Steve Keen’s DebtWatch No 31 February 2009: “The Roving Cavaliers of Credit”

flattr this!

Click on Marx's image to download this post in PDF format

Talk about centralisation! The credit system, which has its focus in the so-called national banks and the big money-lenders and usurers surrounding them, constitutes enormous centralisation, and gives this class of parasites the fabulous power, not only to periodically despoil industrial capitalists, but also to interfere in actual production in a most dangerous manner— and this gang knows nothing about production and has nothing to do with it.” [1]

Ten years ago, a quote from Marx would have one deemed a socialist, and dismissed from polite debate. Today, such a quote can (and did, along with Charlie's photo) appear in a feature in the Sydney Morning Herald—and not a few people would have been nodding their heads at how Marx got it right on bankers.F

He got it wrong on some other issues,[2] but his analysis of money and credit, and how the credit system can bring an otherwise well-functioning market economy to its knees, was spot on. His observations on the financial crisis of 1857 still ring true today:

"A high rate of interest can also indicate, as it did in 1857, that the country is undermined by the rov­ing cav­a­liers of credit who can afford to pay a high inter­est because they pay it out of other people’s pock­ets (whereby, how­ever, they help to deter­mine the rate of inter­est for all), and mean­while they live in grand style on antic­i­pated prof­its.

Simul­ta­ne­ously, pre­cisely this can inci­den­tally pro­vide a very prof­itable busi­ness for man­u­fac­tur­ers and oth­ers. Returns become wholly decep­tive as a result of the loan sys­tem…“[1]

One and a half cen­turies after Marx falsely pre­dicted the demise of cap­i­tal­ism, the peo­ple most likely to bring it about are not work­ing class rev­o­lu­tion­ar­ies, but the “Rov­ing Cav­a­liers of Credit”, against whom Marx quite justly railed.

This month’s Debt­watch is ded­i­cated to analysing how these Cav­a­liers actu­ally “make” money and debt—something they think they under­stand, but in real­ity, they don’t. A sound model of how money and debt are cre­ated makes it obvi­ous that we should never have fallen for the insane notion that the finan­cial sys­tem should be self-regulating. All that did was give the Cav­a­liers a licence to run amok, with the con­se­quences we are now expe­ri­enc­ing yet again—150 years after Marx described the cri­sis that led him to write Das Kapital.

The con­ven­tional model: the “Money Multiplier”

Every macro­eco­nom­ics text­book has an expla­na­tion of how credit money is cre­ated by the sys­tem of frac­tional bank­ing that goes some­thing like this:

  • Banks are required to retain a cer­tain per­cent­age of any deposit as a reserve, known as the “reserve require­ment”; for sim­plic­ity, let’s say this frac­tion is 10%.
  • When cus­tomer Sue deposits say 100 newly printed gov­ern­ment $10 notes at her bank, it is then obliged to hang on to ten of them—or $100—but it is allowed to lend out the rest.
  • The bank then lends $900 to its cus­tomer Fred, who then deposits it in his bank—which is now required to hang on to 9 of the bills—or $90—and can lend out the rest. It then lends $810 to its cus­tomer Kim.
  • Kim then deposits this $810 in her bank. It keeps $81 of the deposit, and lends the remain­ing $729 to its cus­tomer Kevin.
  • And on this iter­a­tive process goes.
  • Over time, a total of $10,000 in money is created—consisting of the orig­i­nal $1,000 injec­tion of gov­ern­ment money plus $9,000 in credit money—as well as $9,000 in total debts. The fol­low­ing table illus­trates this, on the assump­tion that the time lag between a bank receiv­ing a new deposit, mak­ing a loan, and the recip­i­ent of the loan deposit­ing them in other banks is a mere one week.

This model of how banks cre­ate credit is sim­ple, easy to under­stand (this ver­sion omits the fact that the pub­lic holds some of the cash in its own pock­ets rather than deposit­ing it all in the banks; this detail is eas­ily catered for and is part of the stan­dard model taught to econ­o­mists),… and com­pletely inad­e­quate as an expla­na­tion of the actual data on money and debt.

The Data ver­sus the Money Mul­ti­plier Model

Two hypothe­ses about the nature of money can be derived from the money mul­ti­plier model:

1. The cre­ation of credit money should hap­pen after the cre­ation of gov­ern­ment money. In the model, the bank­ing sys­tem can’t cre­ate credit until it receives new deposits from the pub­lic (that in turn orig­i­nate from the gov­ern­ment) and there­fore finds itself with excess reserves that it can lend out. Since the lend­ing, deposit­ing and relend­ing process takes time, there should be a sub­stan­tial time lag between an injec­tion of new government-created money and the growth of credit money.

2. The amount of money in the econ­omy should exceed the amount of debt, with the dif­fer­ence rep­re­sent­ing the government’s ini­tial cre­ation of money. In the exam­ple above, the total of all bank deposits tapers towards $10,000, the total of loans con­verges to $9,000, and the dif­fer­ence is $1,000, which is the amount of ini­tial gov­ern­ment money injected into the sys­tem. There­fore the ratio of Debt to Money should be less than one, and close to (1-Reserve Ratio): in the exam­ple above, D/M=0.9, which is 1 minus the reserve ratio of 10% or 0.1.

Both these hypothe­ses are strongly con­tra­dicted by the data.

Test­ing the first hypoth­e­sis takes some sophis­ti­cated data analy­sis, which was done by two lead­ing neo­clas­si­cal econ­o­mists in 1990.[3] If the hypoth­e­sis were true, changes in M0 should pre­cede changes in M2. The time pat­tern of the data should look like the graph below: an ini­tial injec­tion of gov­ern­ment “fiat” money, fol­lowed by a grad­ual cre­ation of a much larger amount of credit money:

Their empir­i­cal con­clu­sion was just the oppo­site: rather than fiat money being cre­ated first and credit money fol­low­ing with a lag, the sequence was reversed: credit money was cre­ated first, and fiat money was then cre­ated about a year later:

There is no evi­dence that either the mon­e­tary base or M1 leads the cycle, although some econ­o­mists still believe this mon­e­tary myth. Both the mon­e­tary base and M1 series are gen­er­ally pro­cycli­cal and, if any­thing, the mon­e­tary base lags the cycle slightly. (p. 11)

The dif­fer­ence in the behav­ior of M1 and M2 sug­gests that the dif­fer­ence of these aggre­gates (M2 minus M1) should be con­sid­ered… The dif­fer­ence of M2M1 leads the cycle by even more than M2, with the lead being about three quar­ters.” (p. 12)

Thus rather than credit money being cre­ated with a lag after gov­ern­ment money, the data shows that credit money is cre­ated first, up to a year before there are changes in base money. This con­tra­dicts the money mul­ti­plier model of how credit and debt are cre­ated: rather than fiat money being needed to “seed” the credit cre­ation process, credit is cre­ated first and then after that, base money changes.

