We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.

[nextpage title=”Introduction”]

It’s been one year since AMD announced that it would acquire ATI and we decided to make an in-depth analysis from both the technical side and the financial side of this operation one year later. We found out some concerning figures. Read on.

AMD buying ATI was probably the most commented news from the past 12 months. Even though there were several other acquisitions, mergers and bankrupcies on the market, none was on the size of this one, officially valuated as a USD 5.4 billion deal.

Even though it was announced on July 24th, 2006, the deal was only concluded at October 25th, 2006, three months later. So it’s not been truly one year since AMD bought ATI; the 1-year anniversary is for the announcement of the deal.

For the PC hardware community AMD buying ATI made sense: it was a CPU manufacturer buying a GPU manufacturer, thus increasing its position on the PC hardware industry. As AMD has plans to release a CPU with an integrated GPU, buying a company specialized in GPUs makes a lot of sense.

In reality ATI wasn’t really a GPU manufacturer, but a GPU designer, since it didn’t have any factory. They designed the chips and sent the project to chip manufacturers like TSMC and UMC for production (the same thing happens with NVIDIA, for the record). Because of that there is a huge delay between projecting a chip and delivering it to the market. Since AMD has its own factories, they could start manufacturing GPUs, shortening the time between development and delivery. Not only that, having its own factories allow several fine tunings that improve both design and production.

But several questions arouse at the time. Would this merger make AMD a stronger company? Would it increase ATI’s and/or AMD’s market share? And what about their direct competitors, NVIDIA and Intel? Would they join forces somehow? Would AMD and ATI merger increase or decrease NVIDIA’s or Intel’s power? These are some of the questions we would like to analyze in this article.

“It makes sense” was – and probably still is – the prevalent feeling of the PC hardware community and industry. However what is the feeling of Mr. Market?

[nextpage title=”Market Analysis: ATI and AMD”]

A lot of people may think that stock market analysis belongs solely to financial media, and that we should not cover this kind of analysis. However, we are from a different opinion. Financial market matters – a lot.

Issuing new stock is the cheapest way for a company to finance its expansion, as the company does not need to pay interest on the money they got from the stock issue. The company may pay dividends over the stock, but this is up to the company – the company isn’t obliged by any governing body to pay dividends and they may not pay any dividends at all, if they wish.

People will only buy the company’s stock if they think they will make money. Let’s be honest. Nobody will buy AMD stock only because they like green or only because they saw AMD logo on a Ferrari Formula 1 racing car. They want money in their pockets, through increase on the stock price and/or through dividends payment.

So if the market isn’t seeing a particular stock as a good stock for buying, the company may sell less stock than they had planned to when issuing new stock, frustrating their plans.

The price of a stock is an opinion of the people on the stock market on how much the company is worth at a given time. Multiplying the stock price by the number of issued stocks we have how much the market thinks the company is worth, i.e., at that particular time, how much money is needed to buy all the stocks. This is called market capitalization or simply market cap.

The stock price and the market cap are the two metrics we are going to analyze, because we were interested to know how much money an individual investor like you could make (or lose) if you had invested on AMD and also because we wanted to know how much the market thought AMD was worth before and after it bought ATI.

We are going to take a look at four particular dates: July 21st, 2006, the Friday before AMD publicly announced its intention to buy ATI; July 24th, 2006, the day AMD announced its intention to buy ATI; October 24th, 2006, the last day of ATI stock trading; and July 20th, 2007, last Friday, one year after AMD had first announced that it would buy ATI.

The scenario for ATI stock was the following. On July 21st, 2006 it closed at $16.56, closing at $19.67 on July 24th, 2006 after AMD announced that they would buy ATI. That was an impressive 19% increase on the company stock price, however you have to keep in mind that at the closing of the deal (which happened Oct 25th, 2006) AMD would buy each ATI stock paying $16.40 cash plus 0.2229 shares of AMD stock for people still holding ATI stock. Since AMD closing price on July 24th was $17.39, each ATI stock was actually valuated at $20.28 ($16.40 + $3.876, which is $17.39 x 0.2229), so even with this price increase it was still quoted a little bit below its “real” value.

