Saturday, July 12, 2008

Yet another stupid acquisition

One of the things that I teach my strategy undergraduates is that most acquisitions destroy value. Related diversifications realize less synergy than claimed, vertical integration locks a firm into a substandard supplier (or customer), any merger has culture, integration, and strategic defocusing problems.

Acquisitions make sense as an exit strategy for little companies: sometimes they even work, but either way they get to cash out. But, as I tell students, most big acquisitions are about the CEO moving up the rankings in the Fortune 500 (or increasing the base for a bonus based on revenues or net income).

There are always nominal reasons for the acquisition, but they are filtered and twisted to support what executives want to do. In the principal-agent problem, the principal (shareholders) gives the agent too much discretion and then shouldn’t be surprised when that delegation is abused.

This morning’s Exhibit 2,423 in the Stupid Acquisitions Hall of Fame comes from AMD. AMD admits that the $5.6 billion it paid 21 months ago for videochip maker ATI is now worth only $3.1 billion. Friday’s write-down of $880 million comes on top of a $1.6 billion write down in January. Despite punditry endorsing the deal two years ago, the company is struggling to service its $5 billion in debt and shares fell to a 16 year low. Value destruction doesn’t get any more clear or convincing than that.

The fundamental problem of acquiring public companies is that you have to pay more than the market price — so the claim is either you know better than the market (never true) or that you will realize synergies that increase the value of the acquired company (almost never true). So the choice is between buying overpriced good companies, or troubled companies not worth buying at any price. Acquiring a troubled company means you acquire their troubles — whether it’s exposure to an industry past its peak (AOL Time Warner, Viacom-Blockbuster) or a company with a justifiably lousy market position (Daimler Chrysler).

There's one other problem with megamergers — small numbers. Cisco makes more acquisition mistakes than any tech company in the world, but since they have more at-bats, they have more successes too. (Evidence suggests their batting average is much better than most). Big mergers are almost no upside and huge downside: think Sprint Nextel, or any of the recent Oracle acquisitions.

The poster child for diversification was once General Electric. Its extreme diversification has always meant a lack of financial transparency, which meant that investors were putting blind faith in management. That faith no longer appears justified: GE has missed projections two quarters in a row, and the stock is back where it was in May 2003 (and October 1998). Normally I don't like to inflict new (expensive) editions of the textbook on students, but I think we need one that drops the praise of GE’s diversification skills — which apparently only worked when management skills were scarce among its competitors.


Kevin M said...

The AMD/ATI merger is a bad example of what you're arguing. The problem is that it usually takes between 4-6 years for a processor to go from concept to the first finished product. You're not going to see any of the fruits of the merger (outside of consolidation in the HR and Accounting departments) for a couple of years yet.

The rationale behind the merger was to allow AMD to continue to compete with Intel by integrating GPU functionality into their CPU cores, and those products are still a couple of years out. Intel is also working on similar products, and their version is also a couple of years out. While AMD is definitely taking a financial hit today from the merger, without the merger they would essentially be out of business in 3-5 years. So is it a bad/failed merger? Ask again later.

Abby said...

Hi! I'm an editor for Seeking Alpha. Please contact me at your earliest convenience at Abby