It doesn’t take sophis­ti­cated sta­tis­tics to show that the sec­ond pre­dic­tion is wrong—all you have to do is look at the ratio of pri­vate debt to money. The the­o­ret­i­cal pre­dic­tion has never been right—rather than the money stock exceed­ing debt, debt has always exceeded the money supply—and the degree of diver­gence has grown over time.(there are atten­u­at­ing fac­tors that might affect the prediction—the pub­lic hoard­ing cash should make the ratio less than shown here, while non-banks would make it larger—but the gap between pre­dic­tion and real­ity is just too large for the the­ory to be taken seriously).

Aca­d­e­mic eco­nom­ics responded to these empir­i­cal chal­lenges to its accepted the­ory in the time-honoured way: it ignored them.

Well, the so-called “main­stream” did—the school of thought known as “Neo­clas­si­cal eco­nom­ics”. A rival school of thought, known as Post Key­ne­sian eco­nom­ics, took these prob­lems seri­ously, and devel­oped a dif­fer­ent the­ory of how money is cre­ated that is more con­sis­tent with the data.

This first major paper on this approach, “The Endoge­nous Money Stock” by the non-orthodox econ­o­mist Basil Moore, was pub­lished almost thirty years ago.[4] Basil’s essen­tial point was quite sim­ple. The stan­dard money mul­ti­plier model’s assump­tion that banks wait pas­sively for deposits before start­ing to lend is false. Rather than bankers sit­ting back pas­sively, wait­ing for depos­i­tors to give them excess reserves that they can then on-lend,

In the real world, banks extend credit, cre­at­ing deposits in the process, and look for reserves later”.[5]

Thus loans come first—simultaneously cre­at­ing deposits—and at a later stage the reserves are found. The main mech­a­nism behind this are the “lines of credit” that major cor­po­ra­tions have arranged with banks that enable them to expand their loans from what­ever they are now up to a spec­i­fied limit.

If a firm accesses its line of credit to, for exam­ple, buy a new piece of machin­ery, then its debt to the bank rises by the price of the machine, and the deposit account of the machine’s man­u­fac­turer rises by the same amount. If the bank that issued the line of credit was already at its own limit in terms of its reserve require­ments, then it will bor­row that amount, either from the Fed­eral Reserve or from other sources.

If the entire bank­ing sys­tem is at its reserve require­ment limit, then the Fed­eral Reserve has three choices:

  • refuse to issue new reserves and cause a credit crunch;
  • cre­ate new reserves; or
  • relax the reserve ratio.

Since the main role of the Fed­eral Reserve is to try to ensure the smooth func­tion­ing of the credit sys­tem, option one is out—so it either adds Base Money to the sys­tem, or relaxes the reserve require­ments, or both.

Thus cau­sa­tion in money cre­ation runs in the oppo­site direc­tion to that of the money mul­ti­plier model: the credit money dog wags the fiat money tail. Both the actual level of money in the sys­tem, and the com­po­nent of it that is cre­ated by the gov­ern­ment, are con­trolled by the com­mer­cial sys­tem itself, and not by the Fed­eral Reserve.

Cen­tral Banks around the world learnt this les­son the hard way in the 1970s and 1980s when they attempted to con­trol the money sup­ply, fol­low­ing neo­clas­si­cal econ­o­mist Mil­ton Friedman’s the­ory of “mon­e­tarism” that blamed infla­tion on increases in the money sup­ply. Fried­man argued that Cen­tral Banks should keep the reserve require­ment con­stant, and increase Base Money at about 5% per annum; this would, he asserted cause infla­tion to fall as people’s expec­ta­tions adjusted, with only a minor (if any) impact on real eco­nomic activity.

Though infla­tion was ulti­mately sup­pressed by a severe reces­sion, the mon­e­tarist exper­i­ment over­all was an abject fail­ure. Cen­tral Banks would set tar­gets for the growth in the money sup­ply and miss them completely—the money sup­ply would grow two to three times faster than the tar­gets they set.

Ulti­mately, Cen­tral Banks aban­doned mon­e­tary tar­get­ting, and moved on to the mod­ern approach of tar­get­ting the overnight inter­est rate as a way to con­trol infla­tion.[6] Sev­eral Cen­tral Banks—including Australia’s RBA—completely aban­doned the set­ting of reserve require­ments. Others—such as America’s Fed­eral Reserve—maintained them, but had such loop­holes in them that they became basi­cally irrel­e­vant. Thus the US Fed­eral Reserve sets a Required Reserve Ratio of 10%, but applies this only to deposits by indi­vid­u­als; banks have no reserve require­ment at all for deposits by com­pa­nies.[7]

How­ever, neo­clas­si­cal eco­nomic the­ory never caught up with either the data, or the actual prac­tices of Cen­tral Banks—and Ben Bernanke, a lead­ing neo­clas­si­cal the­o­reti­cian, and unabashed fan of Mil­ton Fried­man, is now in con­trol of the Fed­eral Reserve. He is there­fore try­ing to resolve the finan­cial cri­sis and pre­vent defla­tion in a neo­clas­si­cal man­ner: by increas­ing the Base Money supply.

Give Bernanke credit for try­ing here: the rate at which he is increas­ing Base Money is unprece­dented. Base Money dou­bled between 1994 and 2008; Bernanke has dou­bled it again in just the last 4 months.

If the money mul­ti­plier model of money cre­ation were cor­rect, then ulti­mately this would lead to a dra­matic growth in the money sup­ply as an addi­tional US$7 tril­lion of credit money was grad­u­ally created.

