Fractional Reserve Banking: A Response To Mish

On Tuesday I wrote an article, “Fractional reserve Banking: The Real Story,” in which I discussed the workings of fractional reserve banking. Mish (Mike Shedlock) responded critically to my article on his blog, “Central Bank Authorized Fraud; Fractional Reserve Lending Problems Go Far Beyond “Duration Mismatch“. This post is a response to his comments. I think there can be no more important topic than the discussion of a proper gold standard, and integral to this topic is the question of whether fractional reserve banking is perfectly fine, or whether it is by nature fraudulent.

Let me first address a few things that I could have made clearer in my paper.  First, I consider myself a student of the Austrian School.  My disagreement is with the (modern day) notion that fractional reserve banking is intrinsically fraudulent, which I don’t think is part of the Austrian School proper.

Second, when I said I was not aware that others had written about duration mismatch, I should have stated clearly: duration mismatch as the root (or one of the main roots) of systemic instability.  I read your blog every day, and I have seen you discuss the problem but to my recollection you have addressed this as a problem for the lender, which I agree it is.  But I think we both agree that it is also a problem for the financial system and the currency itself.  (By the way, I looked at your 17 links search results page, which has now proliferated to 5 pages, and most of the links are to unrelated posts that have a link to your present post in the sidebar.)

Third, the lending of money that is supposed to be available on demand is in fact a case of duration mismatch (the duration of the liability is zero and the duration of the asset is whatever the term of the loan is).

Fourth, I certainly agree with you that there are other kinds of illegitimate lending whereby the bank lends money it does not have the right to lend for reasons other than duration.  In my paper, I did not try to address this.

Fifth, the central bank allows (and encourages and incentivizes and in some cases forces) banks to do all sorts of illegitimate things.  Again, I was not trying to address this.  I deliberately choose to look at a gold standard without a central bank (which I could have made more explicit).  My goal was to look at the essence of fractional reserve banking as such, vs. the present system of irredeemable paper money + a central bank + FDIC moral hazard insurance + GSE’s + social-engineering-masquerading-as-tax-policy + hundreds of thousands of pages of regulation + deliberately inflationary monetary policy + massive deficits + “open market operations” + US Government bonds crowding out other investments + Fed-authorized “sweeps” + too many other things to mention.  I agree 100% that those things (and many others) are illegitimate, coercive, fraudulent, and contribute to systemic risk.

The scope of our disagreement might be as narrow as your statement “…I cannot lend you $1,000 if I only have $1.98.  However, and this may surprise many people, banks can.”

I think I showed how fractional reserve banking (stripped of central banks and other accidental stuff), in its essence, does not provide a way for banks to lend more money than they have.  Of course, in today’s irredeemable debt-backed currency system, there is no distinction between money and credit.  In my examples, the difference between a gold coin vs. a claim to a gold coin to be delivered in the future is black and white.  In a free market with a proper gold standard, banks can multiply *credit* but not *money*, and if they are prohibited from borrowing short to lend long, then they cannot fraudulently extend the duration of a loan beyond the specified preference of the actual owner of the gold coin: the depositor.

I want to emphasize one last–but vitally important–point.  There is no such thing as lending without fractional reserves.  If a bank takes in a deposit and lends the smallest portion of it, that is a fractional reserve (i.e. less than 100%).  A so-called 100% gold-backed deposit means that the bank is relegated to the roles only of payment transfers plus safe depository.  Clearly both of these are legitimate services and would be available in the free market.  Savings as well as hoarding (storing gold without being willing to lend it) both play crucial roles in a proper gold standard, and there are important arbitrages between them.  But this is outside the scope of my paper.

I look forward to Mish’s response.


24 comments to Fractional Reserve Banking: A Response To Mish

  • Keith: FWIW I agree that fractional reserve banking does not make the lending bank insolvent per se. However . . .

    Completely separate from the issue of temporal mismatch of assets/liabilities, I always thought the main issue was not that fractional reserve banking was inherently improper/fraudulent from an individual bank standpoint, and as per your proof that it does not involve insolvency per se in your initial post. I thought that the most bothersome issue was the instability associated with the eventual creation of (say) $900 in credit (money) beginning with $100 in base money, with successive banks lending out (say) 90% of each deposit.

    Even if all assets and liabilities were properly matched from a time maturity standpoint, my concern over the way fractional reserve banking is practiced today is one I’m not sure you addressed. It is that the “system” becomes unstable if even one bank lends badly (fraudelently or just imprudently or even just unluckily) and fails, imperiling depositor confidence in general and leading to runs as well as cross-bank impairment of assets. Since banks nowadays lend out almost all their deposits, your extreme example of a bank that lends out 0.2% of its assets is not the issue.

