Texas Instruments vs. Intel

First off, I'm sure many of you spent your Saturday night at home, curled up in front of the TV watching The Motley Fool Money-Making Life-Changing Special on PBS, right? Your eyes glued to the screen with a pad of paper and pencil in hand, ready to take notes on the investing and personal finance savvy pouring forth from the brothers Gardner -- Tom and Dave. But, in case you missed it, it is worth viewing -- really. I work here and there were some things I learned for the first time. But more importantly, I was entertained. Tom and Dave are at their best with the PBS hosts in between segments. Check out the schedule of airtimes to see when you can catch the show in your area.

While digging through the financials of all sorts of well-known companies in search of contenders for the Top 25 Rule Maker report (free with our ongoing seminar), one company surprised and disappointed me the most -- Texas Instruments (NYSE: TXN). I was so ready to make the case that this leader in programmable digital signal processors (DSPs) was a worthy Rule Maker candidate. But I ran into some trouble with a little thing called cash. More on that later, but first, let me explain why Texas Instruments (TI) was in contention.

TI is a leader in DSPs and high-performance analog integrated circuits. A CNET news story defined these products well, "Digital signal processors are chips that refine digital audio and video data so they can be efficiently sent across computerized networks. An analog converter actually translates real-world sounds and images into ones and zeroes while the DSP removes noise and creates a more compact file." Put another way, any sounds or pictures created from the real world (which is analog by nature) requires a DSP to get into the digital world. DSPs are used in devices like cell phones, modems, and most things wireless.

Last year, nearly 60% of all handsets sold used Texas Instrument's DSPs. With wireless juggernaut Nokia (NYSE: NOK) predicting global mobile phone sales for 2001 in the 500 to 550 million range, things looked good. Wireless device sales are lapping PC sales and TI is at the heart of it all. And even though market data providers can't agree on TI's exact share of the market, they do agree on one thing: TI dominates, and it has approximately half the DSP market share.

TI's bread and butter is the general-purpose programmable-DSP. This product has seen phenomenal growth over the last decade, concluding last year with 37% growth according to IC Insights. That pace is expected to slow to the single digits to mid-teens in 2001, but even a dramatic slowdown this year can't stop the digitization of the world. The way I see it, this downturn is just a hiccup in the cyclical semiconductor market, and with TI trading at $35, down from a 52-week high of $99, what better time to invest than now? Well, not exactly.

What I found was a great company in a great industry with very little free cash flow (FCF). (Free cash flow is cash flow from operations less capital expenditures, and we use it to measure a company's earnings because it is less subject to accounting manipulation by Wise management. I also adjusted FCF by removing the tax benefit from employee stock plans to more closely represent operating results. For more on this tax benefit and how dramatically different free cash flow and net income can be, see these lessons learned from Lucent (NYSE: LU).)

To get a better sense of how TI stacked up to it competitors in the semiconductor business, I compared its financials to those of Rule Maker and fellow semiconductor maker, Intel (Nasdaq: INTC) -- not an easy match for even the best run company. Below are some key numbers for the past three years. Note, Intel's 2000 numbers are for the first nine months of 2000 since the year-end cash flow statement is not available yet. (All numbers in millions of dollars.)

             Texas Instruments
                         1998    1999    2000
Cash Flow from Ops.     $1386    2357    2185
Net Capital Exp.         1067    1398    2762
FCF                       319     959    -577
Net Income                452    1451    3058

                   Intel
                         1998    1999    2000*
Cash Flow from Ops.     $9191  11,335    9449
Net Capital Exp.         3557    3403    4251
FCF                      5634    7932    5198
Net Income               6068    7314    8342
* 9 months ended 9/30/00

First, let's look at the Cash King Margin (CKM). As Rich explained last week, CKM is similar to the net profit margin but uses free cash flow instead of net income. It is calculated as follows: free cash flow divided by sales.

      Cash King Margin
           1998    1999    2000
TI         3.6%   10.0%   -4.9%
Intel     19.9%   25.3%   17.4%*
* 9 months ended 9/30/00

Even though Intel's CKM declined last year, anything above 10% for a capital-intensive manufacturing company is impressive and anything close to 20% is simply outstanding, especially when you consider that a new microprocessor plant can run upwards of $2.0 billion. Even other dominant companies with deep roots in manufacturing high-tech electronic devices like EMC (NYSE: EMC), Nokia, and Sun Microsystems (Nasdaq: SUNW) can't stand up to Intel's CKM. Their CKMs for the nine months ended September 30, 2000 were 12.1%, 5.6% and 11.0%, respectively.

TI's negative CKM in 2000 makes me nervous and I don't care for the large fluctuations either. So, I went back a few more years, but found the same story -- low or negative CKMs. A recent CFO Magazine article shed some light on TI's high capital expenditures. In it, TI's CFO noted that his company has shied away from outsourcing its manufacturing operations in order to maintain control over the quality and supply of its products, and to benefit from the "synergies we get from integrating process technology with product technology." In fact, he even went so far as too say that TI had been light in its capital expenditures and that it plans to continue to invest heavily. Uh oh.

Now, don't get me wrong. I understand the need for capital expenditures. They're a necessary evil. Computers need to be upgraded and machinery needs to be oiled and replaced. Capital expenditures are also vital to maintaining a competitive position and to keeping up with the latest manufacturing processes and technologies. In fact, Intel recently announced that it is upping its capital spending in 2001 by 12% to $7.5 billion, and I applaud that decision. The difference between Intel's and TI's capital expenditures is that Intel's are largely value-added and discretionary while TI's are not.

I've seen how Intel produces solid free cash flow in the normal course of business as well as upgrade cycles, but TI simply does not have the wiggle room in upgrade cycles because it does not have the free cash flow from normal operations. TI just does not have the gross margins to support its high level of capital expenditures. There are few companies that do.

The nature of the semiconductor business is such that these companies have to be able to invest in billion-dollar factories. Unfortunately, TI has not proven it can do that and still produce a healthy level of free cash flow. Perhaps some day TI's spending spree will pay off, but until it does, free cash flow will continue to suffer and this investor will stay on the sidelines.

Todd Lebor is a not Tiger Woods' caddy, but he did shake hands with Ben Crenshaw once. At the time of publication, he owned shares of Intel and Sun Microsystems and was secretly negotiating a hat, shirt, ball, and bag deal with Nike, despite Titleist. Todd's other holdings can be found online, as can the Fool's complete disclosure policy.