On the day the deal was closed, Oct 25th, 2006, ATI was valuated at $5 billion ($19.53 closing price on Oct 24th, 2006 times 258.26 million outstanding shares). AMD paid $4.2 billion in cash for it plus gave away 57 million AMD shares. Keep these numbers.

ATI Stock PerformanceFigure 1: ATI stock performance.

And how about AMD? On July 21st, 2006 it closed at $18.26. On July 24th, 2006 it closed at $16.90, an 8% loss on the day AMD announced it would buy ATI (what is quite intriguing is that on July 20th, 2006 AMD closed at $21.65, suggesting that people knew what was going to happen). On October 24th, 2006, last day of ATI stock trading AMD stock closed at $20.32. And last Friday AMD closed at $15.50.

So if you had bought AMD stock on the day it announced they would buy ATI, you would have 8.3% less money today. If you had bought AMD stock on the day it finished buying ATI (October 25th, 2006) you would have 25.60% less money today.

On the last business day before AMD announced it would buy ATI it was a company worth $9 billion, and one day before it was worth $10.65 billion, two times more than ATI. On the day it announced it was buying ATI, its market value dropped to $8.3 billion (values based on 492 million outstanding shares). And today, one year later, AMD is worth only $8.5 billion (value based on 550 million outstanding shares).

This is the math we can’t understand. AMD bought a 5 billion dollar company, paid 4.2 billion dollars in cash for it and instead of becoming a $15 billion company the market says that today the company is worth less than when it started the merger process. Once again, two business days before announcing it was going to acquire ATI, AMD was worth $10.65 billion, its value dropped to $9 billion on the day before the announcement and today it is worth $8.5 billion.

AMD stock performanceFigure 2: AMD stock performance.

But this story is even more intriguing when we analyze what happened to NVIDIA and Intel during the same time period.

[nextpage title=”Market Analysis: NVIDIA and Intel”]

So what happened to NVIDIA, ATI’s archenemy, during this same time period?

On the day AMD announced it was going to buy ATI, NVIDIA stock closed at $19.56, a 10% increase over the price of the previous business day. On the day after the last trading day of ATI stock, October 25th, 2006, NVIDIA stock closed at $32.80 – a 67.7% increase! And July 20th, 2007 NVIDIA stock closed at $45.10, an unbelievable 130% increase in just one year. Who bought $5,000 in AMD shares one year ago would have today $4,585, while if you had bought NVIDIA you would have $11,500!

How about the size of the company? On July 21st 2006 the market evaluated NVIDIA as worth $6.2 billion, on the day AMD announced they were going to buy ATI NVIDIA market cap increased to $6.9 billion (values based on 352.5 million outstanding stocks). Today the market says NVIDIA is worth $16.4 billion (based on 362.9 million outstanding stocks).

What!?!? Wait a moment! That was exactly what should have happened to AMD! Oh-oh. Backfire! So what happened is that even though AMD buying ATI “made sense”, the market thinks that the company that benefited the most from this merger was NVIDIA!

That is even more impressive if we remember that ATI was a $5 billion company when it was sold to AMD.

NVIDIA stock performanceFigure 3: NVIDIA stock performance.

And how about Intel? Well, Intel is a different story, because it is far bigger than AMD, as it designs and develops several other products besides PC processors.

Anyway, On July 21st 2006 Intel stock closed at $17.15, closing at $17.48 on the day AMD announced it was going to buy ATI. However, on July 19th, 2006 Intel stock closed at $18.49, dropping 7.25% in two days. On the last day of trade of ATI stock Intel closed at $21.62. And July 20th, 2007 Intel closed at $24.55.

If you bought Intel instead of AMD on the day AMD announced they would buy ATI, you would have accumulated a 40.45% profit in one year. That is amazing.

The market valuated Intel on July 21st 2006 as a $99 billion company. On the last day of ATI trading Intel was valuated as a $125 billion company. And today Intel has a market cap of $142 billion, being almost 17 times more valuable than AMD (Intel has 5.8 billion outstanding shares).

Intel stock performanceFigure 4: Intel stock performance.