    The daisy chain of highly leveraged financial institutions creates much more inherent risk than naive depositors tend to believe is present.

    Do you have any comments on that topic?

    • Keith Weiner

      You are correct, I did not want to address the present system directly. I wanted to address the concept of fractional reserve banking per se, separated from the present system (which is so illegitimate in so many ways it’s hard to know where to even start dissecting it!) For a number of reasons, the system we have today isn’t going to last too much longer. So I’d like to confine my comments to a proper gold standard.

      I think you are right that financial institutions are and would be highly interconnected. If one fails, the others are exposed. What I want to focus on at the moment is the difference in quality of credit expansion if durations are matched. Depositors bring in their gold coin and, wanting to earn interest, agree to time deposits of varying durations. The banks have the right to lend this money for those (or shorter) durations.

      This system has a different kind of quality to it. Because the durations are matched, the bank is not exposed to “rollover” risk: what if the depositors demand their money? Depositors can come to the bank and ask the bank what the cost would be to sell the loans and recover their money. The bank would look at the bid on such loans, and come back to the customer with an answer (of course keeping a small fee for this service). The customer can decide whether the loss is worthwhile to him or not. Presumably, the collapsing bid on loan assets would serve as a check to curb depositors’ enthusiasm for early withdrawal. If not, it does not impact the solvency of the banks but of the depositors.

      So the bank retains credit risk: what if the borrower cannot repay? But not interest rate and duration risks. So the bank could still become insolvent as a result of poor credit judgement. This is where a prudent bank would keep reserves. I would assume that it is possible to determine a conservative reserve ratio to cover credit risk, if one is an expert underwriter (I certainly am not!)

      An interesting point (which I addressed in reply to a comment on my earlier post) is that there is no particular reserve ratio that can cover duration mismatch risk. As credit expands to whatever arbitrary level, it will eventually become larger than any bank’s reserves. Although a bank with larger reserve ratios would survive a level credit contraction that would wipe out its more aggressive peers, every bank that plays this game can and eventually would fail.

      I think there are a few additional factors that would add to stability in a gold standard. One is that gold is the extinguisher of debt. Debt would not tend to accumulate until a catastrophic systemwide failure (as in our system). Another is that the marginal saver can withdraw his gold coin if he thinks interest rates are too low, or that credit is being extended to poor credit risks. This will tend to force interest rates higher (in our present system an individual can of course sell dollars to buy gold, but the seller of that gold buys his dollars and deposits them back into the banking system). In engineering parlance, we would call the present banking and financial system a closed loop. In gold, it would be an open loop whereby gold can enter and leave the system. On the ceiling of interest rates, the marginal entrepreneur will liquidate his capital and buy bonds if the interest rate is higher than what he is earning in his business.

      Does this answer your question?

  • Inquisitor

    I pretty much second DoctoRX’s comment.

  • Sundar


    I think one of the important questions that has to be understood and answered properly in the case of fractional reserve banking is this:

    Who determines the fraction of reserves kept on demand and the remaining lent out as time deposits?

    The answer is: It has to be the individual who deposits the money in the bank. It has to be a purely subjective decision and every individual has the right to choose the fraction he wishes to lend and the bank should honor such a decision. The important thing to note here is that it is purely subjective and it can vary quite widely. In this case, the leverage in the financial system exists only as much as the combined risk appetite of all the depositors in that bank.

    The Mises School characterizes that fractional reserve banking process itself is fraudulent. As it exists today, I think we all agree that it is fraudulent. I don’t think the problem lies in fractional reserve lending. The real reason is that the individual’s subjective risk appetite has been usurped and completely taken over by the monetary authorities.

    Would be glad to hear other thoughts on my observation.

    • Keith Weiner

      Sundar, I agree. You make an excellent point: there is no one intrinsically right ratio of demand to time deposits. I don’t use the word “subjective” (a debate for another day please!) but it is certainly personal.

      Getting back to the question posted by DoctoRX, this is another stabilizing factor. If, instead of having Sheila Bair or Tim Geithner or Ben Bernake (I don’t know which has control over this variable, and it doesn’t matter) dictate that all banks shall hold 10% of deposits on demand (let alone “sweep” them every night)… if instead of this, each individual specified it, then the risk of a bank holding insufficient money on demand is eliminated.

      • Sundar


        No debating :), but just wanted to state the reason why I use the word ‘subjective’ is really attributed to Carl Menger.

        He said ‘Value does not exist outside the consciousness of men’.

