Galtland Index

100% Reserve Banking

12 min read

100% Reserve Banking

Murray N. Rothbard treats banking as two separable activities. Lending out savers’ time-deposits is ordinary intermediation and unobjectionable. Issuing claims against deposited specie — bank notes or demand deposits — is warehouse-receipt business and must be backed 1:1. Issuing such claims beyond the specie on hand creates “pseudo warehouse receipts” against money that doesn’t exist, and is the operational definition of inflation in his system.

The process of issuing pseudo warehouse receipts or, more exactly, the process of issuing money beyond any increase in the stock of specie, may be called inflation.

— Rothbard, Man, Economy, and State, Ch. 12 §A

The property-rights argument

The Rothbardian case for 100% reserves is not primarily empirical. It runs through Nonaggression and Property Rights. When a depositor hands the bank specie and receives a demand claim, the bank holds the specie in trust. The depositor still owns it — that’s what “demand deposit” means. The bank may then either (a) lend out reserves that are owed on demand to the original depositor, or (b) issue receipts against specie that is not in the vault. Either move creates obligations the bank cannot simultaneously honor; both are pseudo warehouse receipts in Rothbard’s sense.

This is structurally identical to grain-warehouse fraud or coat-check fraud: the warehouse cannot legitimately issue more receipts than the goods on hand. That a fractional-reserve bank “usually” gets away with it — because most depositors don’t withdraw simultaneously — does not convert the underlying double-claim into a legitimate one. The fraud is in the issuance, not in the eventual default.

The monetary-theory argument

The companion case is monetary-economic. New bank-issued claims beyond reserves are fiduciary media in the Mises sense (see The Theory of Money and Credit and Ludwig von Mises). They are also the single operational lever that powers Austrian Business Cycle Theory. Forbid fiduciary media — which is what 100% reserves does — and there is no Austrian business cycle entering through the banking channel. The mechanism the regime eliminates runs in five steps.

1. The bank must drop the loan rate to place the new credit. Fiduciary media are an additional supply of credit, not an offering against accumulated savings. The market will not absorb that additional supply at the prevailing natural rate of interest — the rate the underlying time-preference structure establishes — because the demand for credit at that rate is already fully met by genuine savers. The only way the new credit gets borrowed is to price it below the natural rate. Mises is explicit:

If they demand less than the natural rate of interest - and they must do this if they wish to do any business at all with the new issue of fiduciary media; it must not be forgotten that they are offering an additional supply of credit to the market - then these requests will increase.

— Mises, The Theory of Money and Credit, Part III, Ch. XX

The same passage names the banking-channel power: a bank of issue’s interest-rate setting is the credit policy, and the demand for its credit fluctuates only with that policy. The lever is real, internal, and asymmetric — though Mises is also clear that a single bank acting alone is constrained by competing banks and depositor confidence; the strong system-wide rate-suppression claim depends on parallel issue across the banking system.

2. The gross market rate falsifies entrepreneurial calculation. Entrepreneurs, including the most conservative, take the observable loan rate as a fact and compute net present values, project lifetimes, and capital-budget decisions against it. When that rate is suppressed below the natural rate by fiduciary expansion, the calculation returns projects that look profitable but in fact rest on resources — savings — that do not exist at the implied size. The error is not forecasting; it is the rational output of computing against a falsified input. The full statement of the calculational point is in Mises on Credit Expansion.

3. The production structure lengthens. The suppressed rate makes higher-order capital goods — projects far from final consumption — disproportionately more attractive than they would be at the natural rate. The economy reallocates labor and intermediate goods toward longer-duration projects: construction, heavy machinery, raw-materials extraction, R&D, anything interest-rate-sensitive. Consumer goods (lower-order) draw down. The visible footprint is exactly the sectoral asymmetry Rothbard documents in America’s Great Depression and Rothbard on Fed-Induced Booms.

4. The boom must end. Time preferences are unchanged. Real savings are unchanged. The capital structure now in motion cannot be completed with the resources actually available. Either the bank keeps issuing fresh fiduciary media to extend the artificial rate (postponing the reckoning at the cost of a larger eventual correction — and, in the limit, a flight from the currency that destroys it), or it stops and the natural rate reasserts itself, at which point the longest-duration projects are revealed as unprofitable. They are liquidated. That liquidation is the bust.

5. 100% reserves removes step 1, which removes the rest. Under 100% reserves, banks cannot offer credit in addition to what their customers have saved. Lending is restricted to time-deposit funds — what Mises calls commodity-credit and distinguishes from circulation-credit in Credit and Deferred Payment. Commodity-credit reflects an actual transfer of present-good claims from savers to borrowers; the loan rate clears the market between the actual savers and the actual borrowers, with no fiduciary-media wedge between the loan rate and the underlying time-preference structure. There is no false signal to lengthen the capital structure beyond what voluntary savings can sustain, no boom built on resources that do not exist, and therefore no bust required to correct it.

This is the operational sense in which 100% reserves is the Austrian-sound-money position on banking. It is not “tight money” in the Keynesian sense — it permits any amount of intermediation backed by genuine savings, at whatever interest rate the time-preference market clears. It is a structural prohibition on the single monetary lever the Austrian cycle theory identifies as load-bearing. Rothbard reinforces the link as a definitional matter:

“Inflation” is here defined as an increase in the money supply not consisting of an increase in the money metal.

