Submission to the Senate Economics Committee Post-GFC Banking Inquiry

Flattr this!

In a very time­ly move, the Aus­tralian Sen­ate Stand­ing Com­mit­tee on Eco­nom­ics estab­lished a hear­ing into the post-GFC bank­ing sec­tor. I was pleased to be invit­ed to make a sub­mis­sion, which I com­plet­ed just before leav­ing Aus­tralia for one mon­th’s research into mon­e­tary eco­nom­ics with math­e­mati­cians at the Fields Insti­tute in Toron­to.

The sub­mis­sions have just been made pub­lic. Mine is avail­able here in PDF, and is also repro­duced below. I have improved on my mod­el­ing of mon­ey cre­ation in the last two weeks at the Fields, but I’ll leave details of that for a lat­er post.

Pro­fes­sor Steve Keen, Uni­ver­si­ty of West­ern Syd­ney

This is a long sub­mis­sion, and I apol­o­gise in advance for its length and com­plex­i­ty. How­ev­er these are com­plex times, and since my analy­sis and rec­om­men­da­tions are well out­side the main­stream of eco­nom­ic advice, I need­ed to cov­er them in detail.


The eco­nom­ic cri­sis occurred because of a fail­ure of both eco­nom­ic the­o­ry and eco­nom­ic man­age­ment. Eco­nom­ic the­o­ry is dom­i­nat­ed by the “Neo­clas­si­cal” school of thought, and this school’s under­stand­ing of banks, mon­ey and debt is seri­ous­ly defi­cient. Attempts to con­trol the macro­econ­o­my have been based upon this the­o­ry, and have there­fore failed. Poli­cies to con­trol the bank­ing sys­tem need to be based on a real­is­tic mod­el of how it oper­ates, and this is the mod­el of endoge­nous mon­ey.

This mod­el, which is strong­ly sup­port­ed by empir­i­cal research, argues that bank lend­ing (a) is not con­strained by the Cen­tral Bank, (b) adds to aggre­gate demand—so that change in the lev­el of pri­vate debt has impor­tant macro­eco­nom­ic con­se­quences, and © can cause asset bub­bles and finan­cial crises when that lend­ing pri­mar­i­ly finances spec­u­la­tion on asset prices rather than invest­ment.

There are three indi­ca­tors that a finan­cial cri­sis could occur: the lev­el, rate of change and rate of accel­er­a­tion of pri­vate debt. The RBA must mon­i­tor these indi­ca­tors as well as its cur­rent tar­gets of the rate of infla­tion and the rate of unem­ploy­ment.

Mon­e­tary and fis­cal pol­i­cy alone are unable to restrain the ten­den­cy that the bank­ing sec­tor has to move from respon­si­ble to irre­spon­si­ble lend­ing over time. The only effec­tive way to pre­vent asset bub­bles and finan­cial crises in the future is to make lev­ered spec­u­la­tion on asset prices less attrac­tive to the pub­lic. I pro­pose redefin­ing shares in such a way that the sec­ondary mar­ket in shares is great­ly dimin­ished while the pri­ma­ry mar­ket that actu­al­ly rais­es cap­i­tal for firms is enhanced (“Jubilee Shares”), and lim­it­ing lever­age in prop­er­ty pur­chas­es on the basis of the antic­i­pat­ed income from the asset being pur­chased (“The PILL”).

Introduction: How did economists get it so wrong?

Per­haps the most strik­ing fea­ture of the Glob­al Finan­cial Cri­sis (GFC) was the con­fi­dence that all lead­ing econ­o­mists, offi­cial eco­nom­ic advi­sors and fore­cast­ers had in the state of both eco­nom­ic the­o­ry and the glob­al econ­o­my imme­di­ate­ly pri­or to the cri­sis.

The then Pres­i­dent of the Amer­i­can Eco­nom­ic Asso­ci­a­tion, Nobel Lau­re­ate Robert Lucas, was con­fi­dent that the econ­o­my would nev­er again suf­fer from a cri­sis like the Great Depres­sion, because mod­ern macro­eco­nom­ic knowl­edge knew how to pre­vent such calami­ties:

Macro­eco­nom­ics was born as a dis­tinct field in the 1940’s, as a part of the intel­lec­tu­al response to the Great Depres­sion. The term then referred to the body of knowl­edge and exper­tise that we hoped would pre­vent the recur­rence of that eco­nom­ic dis­as­ter. My the­sis in this lec­ture is that macro­eco­nom­ics in this orig­i­nal sense has suc­ceed­ed: Its cen­tral prob­lem of depres­sion pre­ven­tion has been solved, for all prac­ti­cal pur­pos­es, and has in fact been solved for many decades.’ (Lucas 2003 , p. 1, empha­sis added)

Writ­ing a year after the cri­sis began, Olivi­er Blan­chard, the Chief Econ­o­mist of the IMF and the found­ing edi­tor of the Amer­i­can Eco­nom­ic Review: Macro, assert­ed that macro­eco­nom­ic the­o­ry was set­tled and well-ground­ed:

there has been enor­mous progress and sub­stan­tial con­ver­gence… large­ly because facts have a way of not going away, a large­ly shared vision both of fluc­tu­a­tions and of method­ol­o­gy has emerged…The state of macro is good. (Blan­chard 2009, p. 210; empha­sis added).

Now Fed­er­al Reserve Gov­er­nor Ben Bernanke was con­vinced that the reduc­tion in eco­nom­ic volatil­i­ty in the two decades pri­or to the cri­sis was an endur­ing fea­ture of the econ­o­my, for which Cen­tral Bankers—those in charge of mon­e­tary policy—could take cred­it:

The sources of the Great Mod­er­a­tion remain some­what con­tro­ver­sial, but as I have argued else­where, there is evi­dence for the view that improved con­trol of infla­tion has con­tributed in impor­tant mea­sure to this wel­come change in the econ­o­my. (Bernanke 2004, empha­sis added).

The OECD, the the world’s pre­miere eco­nom­ic fore­cast­ing organ­i­sa­tion, fore­cast In June 2007 that the out­look for the glob­al econ­o­my was “quite benign”:

Recent devel­op­ments have broad­ly con­firmed this prog­no­sis. Indeed, the cur­rent eco­nom­ic sit­u­a­tion is in many ways bet­ter than what we have expe­ri­enced in years. Against that back­ground, we have stuck to the rebal­anc­ing sce­nario. Our cen­tral fore­cast remains indeed quite benign: a soft land­ing in the Unit­ed States, a strong and sus­tained recov­ery in Europe, a sol­id tra­jec­to­ry in Japan and buoy­ant activ­i­ty in Chi­na and India. In line with recent trends, sus­tained growth in OECD economies would be under­pinned by strong job cre­ation and falling unem­ploy­ment. (Cotis 2007, p. 7, empha­sis added)

On August 12, 2007, less than 2 months after this rosy fore­cast was pub­lished, the BNP shut down 3 of its funds that were heav­i­ly exposed to the US sub­prime mort­gage mar­ket, and what Aus­tralians now call the GFC began. Two years lat­er, US unem­ploy­ment peaked at 10.2% of the work­force, which was only not a post-WWII record because of changes to the def­i­n­i­tion of unem­ploy­ment after 1990.