        I believe that this is the foundation of all human action.

        Value as a concept is ‘subject’ to varied interpretation depending on who is actually doing it.

  • Keith Weiner

    May I email you to take this offline (I get your email address in my notifications as an author of this blog entry)?

  • Sundar

    Sure, Keith.

  • Stephen

    “There is no such thing as lending without fractional reserves.”


    Lending could very well be organized as a business that sells shares. The business issues shares and uses the proceeds to lend. As the performance of the business improves, share values rise, as it falls, share values fall. You wouldn’t have a traditional collapse due to a run on the bank…there might be a revelation that the load business is not performing well and people would sell of shares, but eventually the price of the shares would get to a point were it accurately reflected the true performance of the load portfolio. People would own shares instead of demand deposits and they would be under no illusions regarding the potential for the value of those shares to both rise and fall in tandem with the performance of the loan portfolio.

    • Keith Weiner

      That is another way of allowing personal preference to establish the ratio of demand to time deposits. Instead of each person specifying to the bank how to apportion his deposit into time vs. demand, the person could bring some of his money to a safe depository and the rest of his money to buy the equity of a joint stock loan company.

      I think it would be a cumbersome setup.

      The results would be:
      – probably used only by wealthier and more sophisticated investors
      – relegate the majority of people to paying for safe depository services
      – much wider bid-ask spreads on bonds (i.e. interest rates)
      – vastly higher costs of credit and less availability of credit to worthy business projects
      – higher interest rate volatility, and thus destruction of capital

      • Stephen

        These are all just justifications for the fraud. You may be right (higher interest rates, higher lending costs, etc), but that doesn’t change the fact that fractional reserve banking is simply a fraud that tricks people into lending under false pretenses (not to mention the fact that, as you point out, is pretty much guaranteed to fail). In any case, I think you’re wrong. The free market will quickly learn that they are far better off allocating a portion of their savings to lending/investment than letting it sit in 100% backed demand deposits. Your points are conjecture…what’s needed is empirical evidence and controlled experimentation.

        • Keith Weiner

          I don’t see how you can use the words “fraud” or “false pretenses” after reading my paper.

    • What you are describing is a mutual fund. The value of your deposit would be relatively unstable and thus people would likely rather not own such shares. Without a banking intermediary, people would hold cash in their homes instead of lending it out, to have some stability.

      • Stephen

        Yes, very much like a mutual fund. The dollar itself is very unstable. You are merely speculating without empirical evidence that people wouldn’t want to own such shares. People would gravitate toweard those funds that are more stable and offer decent returns. But fractional reserve banking is a quasi-fraud is fails on any sufficiently long time horizon.

  • Thanks, Keith. Nice to see such an informed discussion.

  • Linus Huber

    I love to follow these kind of discussions. I have a slight different aspect to address on this subject. For me the aspect of fraudulency arises from the aspect of loans being issued for extensive periods of time. In my opinion, there should not be loans allowed in excess of maybe 7 years, after which it will automatically become nil and void. This will allow the system to continuously clean itself therefore avoiding a build up respective ponzi scheme in the area of credit. For Companies and Countries this period may be doubled or something, but no roll over should be allow.

    I am interested for any comment as I love to learn and understand.

    • Keith Weiner

      Linus: you touch on a topic for a whole ‘nother paper, which is why debt accumulates in our regime of irredeemable paper money. It is a ratchet that only goes up (until it crashes catastrophically). In gold, this is not true–and one does not need a jubilee every 7 years. There are many reasons, but the most important is that gold is the ultimate extinguisher of debt. In a gold standard, if you pay a creditor in gold, then the debt is eliminated. Today, if you pay a creditor in Federal Reserve Notes, or better yet by banking credit, you merely transfer the debt to another party.

      One party’s debt is another party’s money, today. The lack of extinguisher of debt is not a bug. It’s a feature.

  • Keith,

    I’m a little unclear on your use of the word duration. Are you talking about maturity per se, or duration per se insofar as duration (e.g. Macaulay) is a measure of interest rate risk?

    Second, in support of the hypothesis that fractional reserve banking is not inherently stable: the Scottish free banking episode was characterized by exceptionally low reserve ratios (2%, gold backing) but this was a very stable period.

    On the other hand, HK banks had a 100% reserve requirement, when they had free banking until 1965.

    The question is to what does reserve apply? And this is not fundamentally clear from the Misesian standpoint. You can have 100% reserves backing notes, but that wouldn’t reduce the ability to have fractional reserves against deposits. It’s a very interesting and convoluted issue, and every attempt to deal with it is valuable. Very nice post.

