October 2017

Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31        
Blog powered by Typepad

« Making Poor Nations Rich | Main | In Praise of Uncertainty?! »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Prof Horwitz,

What do you think of the argument that banks have incentives to lend out whatever they can because no single bank suffers the full cost of a credit crunch? I cannot remember which paper I read it in now, but it was modeled as a prisoner's dilemma with each bank's logical choice being to make a loan even if it might be defaulted on. I believe Huerta de Soto's book also says something about it, although I haven't gotten close to reading the whole thing. This argument always appealed to me more than banks simply expecting to be bailed out.

I suppose the incentives would be even greater if flipping of mortgage securities was possible, although I don't fully understand the economic conditions which enable flipping (is it a massive decrease in time preference?).

Ah, found the paper:
http://www.gmu.edu/rae/archives/VOL14_4_2001/4_carilli&dempster.pdf
In it, the acceptance of "bad loans" is also modeled as a commons (prisoner's dilemma) problem. I'm not really sure I buy that myself - most executives I've talked to seem to blame short-sighted decisions on quarterly-focused shareholders, which would seem to be indicative of a different problem to me.

While it would be interesting to show that banks have an incentive to lose money, It don't think it adds much to the common sense notion that in an uncertain world, banks intentionally make loans knowing that some will lose money. Even if there is a 99% chance that each loan will be fully paid, make enough loans and the probability that all loans will be repaid will become vanishingly small. Further, if some of the costs of default are born by others, then banks will take too much risk.

But... there may be other forces that cause banks to take too little risk. Determining whether credit markets generate too much or too little risk is impossible, much less determining the net balance and figuring out the proper regulation. And, of course, why do we think politicians will actually generate a scheme for perfecting markets?

It appears to me that the core of the "law & economics" debate between 100% banking (100B) and fractional reserve banking (FRB) is whether or not I can sell the full and immediate availability of one good more than one time. Being this a logical impossibility, it is not the same as credit risk: the borrower has time to find money, the depositor must yield it at demand.

FRB seems like overbooking in airline companies: they have 80 seats and sell 85, hoping 5 clients will not present themselves at check-in.

Contrary to 100B defenders, I believe that there is nothing strange with this contract, provided that the non-depositary nature of the contractual relation is explicitly revealed to the clients of the bank, to the borrowing firms, and to everybody who happens to use the money, otherwise it would be fraud.

Contrary to FRB defenders I hold that, if the nature of FRB is explicitly revealed to potential users, fiduciary media will most likely circulate at discounted price, like credit money, and I wonder if someone would ever accept a piece of paper that could be converted into money only if one is quick enough in rushing toward the bankrupt bank (and bankruptancy is inherent in fractional expansion, as de Soto has convincingly argued when talkin about non-insurable risks)... it would be much better the "loan-turned deposit" option of denying clients immediate convertibility at banker's will, which, being an option, has a market value, thus causing some discount anyway.

Contrary to 100B defenders, I don't believe that property rights violation and fiduciary media are two sides of the same coin.

Contrary to FRB, I'm almost sure their money, if no fraud occurs, would circulate at discount, and it might be that it wouldn't circulate, if people knew, but I don't know much about monetary history to be sure about that.

I'm further disappointed by the fact that no FRB defender has addressed the Austrian problem of intertemporal discoordination due to the injection of fiduciary media: White's book on the topic (the theory of mon. inst.) has only one Austrian aspect, the mengerian theory of the origin of money from barter, but has nothing on the relation between money and production; I don't know if there are other books without this limitation (Selgin's, Dowd's...).

A problem I have with 100B defenders is that it appears they confuse validity and efficacy of laws, like in legal positivism: the mere fact that a practice is declared illegal does not say much about the actual respect of that legal principle. However, punishment for fraud, when there is fraud (and not when there is everything is explicit), would surely help.

LF writes:

"I'm further disappointed by the fact that no FRB defender has addressed the Austrian problem of intertemporal discoordination due to the injection of fiduciary media:"

By this, do you mean ANY fiduciary media, or fiduciary media in excess of the demand to hold such media at the current price level? If the latter, then I certainly address it in my 2000 Routledge book. If the former, I think I address it as well as the whole point of my chapter on monetary equilibrium theory is to argue that as long as issuances of fiduciary media are not in excess of the demand to hold them, there are no intertemporal coordination problems created. Or at least no systematic ones - banks can be wrong about their loans, but there will be no pattern of intertemporal discoordination if their issuances are in sync with the demand to hold them.

Or have I misunderstood the concern you have LF?

Thanks for the answer.

I meant both types of fiduciary medias: both if the distinction between the two is considered relevant or not, I still haven't read a paper or a book which explains it, or may be I didn't notice it talked about it.

On Selgin's and White's "...we are misesians" and in Hayek's "denationalization..." there are traces of a connection between the price level and intertemporal coordination but I didn't extract a full argument.

At the moment, I don't understand why there should be any difference: fiduciary media are always injected in the credit market, and demand for money for investment purposes can, in normal conditions, always be stimulated by easy money, so it appears to me that there is always discoordination.

So, the baseline misesian argument that an increase in fiduciary media is self-reverting because if flows from credit markets to the rest of the economy, causing a Wicksell effect on real rates, appears to occur for every kind of fiduciary media. It is true that individuals may decide not to consume the money they will finally receive (thus making the initial investments sustainable), but their desire to save their future income should lower the interest rate now, without any need of fiduciary media whatsoever.

