How can banks be "right-sized" to avoid inefficient small banks, yet not create banks that are "too big to fail"?
I'm reading "Fragile by Design" by Calomiris and Haber. I'm only partway through the book, and in the early chapters they argue that small unit banks create economic systems that are under-banked and slow economic development. They say that unit banks cannot effectively spread risk over geographic areas and business sectors. They cannot achieve scales of economy. In short, they are deeply problematic.
On the other hand, in the 2008 global financial crisis, one thing I heard quite a bit was that bank bailouts were necessary because the banks were too big to fail. Economists and public policy people seemed to widely agree that Bank of America, JP Morgan, and Goldman Sachs were all too big to fail, and that the damage of the 2008 crisis would have been even worse if bank bailouts hadn't occurred.
I imagine that the answer is that banks should be "somewhere in-between" but what does that mean in a quantitative way? What is the "right size" for a bank? Can that be quantitatively calculated or in some other way well-defined?
If it matters, I'm a physics person so I can handle (probably would even prefer) a mathematically rigorous model.