Regular readers of this site may be aware of a debate relating to the contractual devices that banks might use to ensure that they are solvent. One of the terms that has been used is a “notice of withdrawal clause”, but what is this?
It might be argued that a notice of withdrawal clause (or a “withdrawal clause”) is merely another term for the more often invoked “option clause”. This has received extensive treatment in the “free banking” literature (for example Dowd (1988), Selgin & White (1994, 1997)), and we can use the following definition:
option clauses… give banks the option of deferring redemption of their notes provided that they later pay compensation to the noteholders whose demands for redemption are deferred” (Dowd, 1998, p.319)
The confusion may stem from the fact that in some instances option clauses and withdrawal clauses are used interchangeably. For example Selgin & White (1997) say:
one possible run-proofing device discussed in the literature is an “option clause” or “notice of withdrawal clause” allowing a bank temporarily to suspend the redeemability of some or all of its liabilities (notes or demand deposits) provided the bank pays a pre-specified (penalty) rate of interest on the suspended liabilities
However, I believe there is stronger textual support for the idea that they are distinct devices. In an earlier article Selgin & White (1994) say the following:
a bank might contractually reserve the option to suspend for a limited time the redeemability of its notes or demand deposits, as Scottish banks did with banknotes before 1765 (when the practice was outlawed) and as banks do today when they include “notices of withdrawal” clauses in deposit contracts
My reading is that they are often used interchangeably (or perhaps as though a withdrawal clause is a type of option clause), because they perform the same economic function. But a detailed reading would reveal them to be different.
In my working paper on the sound money debate I define a withdrawal clause as follows:
In addition to the option clause banks might also offer (and historically did offer) a “notice of withdrawal” clause, specifying that their customers were required to give 30 days notice prior to making a redemption claim. The fact that this clause existed (to protect the bank from a legal point of view if it were ever to suffer a liquidity crisis) does not mean it is always invoked, and banks could routinely not enforce this rule and satisfy immediate redemption requests.
Firstly, note that this is presented as a different clause to the option clause. But secondly, we can see that it differs from the option clause in terms of the default nature of the contract.
Recollect that an option clause allows banks – under certain conditions – to convert a demand deposit into a timed deposit (thus giving them time to generate liquidity whilst avoiding firesale losses). This is seen to be good for the banks (obviously!) but also good for the customers (since it’s better to receive the deposit plus interest at some point in the future than to see the bank being wiped out).
However in the case of the withdrawal clause there is a notice period written into the contract – it is technically a timed deposit (where the notice period serves as a minimum term). But if the bank wanted to offer an instant access account it can simply publicise the fact that it does not routinely enforce this notice period and that it satisfies redemptions on demand.
I suspect the reason withdrawal clauses received less explicit attention in the literature is that unlike the option clause they are not a uniquely “free banking” concept. Indeed, notices of withdrawal are routinely used in contemporary banking. Investopedia define it as follows:
A notice given to a bank by a depositor. As its name implies, a notice of withdrawal states the depositor’s intention to withdraw funds from an account. This notice applies to both time-deposit and NOW accounts
In short, the option clause means that a de jure demand deposit can be treated as a de facto timed deposit. The withdrawal clause means that a de jure timed deposit can be treated as a de facto demand deposit. They are two sides of the same coin – both allow instant access fractional reserve accounts, the only difference is the default position.
So perhaps provisions such as option clauses and withdrawal clauses allow banks to offer fractional reserve accounts that aren’t fraudulent or reliant on legal privilege, but does that make the 100% reserve argument wrong? Not necessarily. The withdrawal clause in particular “works” precisely because it changes the de jure status of the account. A counter argument might be “if a withdrawal clause applies to a timed deposit then you are admitting that fractional reserve banking is irreconcilable with demand deposits”. From the view of legal theory (and depending on your definitions), this may well be correct. However the de facto status of this account is instant access and redeemable “on demand”. In terms of the economic function of the account it exists exactly as “free bankers” envisage.
- Dowd (1988) “Option clauses and the stability of a laissez faire monetary system” Journal of Financial Services Research, 1:319-333
- Selgin & White (1994) “How would the invisible hand handle money?” Journal of Economic Literature, XXXII:1718-1749
- Selgin & White (1997) “The Option Clause in Scottish Banking” Journal of Money, Credit and Banking, 29(2):270-273