Even though I believe that price-deflation may have some effect on investment, I don't believe that price stability implies investment sustainability, as Selgin and White seem to state at the end of "we are not misesians".

I recently re-read Hayek's "denationalizatin" and he made similar claims, but I failed to obtain an argument on this specific topic, to accept of refuse, out of that book.

I'm a little bit embarassed because I didn't know your book was (also?) about this topic, I was looking for it weeks ago on Amazon but I was a little afraid for the price, even though I buy in euros. :-)

LF:

We probably need to specify the institutional arrangement in play. I agree that fiduciary media *in a central banking* system are highly problematic because they are likely to be overissued. Not only do central banks face the usual knowledge problems in determining the demand for their liabilities, they also face incentives to inflate to reduce the real value of government debt. And, last but not least, their fiduciary media are not likely to be based on a commodity, thus reducing the costs to them of overissuing. But these are not problems inherent in fiduciary media, but in central banking.

However, if we are talking about a free banking system with fiduciary media, these concerns go away. The only way such banks can expand their issuances is if their customers wish to hold more of their notes or deposits. If the demand to hold such balances rises, it signals that those folks wish to delay consumption. It also means that the bank now has excess reserves, as the higher money holdings slow the number of redemptions.

Free banks respond by lending out those excess reserves at a marginally lower interest rate. That lower interest rate is NOT discoordinative because the higher demand for money implies lower time preferences (more patience for goods) and the lower interest rate will encourage longer-term investment projects, which is the coordinating response.

Put differently, under free banking, holding more bank liabilities adds to the supply of loanable funds and banks respond to that shift in the supply curve by lowering rates, inducing in new borrowers along the demand for LF curve, leading to the market clearing at a higher quantity of LF being exchanged. By doing so, the banks actually enhance intertemporal coordination. If they did nothing, we'd have more savings but no corresponding investment and the deflationary depression that people like Wicksell, Warburton, and Yeager warn about.

So again, I'm not sure why fiduciary media are inherently problematic in the way I think you are arguing.

Deposits or notes that can be redeemed immediately can only trade at a discount related to the cost of redemption. Reserve holding has nothing to do with it. This is elementary.

If they trade at a discount greater than the cost of redemption, then there are arbitrage profits from buying the discounted checks or banknotes and redeeming them.

In reality, they don't trade at any discount. That is because interbank clearing agreements reduce effective redemption costs to zero. (Unless, of course, it is the 19th century and you have unit banking regulations.)

In reality what happens is that each bank must limit its issue to the demand for holding them at their face value. If no one wants to hold fractional reserve banknotes or deposits, then any banks issue will be rapidly redeemed. If there are people who are willing to hold such deposits or banknotes, then they can and will issue a quantity matching that demand at the face value.

If we assume that people care about bank reserve holding, then we can imagine some people might hold deposits "backed" by 3% reserves paying 2% interest. Others might hold deposits "backed" by 60% reserves paying .5% interest. And some people might insist on 100% reserves and pay a 3% storage fee to the "bank."

The financial innovations that have developed to avoid the 10% reserve requirements of today show exactly how easy it is to avoid a 100% reserve requirement.

While Rothbard claimed somewhere that it is a great rule of banking to always match maturities, I don't know that even he ever claimed that a bank borrowing for 6 months to finance a 1 year loan is committing fraud.

When 6 months pass, the bank could sell the one year loan or (as one would expect) borrow funds to pay off the first loan, say by borrowing for six months a second time. That, of course, is what it means to create liquidity through financial intermediation. And it also creates liquidity risk. If, after six months, it is more difficult to borrow, which would usually be shown by a higher interest rate, then it is possible that the unexpectedly high interest costs (along with the other costs) will be greater than the interest generated by the one year loan. The bank could lose money on the loan. If a bank's losses are greater than its net worth, it fails. The bank cannot pay off all of its creditors in full. This possiblity is the liquidity risk borne by the bank's creditors.

All of this could exist if everyone solely used gold coin for money. (The bank would borrow gold, lend gold, collect repayments in gold, and make payments in gold.)

But if banks can finance a one year loan by borrowing for six months, why not for three months? Or one month? Or one week? Or one day?
One hour? minute? second? Very short terms to maturity would be inconvenient without automatic rollover, but how is that fraud?

While issuing banknotes and deposits "payable on demand" is traditional, there is no need to use that specific formula to create what is an equivalent liquidity risk.

Even if the resulting financial instruments are not checkable or in a form that can be used as currency, any creation of liquidity through finanical intermediation can reduce the demand for money. Moving towards the practical extreme of overnight financial instruments, the possibility for a substantial reduction of money demand is possible.

If money were solely gold coins, then the reduction in the demand for gold coins (and gold) would result in a lower purchasing power of money. (What some of us call "inflation.")

My view is that that making these financial instruments checkable or issuing bearer notes on the same terms that can be used as currency makes no difference in terms of liquidity risk, but these things would be what I define as money.

I don't think that any rational person would assume that finanical intermediaries provide free gold storage services. I do think that families, schools, and the like should explain to people what financial intermediaries do. As regulations go, requiring banks to disclose that they are finanical intermediates and not storing money for free isn't too brudensome. But it certainly is silly. And requiring that banks put in a one second clause or something because using the traditional term "payable on demand" creates some kind of legal problem that isn't just reducable to what people understand the contract to mean.. that would be a waste.

Prof. Horwitz: Thanks for the answer, I'll think about it.

The comments to this entry are closed.