Fig­ure 1: Unem­ploy­ment rates before and after the GFC

There are only 2 expla­na­tions for this huge diver­gence between the expec­ta­tions and fore­casts of lead­ing econ­o­mists and the actu­al out­come of the worst eco­nom­ic cri­sis since the Great Depres­sion. Either (a) the cri­sis was a com­plete­ly unpre­dictable, ran­dom event; or (b) there was some­thing seri­ous­ly at fault with the the­o­ries and mod­els that offi­cial eco­nom­ic advi­sors and fore­cast­ers used to analyse the econ­o­my.

After the cri­sis, lead­ing econ­o­mists have defend­ed option (a), thus reject­ing the alter­na­tive expla­na­tion that the fail­ure to antic­i­pate such a huge event indi­cat­ed that there was some­thing wrong with their analy­sis. RBA Gov­er­nor Glenn Stevens described the GFC as a “tail event”, and huge ran­dom shock which had enor­mous con­se­quences but was inher­ent­ly impos­si­ble to pre­dict:

I do not know any­one who pre­dict­ed this course of events. This should give us cause to reflect on how hard a job it is to make gen­uine­ly use­ful fore­casts. What we have seen is tru­ly a ‘tail’ outcome—the kind of out­come that the rou­tine fore­cast­ing process nev­er pre­dicts. But it has occurred, it has impli­ca­tions, and so we must reflect on it.(Stevens 2008, p. 7)

Ben Bernanke assert­ed that eco­nom­ic the­o­ry itself was not at fault—only the imple­men­ta­tion of it was to blame:

Eco­nom­ic sci­ence con­cerns itself pri­mar­i­ly with the­o­ret­i­cal and empir­i­cal gen­er­al­iza­tions about the behav­ior of indi­vid­u­als, insti­tu­tions, mar­kets, and nation­al economies. Most aca­d­e­m­ic research falls in this cat­e­go­ry. Eco­nom­ic engi­neer­ing is about the design and analy­sis of frame­works for achiev­ing spe­cif­ic eco­nom­ic objec­tives… With that tax­on­o­my in hand, I would argue that the recent finan­cial cri­sis was more a fail­ure of eco­nom­ic engi­neer­ing and eco­nom­ic man­age­ment than of what I have called eco­nom­ic sci­ence. The eco­nom­ic engi­neer­ing prob­lems were reflect­ed in a num­ber of struc­tur­al weak­ness­es in our finan­cial sys­tem. In the pri­vate sec­tor, these weak­ness­es includ­ed inad­e­quate risk-mea­sure­ment and risk-man­age­ment sys­tems at many finan­cial firms as well as short­com­ings in some firms’ busi­ness mod­els, such as over­re­liance on unsta­ble short-term fund­ing and exces­sive lever­age. In the pub­lic sec­tor, gaps and blind spots in the finan­cial reg­u­la­to­ry struc­tures of the Unit­ed States and most oth­er coun­tries proved par­tic­u­lar­ly dam­ag­ing. (Bernanke 2010)

The impli­ca­tion is that the cri­sis was due to the fail­ure of admin­is­tra­tive and polit­i­cal insti­tu­tions to ful­ly imple­ment the guid­ance giv­en by eco­nom­ic the­o­ry, and there­fore that a future cri­sis can be pre­vent­ed sim­ply by ensur­ing that the guid­ance of eco­nom­ic the­o­ry is more close­ly fol­lowed in the future.

Aus­trali­a’s appar­ent immu­ni­ty to this cri­sis is also tak­en as an indi­ca­tor that our admin­is­tra­tive and polit­i­cal insti­tu­tions did a bet­ter job of imple­ment­ing the con­ven­tion­al and cor­rect wis­dom than their coun­ter­parts in the North Atlantic.

I com­plete­ly reject this con­ven­tion­al wis­dom.

Unpredictable shock?

The longevi­ty of this cri­sis is already wear­ing the thin the argu­ment that the GFC was caused by a large, unpre­dictable neg­a­tive exoge­nous shock. If it were, the econ­o­my should be now have returned to trend growth—and it should also have rebound­ed as the neg­a­tive shocks were fol­lowed by pos­i­tive ones. Instead, the reduc­tions in unem­ploy­ment since the peak of the GFC have been ten­u­ous, with the rate of eco­nom­ic growth has in gen­er­al been below the 3 per­cent lev­el that the rule of thumb known as “Okun’s Law” indi­cates is need­ed to bal­ance ris­ing pop­u­la­tion and labour pro­duc­tiv­i­ty. This is true even for Aus­tralia. Despite its appar­ent avoid­ance of the worst of the GFC, and the min­er­als export boom, growth has con­sis­tent­ly been below 3 per cent (see Fig­ure 2).

Fig­ure 2: Real growth post-GFC has rarely exceed­ed lev­el need­ed to reduce unem­ploy­ment

 “Economic Science” is not OK

Eco­nom­ic sci­ence”, as Bernanke describes it, allowed the cause of this crisis—the exces­sive growth of pri­vate debt—to go on for decades before the cri­sis erupt­ed in 2008. To any lay observ­er, the fact that debt peaks and then rapid declines have coin­cid­ed with two past Depres­sions and our cur­rent cri­sis is strong evi­dence that pri­vate debt bub­bles cause eco­nom­ic crises—the sort of evi­dence one would expect pro­fes­sion­al econ­o­mists to inves­ti­gate (see Fig­ure 1). Yet instead Nobel Prize win­ning econ­o­mists and Aus­tralian reg­u­la­tors dis­miss the impor­tance of this data on pure­ly the­o­ret­i­cal grounds.

Fig­ure 3: Pri­vate debt to GDP ratios from 1860 till today

In his most recent book End This Depres­sion Now!, Nobel Lau­re­ate Paul Krug­man asserts that “debt is mon­ey we owe to our­selves”, and there­fore the absolute lev­el of debt is unimportant—all that mat­ters is its dis­tri­b­u­tion:

It’s true that peo­ple like me believe that the depres­sion we’re in was in large part caused by the buildup of house­hold debt, which set the stage for a Minksy moment in which high­ly indebt­ed house­holds were forced to slash their spend­ing. How, then, can even more debt be part of the appro­pri­ate pol­i­cy response?

The key point is that this argu­ment against deficit spend­ing assumes, implic­it­ly, that debt is debt—that it does­n’t mat­ter who owes the mon­ey. Yet that can’t be right; if it were, we would­n’t have a prob­lem in the first place…. Ignor­ing the for­eign com­po­nent, or look­ing at the world as a whole, we see that the over­all lev­el of debt makes no dif­fer­ence to aggre­gate net worth—one per­son­’s lia­bil­i­ty is anoth­er per­son­’s asset.

