What is liquidity anyhow?

For geeks (and bitcoin fans):

I got a referee report on my liquidity paper earlier this week that said (among other nasty things)

“The definition of liquidity is so vague as to be completely useless. In a related vein, the paper seems to use tradability and liquidity as synonyms which is either a grave mistake or an expositional shortcoming.”

My definition was

“I define a financial claim to be liquid if: 1) there are no legal or practical restrictions on who the claim can be sold to, on when (during normal trading periods) sales can be made, or on the amounts that can be sold (partial sales allowed); and 2) during normal trading times and conditions, market makers quote firm bid and asked prices that are not contingent on any further investigation of the claims or of the identity of the seller or buyer of the claim.”

This was slightly sloppy. I should have said “I define the market for a financial claim” etc.

More importantly, it seems to me that people use the term liquidity in two ways. First there is the liquidity of “money: currency notes, bank deposits, or even post-dated checks or havala receipts. Here the holding is not expected to appreciate — it is enough if it doesnt depreciate. And here tradability indeed has nothing to do with liquidity.

My paper focuses on the liquidity of claims that are held with the hope of appreciation and some tolerance for loss. This kind of liquidity is valuable primarily for its contribution to reducing the costs of diversification and secondarily for enabling investors to sell assets to meet unexpected claims. The key word here is unexpected — a sensible investor would not keep funds for predictable impending payments in volatile assets even if they are highly liquid (eg keep cash needed for payroll in stocks). The matching of longer term assets and liabilities is obviously trickier

The problem arises when people rely on the second kind of liquidity for foreseeable and routine cash needs, as Harvard reputedly did before 2008 when it invested its operating cash in the Harvard endowment because the endowment had (until then) earned double digit returns like clockwork.

Bitcoin seems to have a similar problem: its premise is that it will be used as “money” whereas its “liquidity” is entirely a function of its tradablity.