Economies of Scale in Autos?

For a long time, the conventional wisdom in the auto industry has been that bigger is better. In other words, consolidation occurred as firms sought to achieve economies of scale. We saw huge mergers such as Daimler & Chrysler, as well as acquisitions such as Ford's purchase of Volvo and GM's acquisition of Saab.

Now, we see a reverse in this trend. As GM reorganizes, it is selling off many business units. However, these units are not being purchased by large automakers. Just today, we hear that Koenigsegg Automotive AB, a quite small Swedish super luxury carmaker, has acquired Saab. Magna, a Canadian auto parts supplier, acquired GM's European subsidiary, Opel, last week. With these deals, we see firms operating at what used to be considered suboptimal levels of production. In fact, Fiat bid for Opel in part because their CEO, Sergio Marchionne, believes that his firm needs to get to roughly 6 million cars produced in order to fully capitalize on economies of scale. Below that number, he does not believe that he will have fully exploited scale economies.

Who is correct? Are these smaller firms making sensible moves by choosing to compete at levels of production far below 6 million autos, or is Marchionne correct that one has to achieve that level of production to be cost competitive in the mass market? I tend to believe that some powerful scale economies do exist in autos, but that consolidation over the past two decades went too far. Companies failed to account sufficiently for the possibility of diseconomies of scale, and they didn't fully understand the challenges of cross-border merger integration.