It fol­lows that the lev­el of debt mat­ters only if the dis­tri­b­u­tion of net worth mat­ters, if high­ly indebt­ed play­ers face dif­fer­ent con­straints from play­ers with low debt. And this means that all debt isn’t cre­at­ed equal, which is why bor­row­ing by some actors now can help cure prob­lems cre­at­ed by excess bor­row­ing by oth­er actors in the past.

Think of it this way: when debt is ris­ing, it’s not the econ­o­my as a whole bor­row­ing more mon­ey. It is, rather, a case of less patient people—people who for what­ev­er rea­son want to spend soon­er rather than later—borrowing from more patient peo­ple. The main lim­it on this kind of bor­row­ing is the con­cern of those patient lenders about whether they will be repaid, which sets some kind of ceil­ing on each indi­vid­u­al’s abil­i­ty to bor­row.

What hap­pened in 2008 was a sud­den down­ward revi­sion of those ceil­ings… (Krug­man 2012, pp. 146–147)

The Head of the RBA’s Finan­cial Sta­bil­i­ty Depart­ment, Luci Ellis, dis­missed the idea of reg­u­lat­ing debt ratios:

In Aus­tralia, house­holds in aggre­gate used to have very lit­tle debt against their homes, rel­a­tive to the val­ue of those homes, back when the finan­cial sec­tor was high­ly reg­u­lat­ed and infla­tion erod­ed that debt quick­ly (Graph 6). Obvi­ous­ly this mea­sure of lever­age has risen since then. It would not be desir­able for this ratio to approach that in the Unit­ed States. How­ev­er, we do not think the most effec­tive way to pre­vent that would be to impose a cap just for new bor­row­ers.

Nor do we think it is sen­si­ble to rely on sim­ple rules like a ratio of loan amount to income; nowa­days many Aus­tralian lenders sen­si­bly take bor­row­ers’ oth­er oblig­a­tions and expens­es into account when deter­min­ing how much debt can be ser­viced, and thus how much they will lend. (Ellis 2012, pp. 7–8)

The the­o­reti­cian thus tells us that the lev­el of debt or its growth has no sig­nif­i­cant macro­eco­nom­ic impli­ca­tions. The prac­ti­tion­er tells us that we should leave deter­min­ing the opti­mal ratio of pri­vate debt to income to the lenders. Both these well-inten­tioned argu­ments are pro­found­ly wrong.

The real world of endogenous money

Krug­man’s error—which is com­mon to all main­stream “Neo­clas­si­cal” economists—is to ignore the role of banks in the econ­o­my, and treat lend­ing mon­ey as anal­o­gous to one neigh­bour lend­ing a lawn­mow­er to anoth­er. The loan of course makes no dif­fer­ence to the amount of grass that can be cut on a giv­en week­end: the bor­row­er’s capac­i­ty to mow ris­es but the lender’s capac­i­ty falls.

In fact, banks in a mar­ket econ­o­my are more anal­o­gous to lawn­mow­er fac­to­ries than to neigh­bours: just as lawn­mow­er fac­to­ries make lawn­mow­ers, and an increase in lawn­mow­er sales does increase the amount of grass that can be mowed, banks pro­duce mon­ey by cre­at­ing debt, and an increase in the pro­duc­tion of debt increas­es aggre­gate eco­nom­ic activ­i­ty.

The fact that main­stream econ­o­mists ignore this debt-and-mon­ey-cre­ation role of banks is why they did not see this cri­sis com­ing. Bizarre as it may sound, all accept­ed eco­nom­ic mod­els today ignore the exis­tence of banks. I know this claim sounds ludi­crous to non-econ­o­mists, so I have to thank Paul Krug­man for con­firm­ing this for me in a recent debate between us. In reac­tion to a con­fer­ence paper in which I claimed that banks play an essen­tial role in macro­eco­nom­ics, he stat­ed

In par­tic­u­lar, [Keen] asserts that putting banks in the sto­ry is essen­tial. Now, I’m all for includ­ing the bank­ing sec­tor in sto­ries where it’s rel­e­vant; but why is it so cru­cial to a sto­ry about debt and lever­age?

The rea­son that banks are cru­cial is because a bank loan cre­ates addi­tion­al spend­ing pow­er for the bor­row­er with­out reduc­ing the spend­ing pow­er of exist­ing savers. How­ev­er Krug­man’s mod­el of lending—which is com­mon to Neo­clas­si­cal economics—ignores lend­ing by banks to instead mod­el lend­ing as occur­ring between non-bank indi­vid­u­als. The patient lender can con­sume less because of the loan, the impa­tient bor­row­er can con­sume more, so that in the aggre­gate there can only be a triv­ial change to aggre­gate demand:

Keen then goes on to assert that lend­ing is, by def­i­n­i­tion (at least as I under­stand it), an addi­tion to aggre­gate demand. I guess I don’t get that at all. If I decide to cut back on my spend­ing and stash the funds in a bank, which lends them out to some­one else, this does­n’t have to rep­re­sent a net increase in demand. Yes, in some (many) cas­es lend­ing is asso­ci­at­ed with high­er demand, because resources are being trans­ferred to peo­ple with a high­er propen­si­ty to spend; but Keen seems to be say­ing some­thing else, and I’m not sure what. I think it has some­thing to do with the notion that cre­at­ing mon­ey = cre­at­ing demand, but again that isn’t right in any mod­el I under­stand. (Krug­man 2012)

Table 1 gives an exam­ple of this view of lend­ing, in which banks can be ignored (because lend­ing is just “a case of less patient people—people who for what­ev­er rea­son want to spend soon­er rather than later—borrowing from more patient peo­ple” (Krug­man 2012, p. 146)), fol­low­ing the account­ing con­ven­tion that an increase to a lia­bil­i­ty is shown with a neg­a­tive sign and an increase to an asset is shown with a pos­i­tive sign.

Table 1: Neo­clas­si­cal vision of lend­ing



Bank Assets”




Per­son­al Bal­ance Sheets

Not Rel­e­vant








Ini­tial Val­ue

























In this Neo­clas­si­cal vision of lend­ing, the econ­o­my starts with a giv­en stock of mon­ey (100 in this exam­ple), and lend­ing redis­trib­utes this between agents but does not increase it. Banks are just inter­me­di­aries in this process, and can be ignored—as they are in all main­stream eco­nom­ic mod­els.