If neo­clas­si­cal the­ory was cor­rect, this increase in the money sup­ply would cause a bout of infla­tion, which would end bring the cur­rent defla­tion­ary period to a halt, and we could all go back to “busi­ness as usual”. That is clearly what Bernanke is bank­ing on:

The con­clu­sion that defla­tion is always reversible under a fiat money sys­tem fol­lows from basic eco­nomic rea­son­ing. A lit­tle para­ble may prove use­ful: Today an ounce of gold sells for $300, more or less. Now sup­pose that a mod­ern alchemist solves his subject’s old­est prob­lem by find­ing a way to pro­duce unlim­ited amounts of new gold at essen­tially no cost. More­over, his inven­tion is widely pub­li­cized and sci­en­tif­i­cally ver­i­fied, and he announces his inten­tion to begin mas­sive pro­duc­tion of gold within days.

What would hap­pen to the price of gold? Pre­sum­ably, the poten­tially unlim­ited sup­ply of cheap gold would cause the mar­ket price of gold to plum­met. Indeed, if the mar­ket for gold is to any degree effi­cient, the price of gold would col­lapse imme­di­ately after the announce­ment of the inven­tion, before the alchemist had pro­duced and mar­keted a sin­gle ounce of yel­low metal.

Like gold, U.S. dol­lars have value only to the extent that they are strictly lim­ited in sup­ply. But the U.S. gov­ern­ment has a tech­nol­ogy, called a print­ing press (or, today, its elec­tronic equiv­a­lent), that allows it to pro­duce as many U.S. dol­lars as it wishes at essen­tially no cost.

By increas­ing the num­ber of U.S. dol­lars in cir­cu­la­tion, or even by cred­i­bly threat­en­ing to do so, the U.S. gov­ern­ment can also reduce the value of a dol­lar in terms of goods and ser­vices, which is equiv­a­lent to rais­ing the prices in dol­lars of those goods and ser­vices. We con­clude that, under a paper-money sys­tem, a deter­mined gov­ern­ment can always gen­er­ate higher spend­ing and hence pos­i­tive inflation…

If we do fall into defla­tion, how­ever, we can take com­fort that the logic of the print­ing press exam­ple must assert itself, and suf­fi­cient injec­tions of money will ulti­mately always reverse a defla­tion.” [8]

How­ever, from the point of view of the empir­i­cal record, and the rival the­ory of endoge­nous money, this will fail on at least four fronts:

1. Banks won’t cre­ate more credit money as a result of the injec­tions of Base Money. Instead, inac­tive reserves will rise;

2. Cre­at­ing more credit money requires a match­ing increase in debt—even if the money mul­ti­plier model were cor­rect, what would the odds be of the pri­vate sec­tor tak­ing on an addi­tional US$7 tril­lion in debt in addi­tion to the cur­rent US$42 tril­lion it already owes?;

3. Defla­tion will con­tinue because the motive force behind it will still be there—distress sell­ing by retail­ers and whole­salers who are des­per­ately try­ing to avoid going bank­rupt; and

4. The macro­eco­nomic process of delever­ag­ing will reduce real demand no mat­ter what is done, as Microsoft CEO Steve Ballmer recently noted:  “We’re cer­tainly in the midst of a once-in-a-lifetime set of eco­nomic con­di­tions. The per­spec­tive I would bring is not one of reces­sion. Rather, the econ­omy is reset­ting to lower level of busi­ness and con­sumer spend­ing based largely on the reduced lever­age in econ­omy”.[9]

The only way that Bernanke’s “print­ing press exam­ple” would work to cause infla­tion in our cur­rent debt-laden would be if sim­ply Zim­bab­wean lev­els of money were printed—so that fiat money could sub­stan­tially repay out­stand­ing debt and effec­tively sup­plant credit-based money.

Mea­sured on this scale, Bernanke’s increase in Base Money goes from being heroic to triv­ial. Not only does the scale of credit-created money greatly exceed government-created money, but debt in turn greatly exceeds even the broad­est mea­sure of the money stock—the M3 series that the Fed some years ago decided to discontinue.

Bernanke’s expan­sion of M0 in the last four months of 2008 has merely reduced the debt to M0 ratio from 47:1 to 36:1 (the debt data is quar­terly whole money stock data is monthly, so the fall in the ratio is more than shown here given the lag in report­ing of debt).

To make a seri­ous dent in debt lev­els, and thus enable the increase in base money to affect the aggre­gate money stock and hence cause infla­tion, Bernanke would need to not merely dou­ble M0, but to increase it by a fac­tor of, say, 25 from pre-intervention lev­els. That US$20 tril­lion truck­load of green­backs might enable Amer­i­cans to repay, say, one quar­ter of out­stand­ing debt with one half—thus reduc­ing the debt to GDP ratio about 200% (roughly what it was dur­ing the Dot­Com bub­ble and, coin­ci­den­tally, 1931)—and get back to some seri­ous infla­tion­ary spend­ing with the other (of course, in the con­text of a seri­ously depre­ci­at­ing cur­rency). But with any­thing less than that, his attempts to reflate the Amer­i­can econ­omy will sink in the ocean of debt cre­ated by America’s modern-day “Rov­ing Cav­a­liers of Credit”.

How to be a “Cav­a­lier of Credit”

Note Bernanke’s assump­tion (high­lighted above) in his argu­ment that print­ing money would always ulti­mately cause infla­tion: “under a fiat money sys­tem”. The point made by endoge­nous money the­o­rists is that we don’t live in a fiat-money sys­tem, but in a credit-money sys­tem which has had a rel­a­tively small and sub­servient fiat money sys­tem tacked onto it.

We are there­fore not in a “frac­tional reserve bank­ing sys­tem”, but in a credit-money one, where the dynam­ics of money and debt are vastly dif­fer­ent to those assumed by Bernanke and neo­clas­si­cal eco­nom­ics in gen­eral.[10]

Call­ing our cur­rent finan­cial sys­tem a “fiat money” or “frac­tional reserve bank­ing sys­tem” is akin to the blind man who clas­si­fied an ele­phant as a snake, because he felt its trunk. We live in a credit money sys­tem with a fiat money sub­sys­tem that has some inde­pen­dence, but cer­tainly doesn’t rule the mon­e­tary roost—far from it.

