As already noted, the above criticism applies also to supply-side theories. To explain fully the choice of institutional mode, one needs failing markets and/or differentially efficient firms and/or oligopolistic interaction, or all these together. Transaction costs associated with internalizing (internally generated) specific or nonspecific assets (as the case may be) alongside differential firm efficiency in using (rather than selling) these resources-advantages (as the case may be), in the context of oligopolistic interaction, can explain the choice of institutional mode in specific cases by specific firms. All other factors that reduce unit costs (including labour ones) and/or increase value (added) (quality, differentiation, brand name) can also be of relevance. Alongside our account of endogenous growth advantages, the direction of expansion and the choice of location, we have most elements of the puzzle. The desire to have a genuine ‘all-weather company’ can add an extra, mostly demand-side, factor to explain internationalization. All these provide not a general theory, but the main elements of a framework and tools to be applied in specific cases in order to explain and/or predict specific strategies by specific firms.12