In real­i­ty, and in the alter­na­tive macro­eco­nom­ic the­o­ry that I and oth­er non-Neo­clas­si­cal econ­o­mists are con­struct­ing, banks play a cru­cial role which the Neo­clas­si­cal “Loan­able Funds” view ignores. This is, to quote Schum­peter, “the cre­ation of new pur­chas­ing pow­er out of noth­ing…”:

It is always a ques­tion, not of trans­form­ing pur­chas­ing pow­er which already exists in some­one’s pos­ses­sion, but of the cre­ation of new pur­chas­ing pow­er out of noth­ing…’ (Schum­peter 1934, p. 73)

The state­ment that banks can cre­ate new pur­chas­ing capac­i­ty “out of noth­ing” is part­ly alle­gor­i­cal. It should be tak­en as in con­trast to a com­mod­i­ty, which requires oth­er com­modi­ties as inputs in order to be produced—and if you want more out­put, you have to have addi­tion­al inputs. In con­trast, mon­ey is not pro­duced by means of oth­er com­modi­ties, and there is no strict rela­tion­ship between the phys­i­cal resources a bank employs and the amount of mon­ey pro­duced. Unlike a man­u­fac­tur­ing firm which must have a phys­i­cal fac­to­ry to pro­duce out­put, the key asset that a bank needs in order to be able to cre­ate mon­ey is an intan­gi­ble one: a bank­ing licence. This alone enables it to cre­ate the pri­ma­ry asset from which it earns an income: loans. It nec­es­sar­i­ly also gives banks the capac­i­ty to cre­ate mon­ey by the process of dou­ble-entry book­keep­ing.

They can’t cre­ate cur­ren­cy of course, and their capac­i­ty to cre­ate mon­ey is depen­dent on prof­itabil­i­ty and find­ing will­ing bor­row­ers, but this is oth­er­wise a “licence to print mon­ey” which is not effec­tive­ly con­strained by the reserve oper­a­tions of Cen­tral Banks.

A bank’s capac­i­ty to cre­ate mon­ey can eas­i­ly be illus­trat­ed using dou­ble-entry book­keep­ing. A banks assets and lia­bil­i­ties always sum to zero by the rules of dou­ble-entry book­keep­ing, so when a bank com­mences operations—say with a licence worth $100 million—this asset is bal­anced by lia­bil­i­ties (deposits) which ini­tial­ly it owns, and which are record­ed as a neg­a­tive sum on its bal­ance sheet.

Mak­ing a loan to the non-bank pub­lic involves three steps:

  1. The bank loans from the lia­bil­i­ties it cur­rent­ly owns to a mem­ber of the pub­lic: this is shown as a pos­i­tive entry on its “Vault”—since it reduces the neg­a­tive sum cur­rent­ly record­ed there—and a neg­a­tive entry on the deposit account of the bor­row­er, since it increas­es this bank lia­bil­i­ty;
  2. The increase in its loan assets is record­ed as an addi­tion to the pos­i­tive sum of loans out­stand­ing and a cor­re­spond­ing fall in the resid­ual val­ue of its good­will;
  3. It can then restore its intan­gi­ble good­will asset to the orig­i­nal val­ue, and increase the lia­bil­i­ties it cur­rent­ly owns by the same amount.

All these oper­a­tions are shown in Table 2, with an exam­ple in which the bank has already made a sin­gle loan to a “patient” agent, and there­after lends to the “impa­tient” agent.

Table 2: Endoge­nous mon­ey vision of lend­ing







Assets & Lia­bil­i­ties








Ini­tial Val­ue





















Restore Good­will













Thus though lend­ing is a “zero sum game” as Krug­man points out (“one per­son­’s lia­bil­i­ty is anoth­er per­son­’s asset”), it has macro­eco­nom­ic con­se­quences because the amount of mon­ey in cir­cu­la­tion is equiv­a­lent to the bank­ing sec­tor’s lia­bil­i­ties to the non-bank pub­lic, which are the non-bank pub­lic’s deposits at banks. This sum has been increased by the loan, and hence it has added to aggre­gate demand (see the final col­umn in Table 2).

Both the loan itself and restor­ing the val­ue of its intan­gi­ble asset are at the bank’s dis­cre­tion, but if it exer­cis­es this dis­cre­tion then the lev­el of both bank assets and lia­bil­i­ties rise by the amount of the loan. This means that a bank can use an intan­gi­ble asset of con­stant val­ue to cre­ate, over time, an essen­tial­ly lim­it­less amount of mon­ey (see Fig­ure 4, which sim­u­lates a dynam­ic mod­el derived from Table 2; there are of course many oth­er oper­a­tions that a bank exe­cutes, but none of them can­cel out this effect).

Fig­ure 4: Fixed intan­gi­ble asset (bank­ing licence) can sup­port grow­ing loans & deposits over time

In and of itself, this is no bad thing. The qual­i­ta­tive as well as quan­ti­ta­tive devel­op­ment than pro­duc­tion has under­gone since cap­i­tal­ism became the plan­et’s dom­i­nant social sys­tem in the 18th cen­tu­ry has depend­ed on this gen­er­a­tion of new spend­ing pow­er, which as Schum­peter argued long ago, is the major source of fund­ing for true entre­pre­neurs. Unlike con­ven­tion­al Neo­clas­si­cal econ­o­mists, whose use of assump­tions is best sum­ma­rized by the old “Let’s assume we have a can open­er” joke, Schum­peter explained the growth of cap­i­tal­ism by assum­ing that entre­pre­neurs did not, in gen­er­al, come from exist­ing funds with retained earn­ings, but were essen­tial­ly pen­ni­less. Essen­tial­ly, this makes it hard­er to explain how inno­va­tion occurs. Schum­peter then rea­soned that, to be able to put their ideas into action, entre­pre­neurs had to bor­row mon­ey:

the entre­pre­neur … can only become an entre­pre­neur by pre­vi­ous­ly becom­ing a debtor… his becom­ing a debtor aris­es from the neces­si­ty of the case and is not some­thing abnor­mal, an acci­den­tal event to be explained by par­tic­u­lar cir­cum­stances. What he first wants is cred­it. Before he requires any goods what­ev­er, he requires pur­chas­ing pow­er. He is the typ­i­cal debtor in cap­i­tal­ist soci­ety.’ (Schum­peter 1934, p. 102)

In turn, Schum­peter rea­soned that if lend­ing sim­ply was trans­fer­ring exist­ing spend­ing pow­er from “patient” to “impa­tient” agents, the result­ing drop in demand from savers would counter some of the impe­tus to invest in the first place. If that was the only way for entre­pre­neurs to get mon­ey, then the process would be some­what self-defeat­ing, and progress in cap­i­tal­ism would be very slow. But he knew that the con­ven­tion­al belief that banks sim­ply act as inter­me­di­aries between savers and bor­row­ers was false, because it ignored the capac­i­ty for banks to endoge­nous­ly cre­ate new spend­ing pow­er:

Even though the con­ven­tion­al answer to our ques­tion is not obvi­ous­ly absurd, yet there is anoth­er method of obtain­ing mon­ey for this pur­pose, which … does not pre­sup­pose the exis­tence of accu­mu­lat­ed results of pre­vi­ous devel­op­ment, and hence may be con­sid­ered as the only one which is avail­able in strict log­ic. This method of obtain­ing mon­ey is the cre­ation of pur­chas­ing pow­er by banks… It is always a ques­tion, not of trans­form­ing pur­chas­ing pow­er which already exists in some­one’s pos­ses­sion, but of the cre­ation of new pur­chas­ing pow­er out of noth­ing… (Schum­peter 1934, p. 73)

This the­o­ret­i­cal argu­ment received empir­i­cal sup­port from research by Fama and French. Using the Com­pu­s­tat data­base of com­pa­ny reports from pub­licly-trad­ed US non-finan­cial cor­po­ra­tions between 1951 & 1996, Fama and French cal­cu­lat­ed aggre­gate non-finan­cial cor­po­rate invest­ment, and cor­re­lat­ed it with equi­ty issue, retained earn­ings, and new debt (see Fig­ure 5).