The best place to start to analyse the mon­e­tary sys­tem is there­fore to con­sider a model of a pure credit economy—a toy econ­omy in which there is no gov­ern­ment sec­tor and no Cen­tral Bank whatsoever—and see how that model behaves.

The first issue in such a sys­tem is how does one become a bank?—or a “cav­a­lier of credit” in Marx’s won­der­fully evoca­tive phrase? The answer was pro­vided by the Ital­ian non-orthodox econ­o­mist Augusto Graziani: a bank is a third party to all trans­ac­tions, whose account-keeping between buyer and seller is regarded as finally set­tling all claims between them.

Huh? What does that mean? To explain it, I’ll com­pare it with the man­ner in which we’ve been mis­led to think­ing about the mar­ket econ­omy by neo­clas­si­cal economics.

It has deluded us into think­ing of a mar­ket econ­omy as being fun­da­men­tally a sys­tem of barter. Every trans­ac­tion is seen as being two sided, and involv­ing two com­modi­ties: Farmer Maria wants to sell pigs and buy cop­per pipe; Plumber Joe wants to sell cop­per pipe and buy pigs.

Money sim­ply elim­i­nates the prob­lem that it’s very hard for Plumber Joe to find Farmer Maria. Instead, they each sell their com­mod­ity for money, and then exchange that money for the com­mod­ity they really want. The pic­ture appears more complicated—there are two mar­kets intro­duced as well, with Farmer Maria sell­ing pigs to the pig mar­ket in return for money, Plumber Joe doing the same thing in the cop­per mar­ket, and then armed with money from their sales, they go across to the other mar­ket and buy what they want. But it is still a lot eas­ier than a plumber going out to try to find a pig farmer who wants cop­per pipes.

In this model of the econ­omy, money is use­ful in that it replaces a very dif­fi­cult search process with a sys­tem of mar­kets. But fun­da­men­tally the sys­tem is no dif­fer­ent to the barter model above: money is just a con­ve­nient “numeraire”, and any­thing at all could be used—even cop­per pipe or pigs—so long as all mar­kets agreed to accept it. Gold tends to be the numeraire of choice because it doesn’t degrade, and paper money merely replaces gold as a more con­ve­nient form of numeraire.

Impor­tantly, in this model, money is an asset to its holder, but a lia­bil­ity to no-one. There is money, but no debt. The frac­tional bank­ing model that is tacked onto this vision of bar­ter­ing adds yet another mar­ket where depos­i­tors (savers) sup­ply money at a price (the rate of inter­est), and lenders buy money for that price, and the inter­ac­tion between sup­ply and demand sets the price. Debt now exists, but in the model world total debt is less than the amount of money.

If this mar­ket pro­duces too much money (which it can do in a frac­tional bank­ing sys­tem because the gov­ern­ment deter­mines the sup­ply of base money and the reserve require­ment) then there can be infla­tion of the money prices of com­modi­ties. Equally if the money mar­ket sud­denly con­tracts, then there can be defla­tion. It’s fairly easy to sit­u­ate Bernanke’s dra­matic increase in Base Money within this view of the world.

If only it were the world in which we live. Instead, we live in a credit econ­omy, in which intrin­si­cally use­less pieces of paper—or even sim­ple trans­fers of elec­tronic records of numbers—are hap­pily accepted in return for real, hard com­modi­ties. This in itself is not incom­pat­i­ble with a frac­tional bank­ing model, but the empir­i­cal data tells us that credit money is cre­ated inde­pen­dently of fiat money: credit money rules the roost. So our fun­da­men­tal under­stand­ing of a mon­e­tary econ­omy should pro­ceed from a model in which credit is intrin­sic, and gov­ern­ment money is tacked on later—and not the other way round.

Our start­ing point for analysing the econ­omy should there­fore be a “pure credit” econ­omy, in which there are pri­vately issued bank notes, but no gov­ern­ment sec­tor and no fiat money. Yet this has to be an econ­omy in which intrin­si­cally use­less items are accepted as pay­ment for intrin­si­cally use­ful ones—you can’t eat a bank note, but you can eat a pig.

So how can that be done with­out cor­rupt­ing the entire sys­tem. Some­one has to have the right to pro­duce the bank notes; how can this sys­tem be the basis of exchange, with­out the per­son who has that right abus­ing it?

Graziani (and oth­ers in the “Cir­cuitist” tra­di­tion) rea­soned that this would only be pos­si­ble if the pro­ducer of bank notes—or the keeper of the elec­tronic records of money—could not sim­ply print them when­ever he/she wanted a com­mod­ity, and go and buy that com­mod­ity with them. But at the same time, peo­ple involved in ordi­nary com­merce had to accept the trans­fer of these intrin­si­cally use­less things in return for commodities.

There­fore for a sys­tem of credit money to work, three con­di­tions had to be fulfilled:

In order for money to exist, three basic con­di­tions must be met:

a) money has to be a token cur­rency (oth­er­wise it would give rise to barter and not to mon­e­tary exchanges);

b) money has to be accepted as a means of final set­tle­ment of the trans­ac­tion (oth­er­wise it would be credit and not money);

c) money must not grant priv­i­leges of seignor­age to any agent mak­ing a pay­ment.” [11]

In Graziani’s words, “The only way to sat­isfy those three con­di­tions is …:

to have pay­ments made by means of promises of a third agent, the typ­i­cal third agent being nowa­days. When an agent makes a pay­ment by means of a cheque, he sat­is­fies his part­ner by the promise of the bank to pay the amount due.

Once the pay­ment is made, no debt and credit rela­tion­ships are left between the two agents. But one of them is now a cred­i­tor of the bank, while the sec­ond is a debtor of the same bank. This insures that, in spite of mak­ing final pay­ments by means of paper money, agents are not granted any kind of privilege.

For this to be true, any mon­e­tary pay­ment must there­fore be a tri­an­gu­lar trans­ac­tion, involv­ing at least three agents, the payer, the payee, and the bank.” ( p. 3).