[nextpage title=”Two Different Worlds”]

AMD and ATI have two different corporative cultures. ATI was a Canadian company founded by three Chinese immigrants. AMD is an American company founded by seven engineers that were already working at the semiconductor business.

We have a couple of constructive criticisms to make about ATI, and we hope AMD takes care of them if they really want to grow.

ATI had and so far still has the really bad habit of making paper launches, i.e., “launching” products that aren’t available on the market yet and won’t be for several weeks or even months, only to counter-strike products announced by NVIDIA, which, by the way, always make sure that their products are on the market on the same day of the official release.

The most critical case was with CrossFire technology. “Released” May 30th, 2005 it only came to market by the end of 2005. There are several other cases as well and the most recent one was with Radeon HD 2600 and HD 2400 series. During AMD’s technical sessions to present Radeon HD 2000 to the media in April, they said that Radeon HD 2600 and Radeon HD 2400 would be available in late June. All we got was a paper launch on June 28th, 2007 announcing the two families, but the products delayed at least two weeks to arrive at the market. This is something that AMD should definitely work on, right away.

The second main cultural difference we would like to point out is the lack of more in-depth technical information on ATI products written in clear language available to the public. On a press presentation held on September, 2005, ATI spokesperson said that from that date on all ATI would fully disclosure all the technical details (like block diagrams and in-depth explanations, for example) of their architecture, but this never hit the web, what contrasts with AMD philosophy. AMD and all other microprocessor manufacturers always made available their datasheets with in-depth information about their products. For us that work in a highly technical media, not having access to the technical information right away is really bad.

The most recent example we can give is with AMD 690G chipset (which is, in fact, an ATI product). There is no datasheet for it available on AMD’s website. The same thing happens for all GPUs.

On the other hand, if we think about another buyout AMD did in the past, the future of AMD + ATI is quite promising. In the early 1990’s AMD was struggling on the 5th generation microprocessor war. Intel had released Pentium on March, 1993 and AMD only released a counterpart, K5 (a.k.a. 5K86), on March, 1996 – 3 years later.

At this time AMD was struggling big time to catch up with Intel. Pentium-133 was released on June, 1995, while its competitor from AMD, K5-PR133 (which worked at 100 MHz), was released only on October, 1996, 18 months later. Not only that. At the same time AMD was launching K5-PR133, a 100 MHz part, Intel was launching the whole Pentium MMX line, including the 200 MHz version.

In 1995 AMD decided to buy NexGen, a company that had launched a 5th-generation x86 CPU called Nx586 and were developing a new CPU called Nx686. With this buyout AMD renamed Nx686 to K6 and finally could compete face-to-face with Intel. This also explains the success of K6 over K5.

But at that time AMD paid “only” $550 million for NexGen, not $4.2 billion, and also being a CPU designer, they were culturally closer. On the other hand, AMD stock on the day it announced it was going to buy NexGen closed at $13.06 and one year later it was quoted $8.69, 33.46% less.

Will ATI acquisition be as successful as NexGen’s? Only time will tell.

[nextpage title=”Conclusions”]

So far the big winner from the AMD-ATI deal is clearly NVIDIA. Today NVIDIA is worth two times and a half more than it was one year ago, while AMD is worth less than it was before buying ATI!

This is simply amazing, AMD bought a 5 billion dollar company and instead of becoming a 15 billion dollar company NVIDIA was the one that became such company!

Also during the last 1-year period Intel market value increased 43%.

Of course the market isn’t static and everything can change. It is probably too early to see any practical results from the ATI buyout by AMD, especially because when AMD bought ATI they claimed that the results from this buyout would be seen only in 2008.

Also, if we consider that one year after AMD announced that it would buy NexGen (the company that gave them K6, the CPU that put AMD back on track) their stock had dropped 33.46%, we wouldn’t worry about AMD stock’s 8.3% drop in one year.

We truly hope that AMD takes our constructive criticism into account and changes the “paper launch” policy ATI is using for some many years; this is a bad habit for everyone, especially for the company’s image, as people get the impression that the company is just a big mouth.

We should perform a new analysis one year from now to see what happened. Let’s wait and see.