More on Negative Rates
Here's what I think we can agree on:
1. The zero lower bound is a bound at zero only if there is no cost to storing currency. Since this is not true, as there is some cost to storing currency, the bound is not at zero.
2. Furthermore, it may not particularly useful to think of it as a bound. Rather, different agents in the economy face different costs of switching from deposits to currency. This distribution of costs shapes a supply curve for deposits. As the interest rate is increasingly negative, the supply of deposits dwindles, as more and more find it worthwhile to convert deposits into currency.
3. Krugman is right that the marginal depositor must use deposits only as a means of storage, not as a transaction medium. That is, the first dollars to get converted into currency must be ones that nobody plans on spending anytime soon -- the dollars that would otherwise pass in and out of credit-card accounts remain in deposit form unless rates continue to go negative.
4. That argument, however, also suggests that the value of the convenience that I stressed in my previous piece is not irrelevant for the entire deposit-supply curve, but rather just for the initial segment. The marginal benefit of deposits includes convenience in transactions for a later section of the deposit supply curve.
5. It is not quite clear how large the share of passive deposits are -- i.e. the ones that exist just for storage -- relative to the share of deposits that are, in some sense, money in movement. Excess reserves suggest it could be substantial.
6. Furthermore, as Miles Kimball stressed to me, whether the movement into negative nominal interest rates is expected to be enduring or short-lived matters, because setting up systems to store currency likely involve fixed costs. So it's likely the short-term supply of deposits in negative-rate territory would be greater than the long-term supply.
7. There are many, many ways to maneuver around a negative nominal interest rate. As I suggested at the end of the earlier piece, inventories and working capital may be more important than our currency-focused discussion would suggest.