Fig­ure 5: Cor­re­la­tions of invest­ment to new equi­ty, retained earn­ings and new debt (Fama & French 1999, p. 1954)

They con­clud­ed that “the source of financ­ing most cor­re­lat­ed with invest­ment is long-term debt”:

Fig­ure 3 shows invest­ment and its financ­ing year by year. The fig­ure sug­gests that new net issues of stock do not move close­ly with invest­ment. In fact, when the vari­ables are mea­sured rel­a­tive to book cap­i­tal … the cor­re­la­tion of invest­ment, It, and new net issues of stock, dSt, is only 0.19… retained cash earn­ings move more close­ly with invest­ment. The cor­re­la­tion between It and RCEt is indeed high­er, 0.56, but far from per­fect. The source of financ­ing most cor­re­lat­ed with invest­ment is long-term debt. The cor­re­la­tion between It and dLT­Dt is 0.79. The cor­re­la­tion between It and new short-term debt is low­er, 0.60, but non­triv­ial. These cor­re­la­tions con­firm the impres­sion from Fig­ure 3 that debt plays a key role in accom­mo­dat­ing year-by-year vari­a­tion in invest­ment. (Fama and French 1999, p. 1954)

These are the pos­i­tive aspects of endoge­nous mon­ey cre­ation by banks, but as we are painful­ly learn­ing all over again, there are also neg­a­tive aspects of this capac­i­ty.

Macroeconomics of endogenous money

The key con­clu­sions from the Neo­clas­si­cal vision of lend­ing is that the lev­el of pri­vate debt and its rate of change have no major macro­eco­nom­ic con­se­quences. This explains why Neo­clas­si­cal econ­o­mists not only ignore pri­vate debt, but also ignore expla­na­tions of eco­nom­ic crises in which change in the lev­el of debt play a cru­cial role. This is why Bernanke dis­missed Fish­er’s “Debt-Defla­tion The­o­ry of Great Depres­sions” (Fish­er 1933) (though Bernanke did devise a lim­it­ed way that changes in debt could affect macro­eco­nom­ics in what he termed the finan­cial accel­er­a­tor (Bernanke, Gertler et al. 1996)):

The idea of debt-defla­tion goes back to Irv­ing Fish­er (1933). Fish­er envi­sioned a dynam­ic process in which falling asset and com­mod­i­ty prices cre­at­ed pres­sure on nom­i­nal debtors, forc­ing them into dis­tress sales of assets, which in turn led to fur­ther price declines and finan­cial dif­fi­cul­ties. His diag­no­sis led him to urge Pres­i­dent Roo­sevelt to sub­or­di­nate exchange-rate con­sid­er­a­tions to the need for refla­tion, advice that (ulti­mate­ly) FDR fol­lowed.

Fish­er’s idea was less influ­en­tial in aca­d­e­m­ic cir­cles, though, because of the coun­ter­ar­gu­ment that debt-defla­tion rep­re­sent­ed no more than a redis­tri­b­u­tion from one group (debtors) to anoth­er (cred­i­tors). Absent implau­si­bly large dif­fer­ences in mar­gin­al spend­ing propen­si­ties among the groups, it was sug­gest­ed, pure redis­tri­b­u­tions should have no sig­nif­i­cant macro-eco­nom­ic effects…’ (Bernanke 2000, p. 24)

As I have shown, this Neo­clas­si­cal the­o­ry of lend­ing is wrong, and bank cre­ation of mon­ey adds to aggre­gate demand. Schum­peter put this case when explain­ing how this cre­ation of mon­ey funds entre­pre­neur­ial activ­i­ty:

The entre­pre­neur­ial func­tion is not, in prin­ci­ple, con­nect­ed with the pos­ses­sion of wealth … even though the acci­den­tal fact of the pos­ses­sion of wealth con­sti­tutes a prac­ti­cal advantage….the entre­pre­neur may mort­gage goods, which he acquires with the bor­rowed pur­chas­ing pow­er [but] The grant­i­ng of cred­it comes first and col­lat­er­al must be dis­pensed with, at least in prin­ci­ple, for how­ev­er short an inter­val. From this case the con­cep­tion of putting exist­ing assets into cir­cu­la­tion receives still less sup­port than from the first. On the con­trary it is per­fect­ly clear that pur­chas­ing pow­er is cre­at­ed to which in the first case no new goods cor­re­spond.

From this it fol­lows, there­fore, that in real life total cred­it must be greater than it could be if there were only ful­ly cov­ered cred­it. The cred­it struc­ture projects not only beyond the exist­ing gold basis, but also beyond the exist­ing com­mod­i­ty basis. (Schum­peter 1934, p. 101)

As Schum­peter explains, this leads to cycles—but not to the kind of cri­sis we are expe­ri­enc­ing today. These his stu­dent Hyman Min­sky explained as being large­ly due to bank cred­it financ­ing not just pro­duc­tive invest­ment but also unpro­duc­tive Ponzi spec­u­la­tion.