Thus in a credit econ­omy, all trans­ac­tions are involve one com­mod­ity, and three par­ties: a seller, a buyer, and a bank whose trans­fer of money from the buyer’s account to the seller’s is accepted by them as final­is­ing the sale of the com­mod­ity. So the actual pat­tern in any trans­ac­tion in a credit money econ­omy is as shown below:

This makes banks and money an essen­tial fea­ture of a credit econ­omy, not some­thing that can be ini­tially ignored and incor­po­rated later, as neo­clas­si­cal eco­nom­ics has attempted to do (unsuc­cess­fully; one of the hard­est things for a neo­clas­si­cal math­e­mat­i­cal mod­eller is to explain why money exists, apart from the search advan­tages noted above. Gen­er­ally there­fore their mod­els omit money—and debt—completely).

It also defines what a bank is: it is a third party whose record-keeping is trusted by all par­ties as record­ing the trans­fers of credit money that effect sales of com­modi­ties. The bank makes a legit­i­mate liv­ing by lend­ing money to other agents—thus simul­ta­ne­ously cre­at­ing loans and deposits—and charg­ing a higher rate of inter­est on loans than on deposits.

Thus in a fun­da­men­tal way, a bank is a bank because it is trusted. Of course, as we know from our cur­rent bit­ter expe­ri­ence, banks can dam­age that trust; but it remains the well­spring from which their exis­tence arises.

This model helped dis­tin­guish the real­is­tic model of endoge­nous money from the unre­al­is­tic neo­clas­si­cal vision of a barter econ­omy. It also makes it pos­si­ble to explain what a credit crunch is, and why it has such a dev­as­tat­ing impact upon eco­nomic activity.

First, the basics: how does a pure credit econ­omy work, and how is money cre­ated in one? (The rest of this post nec­es­sar­ily gets tech­ni­cal and is there for those who want detailed back­ground. It reports new research into the dynam­ics of a credit econ­omy. There’s noth­ing here any­where near as poetic as Marx’s “Cav­a­liers of Credit”, but I hope it explains how a credit econ­omy works, and how it can go badly wrong in a “credit crunch”)

How the Cav­a­liers “Make Money”

Sev­eral economists—notably Wick­sell and Keynes—envisaged a “pure credit econ­omy”. Keynes imag­ined a world in which “invest­ment is pro­ceed­ing at a steady rate”, in which case:

the finance (or the com­mit­ments to finance) required can be sup­plied from a revolv­ing fund of a more or less con­stant amount, one entre­pre­neur hav­ing his finance replen­ished for the pur­pose of a pro­jected invest­ment as another exhausts his on pay­ing for his com­pleted invest­ment.” [12]

This is the start­ing point to under­stand­ing a pure credit economy—and there­fore to under­stand­ing our cur­rent econ­omy and why it’s in a bind. Con­sider an econ­omy with three sec­tors: firms that pro­duce goods, banks that charge and pay inter­est, and house­holds that sup­ply work­ers. Firms are the only enti­ties that bor­row, and the bank­ing sec­tor gave loans at some stage in the past to start pro­duc­tion. Firms hired work­ers with this money (and bought inputs from each other), enabling pro­duc­tion, and ulti­mately the econ­omy set­tled down to a con­stant turnover of money and goods (as yet there is no tech­no­log­i­cal change, pop­u­la­tion growth, or wage bargaining).

There are four types of accounts: Firms’ Loans, Firms’ Deposits, Banks’ Deposits, and House­holds Deposits. These finan­cial flows are described by the fol­low­ing table. I’m eschew­ing math­e­mat­i­cal sym­bols and just using let­ters here to avoid the “MEGO” effect (“My Eyes Glaze Over”)—if you want to check out the equa­tions, see this paper:

1. Inter­est accrues on the out­stand­ing loans.

2. Firms pay inter­est on the loans. This is how the banks make money, and it involves a trans­fer of money from the firms deposit accounts to the banks. The banks then have to acknowl­edge this pay­ment of inter­est by record­ing it against the out­stand­ing debt firms owe them.

3. Banks pay inter­est to firms on the bal­ances in their deposit accounts. This involves a trans­fer from Bank Deposit accounts to Firms; this is a cost of busi­ness to banks, but they make money this way because (a) the rate of inter­est on loans is higher than that on deposits and (b) as is shown later, the vol­ume of loans out­stand­ing exceeds the deposits that banks have to pay inter­est on;

4. Firms pay wages to work­ers; this is a trans­fer from the firms deposits to the households.

5. Banks pay inter­est to house­holds on the bal­ances in their deposit accounts.

6. Banks and house­holds pay money to firms in order to pur­chase some of the out­put from fac­to­ries for con­sump­tion and inter­me­di­ate goods.

This finan­cial activ­ity allows pro­duc­tion to take place:

1. Work­ers are hired and paid a wage;

2. They pro­duce out­put in fac­to­ries at a con­stant level of productivity;

3. The out­put is then sold to other firms, banks and households;

4. The price level is set so that in equi­lib­rium the flow of demand equals the flow of output

The graphs below show the out­come of a sim­u­la­tion of this sys­tem, which show that a pure credit econ­omy can work: firms can bor­row money, make a profit and pay it back, and a sin­gle “revolv­ing fund of finance”, as Keynes put it, can main­tain a set level of eco­nomic activ­ity. [13]

These sta­ble accounts sup­port a flow of eco­nomic activ­ity in time, giv­ing firms, house­holds and banks steady incomes:

Out­put and employ­ment also tick over at a con­stant level:

That’s the absolutely basic pic­ture; to get closer to our cur­rent real­ity, a lot more needs to be added. The next model includes, in addi­tion to the basic sys­tem shown above:

1. Repay­ment of debt, which involves a trans­fer from the Firms’ deposit account to an account that wasn’t shown in the pre­vi­ous model that records Banks unlent reserves; this trans­fer of money has to be acknowl­edged by the banks by a match­ing reduc­tion in the recorded level of debt;

2. Relend­ing from unlent reserves. This involves a trans­fer of money, against which an equiv­a­lent increase in debt is recorded;

3. The exten­sion of new loans to the firm sec­tor by the banks. The firms sector’s deposits are increased, and simul­ta­ne­ously the recorded level of debt is increased by the same amount.

4. Invest­ment of part of bank prof­its by a trans­fer from the bank­ing sector’s deposit accounts to the unlent reserves.