    • Keith Weiner

      I am talking about maturity. If a bank borrows via taking deposits into a 1-month CD but extends a loan for 1 year. The bank must “roll over” its funding, i.e. convince those depositors to to re-enter 11 more 1-month deposit agreements. If the bank cannot do that, it has a very big problem.

      I think I wrote this in a reply to a comment on another thread, but if a bank mismatches the durations, I think there is no particular reserve ratio that they can guarantee will protect them against a run on the bank. Over time in FRB (especially in irredeemable paper money) credit can expand to any arbitrary size. Eventually credit outstanding is greater than any bank’s reserves. The unwind that begins with depositors demanding their money back could wipe out any bank who committed this sin.

      But if banks do not mismatch their liabilities and assets, then it is a totally different story. The depositors specify the duration of their deposits. Immediately, there is a difference in that the depositors know their own plans and intentions. Second, if a depositor wants to withdraw early, he asks the bank what it will cost. The bank looks at the market for whatever loans or other assets it holds, and based on the bid in that market (subtracting a small fee for their services, of course) the bank goes back to the depositor and tells him how much he will lose to liquidate early. Obviously the problem is self-correcting. The more depositors want to withdraw prematurely, the more the bid on assets will recede, and this will tend to discourage depositors from further withdrawals. It is negative feedback (damping) rather than positive feedback (spiraling out of control until catastrophe).

      In the case where banks do not mismatch durations, the bank has no exposure to rollover risk or interest rate fluctuations (where the new cost of funding a long-term asset could be higher than the yield on the asset–a problem I did not even address!) The bank only has exposure to credit risk: what if the borrower defaults?

      I think *this* risk (unlike the risk of insolvency due to mismatched durations) can be resolved by setting aside reserves. An expert underwriter (which I am not!) should be able to determine such things as how much collateral to require, how much to keep in reserves vs. lend out, the appropriate ratio of bank equity to loan portfolio, etc.

      • Thanks for clearing that up. Have you read much of Gary Gorton’s work on this issue? He’s telling a similar story – funding was in the repo market and assets were MBS. Maturity mismatch is ever there was one.

  • duke

    …maturity mismatching…can lead to unsustainable booms.

  • OK. I would not argue against those who criticize bankers’ refusal to accept personal accountability. Nor would I deny that bankers are really overpaid small men in big outfits. That is all too often the case.
    But ordinary business risk is inevitable in making loans. If banks never made bad loans, there would be nothing to debate. FRB can – and does – work tolerably well in a well-ordered, stable society; where the money supply is intelligently managed; where contracts are enforceable; where individuals are made accountable for decisions; and whenever fundamental business confidence is robust.
    Compare bank loans to derivatives trading by ordinary ‘bettors’. I can buy stocks on margin. I may also enter into futures contracts and other types of derivatives which theoretically, may expose me to bigger risks than I can finance from my own resources. Futures trading, in this respect, is quite similar, on a smaller private level, to the multi-billion dollar loans of big banks, when loans are not fully under-written by assets. So a lot depends on personal ethics and business governance. Even when a loan is backed by sound assets, there remains a risk that the full value – or any substantial value – of the collateral assets is impossible to realize. The horrifying materialization of that risk appears to explain the banking system’s abject dependency on government support in the USA and the EU.
    But this begs another question: how do you competently and ethically manage risk? Could the Japanese have realistically forecast and mitigated the risks of a record-breaking earthquake? Can private insurance firms guarantee clients against any possible calamity? I think the answers to those questions are obvious.
    There were precedents to the Wall Street earthquake of 2008. One notorious financial commitment brought the undoing of the old and distinguished Barings’ Bank back in 1994: the sordid Ted Leeson affair, and the dishonest employee’s deluded recklessness.
    A weakness in corporate governance was the obvious explanation for Leeson’s disastrous folly, but that conclusion, like the offender’s jail sentence, must be cold comfort to the shareholders who lost all.
    By enforcing contractual protections and obligations, we can control irresponsible financial behavior and mitigate business risks. We can catch the outright law-breakers. We can luckily detect and eject the destructive cowboys from the financial system. But never eliminate them.
    In the end, as an insurance consultant once told me, business means taking risks. If you can’t stand the heat, stay out of the kitchen.

    • -my own correction; that was Nick Leeson, not Ted Leeson. – My own omission; it was also an earthquake in that year which triggered a loss of confidence in the Japanese stock market.
      - my real point: anyone in the business of lending and investing has no choice but to accept a degree of risk. Whether abolishing FRB can overcome such risk is the next stage of this (thankfully) intelligent debate.