— Rothbard, America’s Great Depression, Ch. 1 (footnote in the cycle-theory chapter)

A 100%-reserve regime makes this kind of inflation impossible by construction: every additional money-substitute corresponds to additional specie. Compare with the Mises framing in Cash Holding and the Demand for Money: Mises’s strict definition tests whether the broader money supply outran demand, while Rothbard’s tests whether new claims have specie cover. In a 100%-reserve world the two definitions agree on every banking-issued event (no fiduciary media → no event for either test to flag). They can still diverge on non-banking events: a fresh gold strike that swells specie holdings is not inflation by Rothbard’s definition (the money-metal grew) but is inflation by Mises’s unless demand for money happened to rise in step. In a fractional-reserve world Rothbard’s definition flags more events than Mises’s does, because Rothbard’s offset is specie cover while Mises’s offset is money demand.

What it is not

100% reserve banking is not the same as a deposit-as-safe-deposit-box arrangement where the depositor pays for storage. The depositor can still write checks against the deposit; the bank can still clear payments on instruction. What it forbids is the bank lending the reserve to a third party while simultaneously promising the original depositor immediate availability.

It is also not the same as narrow banking, sovereign-money proposals, or central-bank digital currency (CBDC). Those are state-architecture interventions; 100% reserves is a contractual / property-rights condition that, on the Rothbardian view, ought to bind any institution calling itself a bank, with or without a state.

Q&A

How do you measure credit expansion? Is M2 a good proxy? Or central-bank rates?

The Austrian framework does not have a single observable that maps cleanly to “credit expansion” in the Misesian sense. The theoretical variable is the gap between the gross market rate of interest and the natural rate; neither side of that gap is directly observable. Measurement is necessarily indirect, and different proxies catch different parts of the phenomenon.

  • M2 is the standard broad-money aggregate, widely available, but theoretically muddled for Austrian use. It lumps demand deposits (fiduciary media — what the cycle theory targets per Mises on Credit Expansion) together with time deposits, which are commodity-credit (savings-backed lending) and not inflationary in the Misesian sense — see Credit and Deferred Payment. M2 will tell you something about broad-money trends but it overstates Austrian “credit expansion” by counting commodity-credit alongside fiduciary media.
  • Central-bank policy rates (Fed funds target, ECB deposit rate) are an input, not a measure. They are what the central bank wants the loan rate to be; the resulting expansion of fiduciary media depends on how banks and borrowers respond. Low policy rates predict more expansion but don’t quantify it.
  • Rothbard’s True Money Supply (TMS), refined by Joseph Salerno, is the theoretically clean Austrian aggregate. Full treatment in Monetary Aggregates and Credit Expansion; short version: currency in circulation + demand deposits + government deposits + savings deposits at commercial banks, excluding money-market funds and retail repos. Targets fiduciary media specifically. This is what the framework actually wants you to look at; M2 is a noisy substitute.
  • Sectoral asymmetry is the diagnostic Rothbard documents in America’s Great Depression and Rothbard on Fed-Induced Booms: interest-rate-sensitive sectors (construction, capital goods, long-duration projects) expanding disproportionately is the structural footprint of suppressed rates. A rising sectoral skew is a more theoretically meaningful signal than a single aggregate number.

Practical synthesis: M2 plus the policy-rate path gives a rough first-pass picture, but Austrian rigour calls for TMS or demand-deposit growth specifically, combined with the sectoral signal. The Q “is M2 a good measure” gets a qualified no: it’s a usable proxy with known theoretical drift.

If all debt is paid, every expansion meets a contraction. So in the long run, no net expansion?

Mises addresses this directly in The Theory of Money and Credit, Part III:

When the loans granted by the bank through the issue of fiduciary media fall due for repayment, then it is true that a corresponding sum of fiduciary media returns to the bank, and the quantity in circulation is diminished. But fresh loans are issued by the bank at the same time and new fiduciary media flow into circulation.

— Mises, The Theory of Money and Credit, Part III, Ch. XX

The answer is layered:

  1. Per-loan view: yes, repayment cancels issuance. Each loan that gets repaid extinguishes its specific fiduciary-media balance. The mechanism is symmetric on a per-loan basis.
  2. Aggregate view: depends on bank policy, not arithmetic. Banks roll over: gross new issuance happens at the same time as gross repayment. The aggregate stock of fiduciary media stays flat, grows, or shrinks purely as a function of the bank’s credit policy choice. Empirically banks have grown the stock; the asymmetry is policy, not law.
  3. The cycle theory operates on flow, not stock. Even if the aggregate stock is flat year-over-year, this quarter’s marginal lending still suppresses the loan rate. The malinvestment mechanism cares about active fiduciary-media issuance, not the integral over time. A bank rolling at constant stock is still issuing fiduciary media every quarter and still suppressing rates.
  4. Halting issuance is the bust mechanism. If banks ever stop issuing fresh fiduciary media, the natural rate reasserts itself and the longest-duration malinvestments are revealed as unprofitable. That is the bust. So “long-run no net expansion” is not a steady-state equilibrium — it is the trigger for the correction. The bank’s real long-run options are: (a) keep issuing, in which case the stock grows and in the limit the currency dies in a flight from money; or (b) stop, in which case the cycle’s accumulated malinvestments liquidate.
  5. Even a fully-repaid one-shot loan still misallocates. During the loan’s life, capital was reallocated toward higher-order projects on a falsified rate signal. When the loan is repaid, the monetary event is undone but the real misallocation isn’t. Longer-duration projects funded by the loan may still be in motion when the loan matures; their funding channel evaporates while their resource commitments remain. The repayment cancels the fiduciary media; it does not unwind the structural distortion.

So the elegant short answer is: yes locally, no globally, and the global “yes” is the bust. Repayment is not a free unwinding; it is the moment the discipline returns, and the cycle theory predicts that moment is painful in proportion to how long the expansion ran before it stopped.

See Also

Sources

Graph View

Backlinks