Min­sky began from the same per­spec­tive as did Schum­peter, that the endoge­nous cre­ation of spend­ing pow­er by banks means that aggre­gate demand is greater than demand from income alone:

If income is to grow, the finan­cial mar­kets, where the var­i­ous plans to save and invest are rec­on­ciled, must gen­er­ate an aggre­gate demand that, aside from brief inter­vals, is ever ris­ing. For real aggre­gate demand to be increas­ing, … it is nec­es­sary that cur­rent spend­ing plans, summed over all sec­tors, be greater than cur­rent received income and that some mar­ket tech­nique exist by which aggre­gate spend­ing in excess of aggre­gate antic­i­pat­ed income can be financed. It fol­lows that over a peri­od dur­ing which eco­nom­ic growth takes place, at least some sec­tors finance a part of their spend­ing by emit­ting debt or sell­ing assets. (Min­sky 1963; Min­sky 1982, p. 6; empha­sis added)

He then point­ed out that, as well as fund­ing invest­ment and entre­pre­neurs, banks also fund Ponzi Financiers: indi­vid­u­als whose cash flow from assets they own is less than their debt ser­vic­ing costs, but who can prof­it by sell­ing assets on a ris­ing mar­ket. Since they are actu­al­ly insol­vent between asset sales, they have an insa­tiable demand for more debt:

A Ponzi finance unit is a spec­u­la­tive financ­ing unit for which the income com­po­nent of the near term cash flows falls short of the near term inter­est pay­ments on debt so that for some time in the future the out­stand­ing debt will grow due to inter­est on exist­ing debt… Ponzi units can ful­fill their pay­ment com­mit­ments on debts only by bor­row­ing (or dis­pos­ing of assets)… a Ponzi unit must increase its out­stand­ing debts.’ (Min­sky 1982, p. 24)

These insights—that the change in debt funds invest­ment and speculation—mean that macro­eco­nom­ics is far more com­plex than the sim­ple Neo­clas­si­cal mod­el implies. Banks, debt and money—three aspects of real­i­ty that the Neo­clas­si­cal mod­el ignores—must there­fore be con­sid­ered if we are tru­ly to under­stand how the econ­o­my oper­ates.

Putting Schum­peter’s and Min­sky’s argu­ments more for­mal­ly, a start­ing point of this new macro­eco­nom­ics is that aggre­gate demand equals income plus the change in pri­vate debt, and aggre­gate sup­ply equals net expen­di­ture on goods and ser­vices and on finan­cial assets. This means that the change in debt has a strong impact upon both the lev­el of eco­nom­ic activ­i­ty and the lev­el of asset prices.

When we con­sid­er change in eco­nom­ic activ­i­ty, the pic­ture becomes more com­plex still. The change in aggre­gate demand is the sum of change in income plus the accel­er­a­tion of debt, and this will dri­ve both change in eco­nom­ic out­put and change in asset prices.

These rela­tion­ships between changes in pri­vate debt and the macro­econ­o­my and finance are eas­i­ly illus­trat­ed using data that Cen­tral Banks rou­tine­ly col­lect, but also rou­tine­ly ignore.

As Fig­ure 5 shows, the change in pri­vate debt is strong­ly cor­re­lat­ed with the lev­el of employ­ment: a rise in the rate of growth of debt caus­es a fall in unem­ploy­ment. The rel­a­tive­ly mild increase in unem­ploy­ment that Aus­tralia expe­ri­enced dur­ing the GFC, when com­pared the US expe­ri­ence, is large­ly explained by the fact that the change in debt did not go neg­a­tive in Aus­tralia dur­ing the GFC, where­as it did go neg­a­tive Amer­i­ca.

Fig­ure 6: Cor­re­la­tion of change in debt and unem­ploy­ment (Corr = ‑0.64)

Sim­i­lar­ly, Aus­trali­a’s record dur­ing the 1990s reces­sion was worse than the USA’s, since the change in debt did turn neg­a­tive in Aus­tralia dur­ing the 1990s reces­sion, but it did not do so in the USA.

Fig­ure 7: Cor­re­la­tion of change in debt and unem­ploy­ment (Corr = ‑0.75)

Fig­ure 8 and Fig­ure 9 show the rela­tion­ship between the accel­er­a­tion of debt and change in the lev­el of unem­ploy­ment in Aus­tralia and the USA. This indicator—which Neo­clas­si­cal econ­o­mists would have us ignore—clearly explains the sever­i­ty of the GFC, with the decel­er­a­tion of debt at that time being the sharpest in post-WWII his­to­ry (and in Amer­i­ca, the sharpest ever record­ed).

Fig­ure 8: Cor­re­la­tion of debt accel­er­a­tion with change in unem­ploy­ment in Aus­tralia (Corr = ‑0.65)

The dif­fer­ence between the two economies is large­ly a result of the decel­er­a­tion of debt in Aus­tralia being much low­er than it was for the USA. How­ev­er the Aus­tralian data also indi­cates that the boost our econ­o­my received from accel­er­at­ing debt after the GFC is now peter­ing out. The same trend is also evi­dent in the USA after its huge swing from decel­er­at­ing to accel­er­at­ing debt after the GFC (even though the change in debt is still neg­a­tive in the USA, debt is falling more slow­ly which trans­lates as an accel­er­a­tion of debt).

Fig­ure 9: Cor­re­la­tion of debt accel­er­a­tion with change in unem­ploy­ment in the USA (Corr = ‑0.68)

The accel­er­a­tion of pri­vate debt also plays a key role in dri­ving asset prices—and this rela­tion­ship under­pins my rec­om­men­da­tions for a post-GFC approach mon­e­tary pol­i­cy.

An eco­nom­ic the­o­ry which ignores these fac­tors will be a dan­ger­ous­ly mis­lead­ing guide as to how the econ­o­my actu­al­ly oper­ates. Unfor­tu­nate­ly, this the­o­ry not only dom­i­nates aca­d­e­m­ic eco­nom­ics, it is also the basis on which eco­nom­ic agen­cies like the Trea­sury and the RBA devise poli­cies to man­age the econ­o­my. This is why they did not see the cri­sis com­ing, why their ini­tial con­fi­dence that the cri­sis was behind us was mis­placed, why their advice about how to get out of it now is wrong, and why their rec­om­men­da­tions about how bank­ing should be man­aged in the future should be ignored.

Rely on bank prudence?

As not­ed ear­li­er, the Head of the RBA’s Finan­cial Sta­bil­i­ty Depart­ment, Luci Ellis, recent­ly rec­om­mend­ed against the idea that debt to income ratios should be either mon­i­tored or reg­u­lat­ed. Ellis’s error—which again is rep­re­sen­ta­tive of main­stream eco­nom­ic thinking—is to believe that well-man­aged banks can be relied upon to ensure that the aggre­gate lev­el of debt is not a prob­lem. This belief is a con­se­quence of the Neo­clas­si­cal the­o­ry that dom­i­nates RBA advice to government—advice that was and still is igno­rant of (a) the capac­i­ty of banks to cre­ate addi­tion­al aggre­gate demand by lend­ing, (b) the per­verse incen­tives that banks face which encour­age them to wish to cre­ate as much debt as they can per­suade the non-bank pub­lic to take on. It also ignore © that lend­ing can be not just for pro­duc­tive invest­ment or imme­di­ate con­sump­tion, but also for asset spec­u­la­tion, and (d) that asset-based lend­ing cre­ates a pos­i­tive feed­back between change in debt and the lev­el of asset prices which leads to crises like the one we are now in.

The fact that banks can add to aggre­gate demand by new lending—contrary to con­ven­tion­al theory—in turn leads to the dan­ger that unreg­u­lat­ed banks will suc­cumb to the per­verse incen­tives this gen­er­ates and cre­ate too much debt by financ­ing asset price spec­u­la­tion rather than pro­duc­tive invest­ment.

