Corporate strategist agonistes

Russ makes an interesting point about SBU portfolios and headquarters’ impact on their performance, but from Rumelt on, variance decompositions cover multiple years (that’s how the business unit effect is identified). Business unit effects are therefore persistent phenomena. I don’t see how a transient improvement caused by HQ could be confounded with SBU effects as suggested. It could move some of the variance from the error term to the corporate account, I think.

Possibly one could directly look at corporate acquisitions and divestitures to see if particular parents differ in their impact on the time pattern of SBU ROIs. What would worry me about interpreting such a study would be distinguishing between an HQ that actually improved its acquisitions and an HQ that was just good at buying low and selling high.

It does seem that far too much corporate strategy research is done in a decontextualized large-sample way, so that it’s impossible to be sure whether findings are meaningful given the aggregation bias. Lumping together HQs pursuing very different approaches to adding value reminds me of the old Saturday Night Live commercial for the Bass-o-Matic–all the structure of the fish gets destroyed in the statistical blender. So we might be missing some important things. I’m attracted to disaggregative ideas like the Gould and Campbell taxonomy of HQ rationales, which at least try to make sense out of what the corporate strategy actually is.

My gut sense, however, is that most–though not all–actual American corporate strategy is a waste of time. Corporate constraints on divisional behavior are usually either pointless or counterproductive. The whole reason for the SBU structure is to make each market-competing unit accountable for its own results, with the requisite autonomy to adapt to its particular environment. That militates against exploiting most potential synergies and linkages across businesses. And when we do have businesses that are tightly linked, so that the SBU structure is inappropriate, HQ strategies that try to separate them to “unlock value” are likely to fail as well.

In other words, managing tight synergies isn’t really a separate realm of “corporate strategy” but merely business strategy with scope economies; managing portfolios that lack synergies, on the other hand, doesn’t add value to those portfolios.

Corporate strategy has to play on a thin middle ground of loose synergies. You may find some cross-SBU internal labor market gains, common control system gains, and possibly common culture gains IF you pick a portfolio of businesses whose industry contexts and positioning strategies are similar to one another AND you’re especially good at managing them. I find it hard to believe, though, that these types of effects are large enough to make a big difference most of the time. I would be surprised if better, disaggregated, analyses of corporate strategy, disproved this conjecture. I would enjoy such a surprise.


4 Comments on “Corporate strategist agonistes”

  1. RussCoff says:

    I still think that if firms cash out of strong performers, that would make the corporate effect look smaller — those units are then removed from the sample.

    Actually, I do share your skepticism for the prospects of creating value with corporate strategy. However, given the proliferation of multi-business firms, my gut tells me that even the destruction of value might explain more than 4% of the variance (overpaying in M&A and search of illusive synergies eventually add up to real dollars).

    If not, perhaps I am allocating too much time to teaching corporate and M&A… a distinct possibility but it does seem to correlate nicely with what’s in the news.

  2. RussCoff says:

    Actually, your comment stimulated further search on my part. Adner and Helfat have a nice paper identifying the problem of measuring a stable or average corporate effect across a long period (see http://faculty.insead.edu/adner/research/dynmgrlcapscorpeffectSMJfinal.pdf). In other words, variance decomposition models have not generally measured time varying corporate effects.

    This is important because some corporate decisions occur sporadically but have dramatic effects on business units. They add data on one type of decision, corporate downsizing, and find that they can explain an additional 12% of variance in performance (even when entered last into the model). This is over and above the 7.5% average corporate effect they found.

  3. stevepostrel says:

    That’s an interesting paper. It focuses entirely on large firms in the fossil-fuel energy sector, so the kind of broad corporate strategy issues involved in a GE, Tyco, Berkshire Hathaway, or Hyundai are not discussed. In addition, it uses corporate downsizing decisions as a factor in the variance decomposition and looks at how much of the variance in business-unit earnings is accounted for by these decisions as they occur.

    I don’t think that sort of analysis is valid for the causal purposes they put it to–the downsizing decisions seem triggered by business-unit performances across corporate portfolios, rather than the other way around. Corporations that downsize differently are likely to have business units with different performance levels, both prospectively and retrospectively. That isn’t much of a vindication of corporate strategy as a vocation.

  4. […] Steve Postrel wrote a response to Russ’ post, “corporate strategist agonistes.” […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s