When is an alliance too big? Here’s a possible example. Some members seem to be doing well, but others are losing out. And the deal lasts only till 2016.
Coalitions are always tricky because they lose their advantage when some members drop out. On average the deal might make sense. But life is stochastic– random fluctuations can cause some of the members to suffer while others prosper. And these specific realizations of events may cause participants to wonder if it’s to their advantage to continue. Add a short contract time, and where’s the motivation to stay in the coalition?
It is so hard to form these coalitions that they need to be thought through carefully beforehand, and attention paid to how the random fluctuations will be handled.
Moshe Dror and I have a series of papers dealing with such problems in the inventory pooling context, but cooperative games are equally applicable in any pooling of resources.