Perverse Incentives

The per­verse incen­tives that banks face are eas­i­ly illus­trat­ed by a sim­ple endoge­nous mon­ey mod­el in which banks have 3 ways to increase their income: by turn­ing over the cur­rent stock of mon­ey more rapid­ly; by per­suad­ing bor­row­ers to repay debt more slow­ly; and by cre­at­ing new debt by new lend­ing.

This sim­ple mod­el is derived from Table 3.

Table 3: Sim­ple mod­el to con­sid­er ways banks can increase their incomes

Assets Lia­bil­i­ties Equi­ty
Bank Deposits
Account Good­will Loan Vault Firm Work­er Share­hold­er Safe
Work­ing Cap­i­tal WC -WC
Record Loan -WC WC
Charge Inter­est Int -Int
Record Inter­est -Int Int
Pay Inter­est -Int Int
Record Pay­ment Int -Int
Wages Wage -Wage
Div­i­dends Div -Div
Con­sume -Con­sW Con­sW
-Con­sS Con­sS
-ConsB ConsB
Repay Loans -Repay Repay
Record Repay Repay -Repay
Invest­ment Loan Invest -Invest
Record Loan -Invest Invest
Restore Good­will Invest -Invest



The impacts of dou­bling turnover, dou­bling how long loans take to be repaid, and dou­bling the rate of cre­ation of new debt, are shown in Fig­ure 10. Clear­ly, by far the best way to increase bank income is to cre­ate more debt.

Fig­ure 10: Ways to increase bank income

The dif­fi­cul­ty for banks in achiev­ing this objec­tive lies in find­ing will­ing bor­row­ers. Here is the sec­ond, great per­verse incen­tive that leads to finan­cial crises: the best way to encour­age the pub­lic to take on more debt is to fund asset price spec­u­la­tion.

Funding Asset Speculation

When debt is tak­en on specif­i­cal­ly to finance either imme­di­ate con­sump­tion or gen­uine invest­ment, then in gen­er­al the pub­lic can be relied upon to lim­it the debt it takes on accord­ing to its capac­i­ty to pay. This con­trol is far from per­fect, since bor­row­ing relies upon expec­ta­tions about the future which will almost cer­tain­ly not be realised. But this “Schum­peter­ian” use of debt is like­ly to lead to cycles rather than a Depres­sion.

This is most eas­i­ly indi­cat­ed by com­par­ing bor­row­ing by house­holds for per­son­al con­sump­tion to house­hold bor­row­ing for asset pur­chas­es. Con­sid­ered on its own, per­son­al debt appears to have a strong upward trend, and to be high­ly volatile—see Fig­ure 11.

Fig­ure 11: Unse­cured per­son­al debt to GDP (with rede­f­i­n­i­tion of some busi­ness debt in 1989)

How­ev­er, when seen in the con­text of both mort­gage and busi­ness debt, per­son­al debt has been rel­a­tive­ly stable—see Fig­ure 12.

Fig­ure 12: Per­son­al, mort­gage and busi­ness debt to GDP ratios

House­holds were will­ing to take on more mort­gage debt rel­a­tive to income over time because they believed they were rid­ing a prop­er­ty bub­ble towards pros­per­i­ty. But in fact the main force dri­ving house prices high­er was the accel­er­a­tion of mort­gage debt.

The feedback trap between rising debt and asset prices

As explained in sec­tion 6, since the change in debt is a com­po­nent of aggre­gate demand, and aggre­gate demand is expend­ed on asset pur­chas­es as well as goods and ser­vices, then the accel­er­a­tion of debt is a com­po­nent of the change in aggre­gate demand, and this will play a role in deter­min­ing the direc­tion in which asset prices move.

This gen­er­ates what engi­neers call a pos­i­tive feed­back loop between change in debt and asset prices—“positive” not because it is a good thing, but because a change in the first fac­tor caus­es a change in the sec­ond in the same direc­tion, which thus caus­es a change in the first. Pos­i­tive feed­back loops inevitably lead to a break­down of the sys­tem in which they occur—whether this is an elec­tric cir­cuit, a bridge, or an economy—and much of engi­neer­ing is direct­ed at iden­ti­fy­ing pos­i­tive feed­back loops in equip­ment and elim­i­nat­ing them via intel­li­gent design.

The main fac­tor which has dri­ven house price bub­bles around the world is the pos­i­tive feed­back between change in mort­gage debt and house prices. Ris­ing mort­gage debt caused house prices to rise, and the rise in house prices encour­aged more house­holds to take on more mort­gage debt.

The process had to break down—and thus turn house price bub­bles into house price busts—because noth­ing can accel­er­ate for­ev­er: for mort­gage debt to con­tin­ue accel­er­at­ing indef­i­nite­ly, then ulti­mate­ly the ratio of mort­gage debt to income would be infi­nite. Long before this point is reached, mort­gage accel­er­a­tion will slow down: the increas­ing costs of entry deter new entrants, or the pool of avail­able entrant shrinks too much because of past price ris­es.

The same pos­i­tive feed­back process then works in reverse: falling house prices encour­age cur­rent mort­gagees to attempt to reduce their debt, and the decline in mort­gage debt caus­es house prices to fall. The unwind­ing process, how­ev­er, is not sym­met­ric: while a huge rise in debt occurs in the upswing, a huge fall in prices can result from only a mod­est fall in debt lev­els. Soci­ety is then stuck in a Depres­sion, with debt-ser­vic­ing and con­tin­u­ing debt-delever­ag­ing depress­ing aggre­gate demand as asset prices fall well below the lev­el at which the bub­ble began.

This process is now clear­ly evi­dent in US data, and is also becom­ing man­i­fest in Aus­trali­a’s data. In both coun­tries, the rela­tion­ship between mort­gage accel­er­a­tion and change in house prices is clear. The cor­re­la­tion between mort­gage accel­er­a­tion and house price change in Aus­tralia since 1992 is 0.59 (see Fig­ure 13), while the cor­re­la­tion in the USA from 1990 is 0.78 (see Fig­ure 14).


Fig­ure 13: Cor­re­la­tion of mort­gage accel­er­a­tion and house price change in Aus­tralia (Cor­re­la­tion = 0.59)

Fig­ure 14: Cor­re­la­tion of mort­gage accel­er­a­tion and house price change in the USA (Cor­re­la­tion = 0.78)

The bub­ble in Amer­i­can house prices has clear­ly burst, and prices have now fall­en to only 15 per­cent above the long term aver­age. But with the mort­gage debt over­hang as big as it is, the odds are that the fall in prices will take them well below the mean val­ue of 103 (see Fig­ure 15).

Fig­ure 15: Real house prices and mort­gage debt: USA

The USA is now six years into the burst­ing of its house price bub­ble. Aus­tralian house prices peaked in June 2010, and were 10 per cent below the peak by March 2012. We clear­ly have a long way to go.