5. Vari­able wages, grow­ing labour pro­duc­tiv­ity and a grow­ing population.

The finan­cial table for this sys­tem is:

As with the pre­vi­ous model, this toy econ­omy “works”—it is pos­si­ble for firms to bor­row money, make a profit, and repay their debt.

With the addi­tional ele­ments of debt repay­ment and the cre­ation of new money, this model also lets us see what hap­pens to bank income when these para­me­ters change.

Though in some ways the answers are obvi­ous, it lets us see why banks are truly cav­a­lier with credit. The con­clu­sions are that bank income is bigger:

  • If the rate of money cre­ation is higher (this is by far the most impor­tant factor);
  • If the rate of cir­cu­la­tion of unlent reserves is higher; and
  • If the rate of debt repay­ment is lower—which is why, in “nor­mal” finan­cial cir­cum­stances, banks are quite happy not to have debt repaid.

In some ways these con­clu­sions are unre­mark­able: banks make money by extend­ing debt, and the more they cre­ate, the more they are likely to earn. But this is a rev­o­lu­tion­ary con­clu­sion when com­pared to stan­dard think­ing about banks and debt, because the money mul­ti­plier model implies that, what­ever banks might want to do, they are con­strained from so doing by a money cre­ation process that they do not control.

How­ever, in the real world, they do con­trol the cre­ation of credit. Given their pro­cliv­ity to lend as much as is pos­si­ble, the only real con­straint on bank lend­ing is the public’s will­ing­ness to go into debt. In the model econ­omy shown here, that will­ing­ness directly relates to the per­ceived pos­si­bil­i­ties for prof­itable investment—and since these are lim­ited, so also is the uptake of debt.

But in the real world—and in my mod­els of Minsky’s Finan­cial Insta­bil­ity Hypoth­e­sis—there is an addi­tional rea­son why the pub­lic will take on debt: the per­cep­tion of pos­si­bil­i­ties for pri­vate gain from lever­aged spec­u­la­tion on asset prices.

That clearly is what has hap­pened in the world’s recent eco­nomic his­tory, as it hap­pened pre­vi­ously in the runup to the Great Depres­sion and numer­ous finan­cial crises before­hand. In its after­math, we are now expe­ri­enc­ing a “credit crunch”—a sud­den rever­sal with the cav­a­liers going from being will­ing to lend to vir­tu­ally any­one with a pulse, to refus­ing credit even to those with solid finan­cial histories.

I intro­duce a “credit crunch” into this model by chang­ing those same key key finan­cial para­me­ters at the 30 year mark, but decreas­ing them rather than increas­ing them. Firms go from hav­ing a 20 year hori­zon for debt repay­ment to a 6.4 year hori­zon, banks go from increas­ing the money sup­ply at 10% per annum to 3.2% per annum, while the rate of cir­cu­la­tion of unlent reserves drops by 68%.

There is much more to our cur­rent cri­sis than this—in par­tic­u­lar, this model omits “Ponzi lend­ing” that finances gam­bling on asset prices rather than pro­duc­tive invest­ment, and the result­ing accu­mu­la­tion of debt com­pared to GDP—but this level of change in finan­cial para­me­ters alone is suf­fi­cient to cause a sim­u­lated cri­sis equiv­a­lent to the Great Depres­sion. Its behav­iour repro­duces much of what we’re wit­ness­ing now: there is a sud­den blowout in unlent reserves, and a decline in the nom­i­nal level of debt and in the amount of money cir­cu­lat­ing in the economy.

This is the real world phe­nom­e­non that Bernanke is now rail­ing against with his increases in Base Money, and already the wide­spread lament amongst pol­icy mak­ers is that banks are not lend­ing out this addi­tional money, but sim­ply build­ing up their reserves.

Tough: in a credit econ­omy, that’s what banks do after a finan­cial crisis—it’s what they did dur­ing the Great Depres­sion. This credit-economy phe­nom­e­non is the real rea­son that the money sup­ply dropped dur­ing the Depres­sion: it wasn’t due to “bad Fed­eral Reserve pol­icy” as Bernanke him­self has opined, but due to the fact that we live in a credit money world, and not the fiat money fig­ment of neo­clas­si­cal imagination.

The impact of the sim­u­lated credit crunch on my toy economy’s real vari­ables is sim­i­lar to that of the Great Depres­sion: real out­put slumps severely, as does employment.

The nom­i­nal value of out­put also falls, because prices also fall along with real output.

This fall in prices is dri­ven by a switch from a regime of grow­ing demand to one of shrink­ing demand. Rather than there being a con­tin­u­ous slight imbal­ance in demand’s favour, the imbal­ance shifts in favour of supply—and prices con­tinue falling even though out­put even­tu­ally starts to rise.

The unem­ploy­ment rate explodes rapidly from full employ­ment to 25 per­cent of the work­force being out of a job—and then begins a slow recovery.

Finally, wages behave in a per­verse fash­ion, just as Keynes argued dur­ing the Great Depres­sion: nom­i­nal wages fall, but real wages rise because the fall in prices out­runs the fall in wages.

This com­bi­na­tion of falling prices and falling out­put means that despite the fall in nom­i­nal debts, the ratio of debt to nom­i­nal out­put actu­ally rises—again, as hap­pened for the first few years of the Great Depression.

Though this model is still sim­ple com­pared to the econ­omy in which we live, it’s a lot closer to our actual econ­omy than the mod­els devel­oped by con­ven­tional “neo­clas­si­cal” econ­o­mists, which ignore money and debt, and pre­sume that the econ­omy will always con­verge to a “NAIRU[14] equi­lib­rium after any shock.

It also shows the impor­tance of the nom­i­nal money stock, some­thing that neo­clas­si­cal econ­o­mists com­pletely ignore. To quote Mil­ton Fried­man on this point:

It is a com­mon­place of mon­e­tary the­ory that noth­ing is so unim­por­tant as the quan­tity of money expressed in terms of the nom­i­nal mon­e­tary unit—dollars, or pounds, or pesos… Let the num­ber of dol­lars in exis­tence be mul­ti­plied by 100; that, too, will have no other essen­tial effect, pro­vided that all other nom­i­nal mag­ni­tudes (prices of goods and ser­vices, and quan­ti­ties of other assets and lia­bil­i­ties that are expressed in nom­i­nal terms) are also mul­ti­plied by 100.” [15]

The mad­ness in Friedman’s argu­ment is the assump­tion that increas­ing the money sup­ply by a fac­tor of 100 will also cause “all other nom­i­nal mag­ni­tudes” includ­ing com­mod­ity prices and debts to be mul­ti­plied by the same factor.

