know that INTC really wants ALTR.

Hence, the leak, which I would bet came from bankers representing ALTR.  The idea is–let the market bid up the price/decide what the price should be.

I’m not sure whether the leak makes the situation better or worse.  My guess is that a deal gets done somewhere between $35 and $40.






Intel (INTC0 and Altera (ALTR): implications

What can we conclude from INTC’s interest in acquiring ALTR?

–when I became interested in INTC as a stock a couple of years ago, it seemed to me that the firm could be viewed as having two businesses–a high-growth one selling servers and a low-growth, cash cow one selling chips for PCs.   At the time, I thought the server business alone more than justified the then stock price, and that the PC business was mainly important for its contribution to overhead and its free cash flow generation.  A desire to acquire ALTR seems to confirm that this is also INTC management’s view.

–good companies periodically reinvent themselves.  After a period of stagnation, this appears to be what INTC is doing

–the threat of low power servers run by ARM chips is serious

–my guess is that a bid will take the form of all or mostly INTC stock.  An all or largely cash offer would imply either that INTC thinks its shares are deeply undervalued, that debt financing is too ridiculously cheap to pass up, or that long-suffering ALTR shareholders want  to declare investment victory and move on.

–an INTC-ALTR merger spells trouble for Xilinx (XLNX), the main competitor to ALTR

–the main source of value in ALTR is its software.  Assessing that, thorough accumulated R&D spending, is the key.

Numbers tomorrow.

Intel (INTC) and Altera (ALTR)

Late last Friday afternoon a rumor reached Wall Street that INTC is in talks to acquire ALTR, causing a sharp rally in ALTR shares and a modest one in INTC’s.  In hindsight, INTC appears to have been headed down this path for a couple of years, as a prescient article in the Electronic Engineering Journal (titled “When Intel Buys Altera”) pointed out last June.

Why a deal is potentially crucial for INTC:

1.  what ALTR does:  The company is one of the two dominant makers of Field Programmable Gate Arrays (FPGAs)–the other is Xilinx.   FPGAs are logic devices.  What makes them unusual is that they contain software that can be updated, revised or reprogrammed after the servers or telecom equipment (the two big markets for FPGAs) they’re in have already been built and installed.  The traditional upside of FPGAs is that they allow customized equipment to be put into the field and fine-tuned quickly.  Their downside is they’re more expensive than the pure-hardware alternative, ASICs (application-specific integrated circuits).

2.  INTC and Moore’s Law:  A factory to make current-generation INTC chips costs about $3 billion – $4 billion.  A next-generation factory, using much different equipment, will cost maybe $14 billion.  Samsung has already said it will build one; INTC says it’s too risky to build one by itself.  How, then, does INTC retain its technology/speed advantage over rival chipmakers?

3.  an INTC chip + a FPGA:  as reported in the EE Journal, INTC says linking an INTC microprocessor with a FPGA in a server can boost performance by 10x.  Bind the interface between the two closely enough can double performance again.

In other words, INTC + ALTR = huge step forward in chip performance.

why a merger and not a joint venture?

To my mind, the risk to both parties is too high for a joint venture.  INTC would have its lucrative server business in jeopardy if it committed to the FPGA route and the parties ever parted.  ALTR would have to devote a lot of resources to making its programming tools easier to use, potentially diverting attention from its telecoms customers.

More tomorrow.