You can read his reply here. Note however that my criticisms explicitly are directed at narrow banking more generally, most of all my own (previous) version of the idea, not at the specific version of Kotlikoff's proposal. There is one particular topic I did not deal with, and on it I will quote Kotlikoff reproducing my critique and responding to it:
TC: A lot of what current banks do would be replicated by non-bank commercial lenders and the risk of the banking sector would be transferred somewhere else.
LK: You missed the key point that all incorporated financial intermediaries have to operate as mutual fund companies. There are no “non-bank commercial lenders” unless they operate as proprietorships and partnerships and their owners have their houses and yachts on the line. The risk of the banking sector is reduced because we set it up to eliminate any chance of bank runs and gambling by the banks with the taxpayers’ chips. Recall, the mutual funds are 100 percent equity financed at all times and in all situations.
TC: Ideally, these non-bank lenders would engage in greater “maturity-matching,” but if banks will exploit the moral hazard problem won’t these lenders exploit it too?
LK: The only financial intermediaries who can operate under Limited Purpose Banking according to the current rules of the road are private banks with no limited liability. The lack of limited liability will eliminate the moral hazard problem.
I am not inclined to see unlimited liability as a practical alternative. How many businesses supply commercial credit? Trade credit? Credit by any other name? — namely contracts involving derivatives, annuities, insurance, repurchase agreements, etc., with intertemporal payments and embedded interest rates in the prices. Would they all have to give up limited liability? Or would we end up channeling more financial intermediation through indirect credit transactions, while maintaining limited liability? A version of this dilemma is experienced regularly by systems of equity-based Islamic banking..
Second, unlimited liability creates a pecuniary externality across shareholders. Who wants to be the remaining "fat cat" shareholder? Why should Bill Gates ever invest? Non-mutual fund banks will end up owned by thinly capitalized individuals or entities, thereby defeating the purpose of unlimited liability while at the same time raising transactions costs. Walter Bagehot made this point, see also Joseph Grundfest, here is Hansmann and Kraakman with a reply. Alex very ably surveys the main arguments in an MR post.
Unlimited liability is fine for small-scale, private banking, especially in the international sector where tax evasion is a motive and the banks aren't fully part of any standard regulatory network. It doesn't work to force it on such a large sector of the economy as most commercial credit and non-bank lending.
In sum, I do not believe that narrow banking proposals benefit from being bundled with unlimited liability for other lenders.