What­ever might be the impact on prices of increas­ing the money sup­ply by a fac­tor of 100, the nom­i­nal value of debt would remain con­stant: debt con­tracts don’t give banks the right to increase your out­stand­ing level of debt just because prices have changed. Move­ments in the nom­i­nal prices of goods and ser­vices aren’t per­fectly mir­rored by changes in the level of nom­i­nal debts, and this is why nom­i­nal mag­ni­tudes can’t be ignored.

In this model I have devel­oped, money and its rate of cir­cu­la­tion mat­ter because they deter­mine the level of nom­i­nal and real demand. It is a “New Mon­e­tarism” model, in which money is crucial.

Iron­i­cally, Mil­ton Fried­man argued that money was cru­cial in his inter­pre­ta­tion of the Great Depression—that the fail­ure of the Fed­eral Reserve to suf­fi­ciently increase the money sup­ply allowed defla­tion to occur. But he a triv­ial “heli­copter” model of money cre­ation that saw all money as orig­i­nat­ing from the oper­a­tions of the Fed­eral Reserve:

Let us sup­pose now that one day a heli­copter flies over this com­mu­nity and drops an addi­tional $1,000 in bills from the sky, which is, of course, hastily col­lected by mem­bers of the com­mu­nity. Let us sup­pose fur­ther that every­one is con­vinced that this is a unique event which will never be repeated… [16]

When the heli­copter starts drop­ping money in a steady stream— or, more gen­er­ally, when the quan­tity of money starts unex­pect­edly to rise more rapidly— it takes time for peo­ple to catch on to what is hap­pen­ing. Ini­tially, they let actual bal­ances exceed long— run desired bal­ances…” (p. 13)

and a triv­ial model of the real econ­omy that argued that it always tended back to equilibrium:

Let us start with a sta­tion­ary soci­ety in which … (5) The soci­ety, though sta­tion­ary, is not sta­tic. Aggre­gates are con­stant, but indi­vid­u­als are sub­ject to uncer­tainty and change. Even the aggre­gates may change in a sto­chas­tic way, pro­vided the mean val­ues do not… Let us sup­pose that these con­di­tions have been in exis­tence long enough for the soci­ety to have reached a state of equi­lib­rium…” (pp. 2–3)

One nat­ural ques­tion to ask about this final sit­u­a­tion is, “ What raises the price level, if at all points mar­kets are cleared and real mag­ni­tudes are sta­ble?” The answer is, “ Because every­one con­fi­dently antic­i­pates that prices will rise.” (p. 10)

Using this sim­plis­tic analy­sis, Mil­ton Fried­man claimed that infla­tion was caused by “too many heli­copters” and defla­tion by “too few”, and that the defla­tion that ampli­fied the down­turn in the 1930s could have been pre­vented if only the Fed had sent more heli­copters into the fray:

dif­fer­ent and fea­si­ble actions by the mon­e­tary author­i­ties could have pre­vented the decline in the money stock—indeed, pro­duced almost any desired increase in the money stock. The same actions would also have eased the bank­ing dif­fi­cul­ties appre­cia­bly. Pre­ven­tion or mod­er­a­tion of the decline in the stock of money, let alone the sub­sti­tu­tion of mon­e­tary expan­sion, would have reduced the contraction’s sever­ity and almost as cer­tainly its dura­tion.” [17]

With a sen­si­ble model of how money is endoge­nously cre­ated by the finan­cial sys­tem, it is pos­si­ble to con­cur that a decline in money con­tributed to the sever­ity of the Great Depres­sion, but not to blame that on the Fed­eral Reserve not prop­erly exer­cis­ing its effec­tively impo­tent pow­ers of fiat money cre­ation. Instead, the decline was due to the nor­mal oper­a­tions of a credit money sys­tem dur­ing a finan­cial cri­sis that its own reck­less lend­ing has caused—the Cav­a­liers are cow­ards who rush into a bat­tle they are win­ning, and retreat at haste in defeat.

How­ever, with his belief in Friedman’s analy­sis, Bernanke did blame his 1930 pre­de­ces­sors for caus­ing the Great Depres­sion. In his paean to Mil­ton Fried­man on the occa­sion of his 90th birth­day, Bernanke made the fol­low­ing remark:

Let me end my talk by abus­ing slightly my sta­tus as an offi­cial rep­re­sen­ta­tive of the Fed­eral Reserve. I would like to say to Mil­ton and Anna: Regard­ing the Great Depres­sion. You’re right, we did it. We’re very sorry. But thanks to you, we won’t do it again.” [18]

In fact, thanks to Mil­ton Fried­man and neo­clas­si­cal eco­nom­ics in gen­eral, the Fed ignored the run up of debt that has caused this cri­sis, and every res­cue engi­neered by the Fed sim­ply increased the height of the precipice from which the even­tual fall into Depres­sion would occur.

Hav­ing failed to under­stand the mech­a­nism of money cre­ation in a credit money world, and failed to under­stand how that mech­a­nism goes into reverse dur­ing a finan­cial cri­sis, neo­clas­si­cal eco­nom­ics may end up doing what by acci­dent what Marx failed to achieve by delib­er­ate action, and bring cap­i­tal­ism to its knees.

Neo­clas­si­cal economics—and espe­cially that derived from Mil­ton Friedman’s pen—is mad, bad, and dan­ger­ous to know.

Debt­watch Sta­tis­tics Feb­ru­ary 2009

My dis­cus­sion of the most recent monthly data is abbre­vi­ated given the length of this Report, but it now appears that the debt bub­ble has started to burst. Pri­vate debt fell by $A$5 bil­lion in the last month, the first fall since 2003, and the steep­est monthly fall on record.

As a result, Australia’s Debt to GDP ratio has started to fall.

How­ever, it might rise once more if defla­tion takes hold. This was the Depres­sion expe­ri­ence when the debt to GDP ratio rose even as nom­i­nal debt lev­els fell. Leav­ing that pos­si­bil­ity aside for the moment, it appears that Australia’s peak pri­vate debt to GDP ratio occurred in March 2008, with a ratio of 177% of GDP includ­ing busi­ness secu­ri­ties (or 165% exclud­ing busi­ness securities).


[1] Marx, Cap­i­tal Vol­ume III, Chap­ter 33, The medium of cir­cu­la­tion in the credit sys­tem, pp. 544–45 [Progress Press]. http://www.marx.org/archive/marx/works/1894-c3/ch33.htm. Emphases added.

[2] Notably the “labour the­ory of value”, which argues erro­neously that all profit comes from labour, the notion that the rate of profit has a ten­dency to fall, and the alleged inevitabil­ity of the demise of cap­i­tal­ism; see my papers on these issues on the Research page of my blog under Marx.

[3] Kyd­land & Prescott, Busi­ness Cycles: Real Facts and a Mon­e­tary Myth, Fed­eral Reserve Bank of Min­neapo­lis Quar­terly Review, Spring 1990.

[4] “The Endoge­nous Money Stock”, Jour­nal of Post Key­ne­sian Eco­nom­ics, 1979, Vol­ume 2, pp. 49–70.

[5] Basil Moore 1983, “Unpack­ing the post Key­ne­sian black box: bank lend­ing and the money sup­ply”, Jour­nal of Post Key­ne­sian Eco­nom­ics 1983, Vol. 4 pp. 537–556; here Moore was quot­ing a Fed­eral Reserve econ­o­mist from a 1969 con­fer­ence in which the endo­gene­ity of the money sup­ply was being debated.

[6] This pol­icy “worked” in the sense that Cen­tral Banks were suc­cess­ful in con­trol­ling short run inter­est rates, and appeared to work in con­trol­ling infla­tion; but it is now becom­ing obvi­ous that its suc­cess on the lat­ter front was a coincidence—the era of low infla­tion coin­cided with the dra­matic impact of China and off­shore man­u­fac­tur­ing in gen­eral on con­sumer and pro­ducer prices—and it led to Cen­tral Banks com­pletely ignor­ing the debt bub­ble that has caused the global finan­cial cri­sis. As a result, inter­est rate tar­get­ting is also going the way of the Dodo now.

[7] see Table 10 in Yueh-Yun C. OBrien, 2007. “Reserve Require­ment Sys­tems in OECD Coun­tries” , Finance and Eco­nom­ics Dis­cus­sion Series, Divi­sions of Research & Sta­tis­tics and Mon­e­tary A¤airs, Fed­eral Reserve Board, 2007–54, Wash­ing­ton, D.C;. The US rule implies that the main rea­son for the “reserve require­ment” these days is to meet house­hold demand for cash.

[8] Bernanke 2002: Remarks by Gov­er­nor Ben S. Bernanke Before the National Econ­o­mists Club, Wash­ing­ton, D.C., Novem­ber 21, 2002. Defla­tion: Mak­ing Sure “It” Doesn’t Hap­pen Here. Empha­sis added.

[9] Microsoft resorts to first lay­offs, cut­ting 5,000″, Yahoo Finance Jan­u­ary 22nd 2009.

[10] And, for that mat­ter, by Aus­trian eco­nom­ics, whose analy­sis of money is sur­pris­ingly sim­plis­tic. Though Aus­tri­ans advo­cate a pri­vate money sys­tem in which banks would issue their own cur­rency, they assume that under the cur­rent money sys­tem, all money is gen­er­ated by frac­tional reserve lend­ing on top of fiat money cre­ation. This is strange, since if they advo­cate a pri­vate money sys­tem, they need a model of how banks could cre­ate money with­out frac­tional reserve lend­ing. But they don’t have one.

[11] Graziani A. (1989). The The­ory of the Mon­e­tary Cir­cuit, Thames Papers in Polit­i­cal Econ­omy, Sprin, pp.:1–26. Reprinted in M. Musella and C. Pan­ico (eds) (1995). The Money Sup­ply in the Eco­nomic Process, Edward Elgar, Aldershot.

[12] Keynes 1937, “ Alter­na­tive the­o­ries of the rate of inter­est” , Eco­nomic Jour­nal, Vol. 47, pp. 241–252: p. 247

[13] The para­me­ters were an ini­tial loan of $100, loan rate of inter­est of 5%, deposit rate of 1%, 3 month lag between financ­ing pro­duc­tion and receiv­ing the sales pro­ceeds, 1/3rd of the sur­plus from pro­duc­tion going to firms as prof­its and the remain­der to work­ers as wages, a one year lag in price adjust­ments, a money wage of $1, worker pro­duc­tiv­ity of 1.1 units of phys­i­cal out­put per worker per year, and a one year lag in spend­ing by bankers and a two week lag by workers.

[14] “Non-Accelerating Infla­tion Rate of Unem­ploy­ment”, another one of Milton’s myth­i­cal constants.

[15] Mil­ton Fried­man 1969, “The Opti­mal Quan­tity of Money”, in The Opti­mal Quan­tity of Money and Other Essays, Macmil­lan, Chicago, p. 1.

[16] Mil­ton Fried­man 1969, pp. 5–6

[17] Mil­ton Fried­man and Anna Schwartz 1963, A Mon­e­tary His­tory of the United States 1867–1960, Prince­ton Uni­ver­sity Press, Prince­ton, p. 301.

[18] Remarks by Gov­er­nor Ben S. Bernanke At the Con­fer­ence to Honor Mil­ton Fried­man, Uni­ver­sity of Chicago, Chicago, Illi­nois Novem­ber 8, 2002 On Mil­ton Friedman’s Nineti­eth Birth­day.

About Steve Keen

I am a professional economist and a long time critic of conventional economic thought. As well as attacking mainstream thought in Debunking Economics, I am also developing an alternative dynamic approach to economic modelling. The key issue I am tackling here is the prospect for a debt-deflation on the back of the enormous debts accumulated in Australia, and our very low rate of inflation.
Bookmark the permalink.