Fig­ure 16: Real house prices and mort­gage debt: Aus­tralia

There are his­tor­i­cal rea­sons why the full peri­od from 1880 till now is not a good guide to the long term aver­age real house price for Aus­tralia (Sta­ple­don 2007), but even a com­par­i­son of cur­rent prices to the aver­age since 1975 implies that the fall in Aus­tralian house prices—under the impe­tus of decel­er­at­ing mort­gage debt—has a long way to go.

Fig­ure 17: Com­par­ing Aus­tralian and US real house prices

The Com­mit­tee should not take any com­fort in the “Aus­tralia is dif­fer­ent” that Aus­tralia does­n’t have a house price bub­ble, nor that the price declines will not have seri­ous macro­eco­nom­ic con­se­quences. Sim­i­lar assur­ances were accept­ed by the US Con­gress when prof­fered by US Fed­er­al Reserve Chair­man Alan Greenspan in August of 2005:

Although a “bub­ble” in home prices for the nation as a whole does not appear like­ly, there do appear to be, at a min­i­mum, signs of froth in some local mar­kets where home prices seem to have risen to unsus­tain­able lev­els… Although we cer­tain­ly can­not rule out home price declines, espe­cial­ly in some local mar­kets, these declines, were they to occur, like­ly would not have sub­stan­tial macro­eco­nom­ic impli­ca­tions. (Greenspan 2005)

Two years lat­er, Amer­i­ca entered its deep­est down­turn since the Great Depres­sion. Aus­tralian house prices are also falling at much the same rate as US prices did in the ear­ly years of their decline.

Fig­ure 18: Com­par­ing house price declines from peaks

As an aside, I also find Ms Ellis’s com­ment that “It would not be desir­able for this ratio [debt …, rel­a­tive to the val­ue of those homes] to approach that in the Unit­ed States” rather disin­gen­u­ous, since this ratio ignores the role that ris­ing debt has played in increas­ing house prices. The Amer­i­can ratio was also quite low pri­or to the cri­sis, and for the same rea­son: house prices had not yet crashed. Once they did, this ratio increased dra­mat­i­cal­ly, as I expect it will also in Aus­tralia. A far bet­ter indi­ca­tor of how our lever­age com­pares to the USA’s is a com­par­i­son of our mort­gage debt to GDP ratio to theirs, and on that basis, Aus­tralians are even more lever­aged than Amer­i­cans were at the peak of their bub­ble, and took on that lever­age much more quick­ly than did Amer­i­cans (see Fig­ure 19).

Fig­ure 19: Mort­gage debt to GDP in Aus­tralia and the USA

Rely on Central Bank control?

The pre­ced­ing sec­tions estab­lish that the pri­vate bank­ing sys­tem can­not be relied upon to con­strain lend­ing to lev­els that are con­sis­tent with good macro­eco­nom­ic out­comes. Is it then pos­si­ble that the Cen­tral Bank itself can con­trol this lend­ing, and keep it with­in sen­si­ble bounds?

Eco­nom­ics text­books answer this ques­tion in the affir­ma­tive, as they teach the “mon­ey mul­ti­pli­er” mod­el of mon­ey cre­ation, in which bank loans can only be made after the gov­ern­ment has inject­ed reserves into the sys­tem. Yet almost 50 years ago, the then Vice Pres­i­dent of the New York Fed­er­al Reserve described the view that bank lend­ing was con­strained by the activ­i­ties of Fed­er­al Reserve as “naïve”. Speak­ing in oppo­si­tion to pro­pos­als put for­ward by Mon­e­tarists, Holmes stat­ed that:

The idea of a reg­u­lar injec­tion of reserves … also suf­fers from a naive assump­tion that the bank­ing sys­tem only expands loans after the Sys­tem (or mar­ket fac­tors) have put reserves in the bank­ing sys­tem. In the real world, banks extend cred­it, cre­at­ing deposits in the process, and look for the reserves lat­er… the reserves required to be main­tained by the bank­ing sys­tem are pre­de­ter­mined by the lev­el of deposits exist­ing two weeks ear­li­er. (Holmes 1969, p. 73)

The rela­tion­ship of loans and deposits lead­ing and reserves lag­ging is more pro­nounced today, with the reserve lag now being 30 days (O’Brien 2007, Table 12, p. 52). The Euro­pean Cen­tral Bank has also recent­ly con­firmed that the Post Key­ne­sian posi­tion that “loans cre­ate deposits, and deter­mine reserves with a lag” accu­rate­ly describes pri­vate and Cen­tral Bank pro­ce­dures:

In fact, the ECB’s reserve require­ments are back­ward-look­ing, i.e. they depend on the stock of deposits (and oth­er lia­bil­i­ties of cred­it insti­tu­tions) sub­ject to reserve require­ments as it stood in the pre­vi­ous peri­od, and thus after banks have extend­ed the cred­it demand­ed by their cus­tomers. (ECB 2012, p. 21)

Con­ven­tion­al Cen­tral Bank mech­a­nisms to lim­it pri­vate mon­ey cre­ation have there­fore long been known to be inef­fec­tive, and Aus­tralia is one of many OECD nations that no longer has a reserve ratio (O’Brien 2007). But despite this, Neo­clas­si­cal econ­o­mists still tend to believe that the Cen­tral Bank can con­trol pri­vate bank lend­ing. Krug­man recent­ly argued in his debate with me that pri­vate bank lend­ing was con­strained by the cash print­ed by the Fed­er­al Reserve:

Yes, a loan nor­mal­ly gets deposit­ed in anoth­er bank — but the recip­i­ent of the loan can and some­times does quick­ly with­draw the funds, not as a check, but in cur­ren­cy. And cur­ren­cy is in lim­it­ed sup­ply — with the lim­it set by Fed deci­sions. So there is in fact no auto­mat­ic process by which an increase in bank loans pro­duces a suf­fi­cient rise in deposits to back those loans, and a key lim­it­ing fac­tor in the size of bank bal­ance sheets is the amount of mon­e­tary base the Fed cre­ates — even if banks hold no reserves. (Krug­man 2012)

I leave it to the Com­mit­tee to imag­ine the pub­lic’s reac­tion if cash in ATMs sud­den­ly ran out. The Cen­tral Bank’s reac­tion would be obvi­ous: notes would be print­ed in a gale and rapid­ly deliv­ered to ATMs. There is thus no way that bank loan offi­cers would even con­sid­er the amount of bank notes cur­rent­ly in the sys­tem as a con­straint on their lend­ing. Krug­man’s bizarre argu­ment here shows how far Neo­clas­si­cal eco­nom­ics is removed from real­i­ty.

We there­fore can­not rely upon pri­vate banks to lim­it their lend­ing to respon­si­ble lev­els, nor can we rely upon the cur­rent or past con­trol mech­a­nisms of the Cen­tral Bank. Clear­ly, a new approach to bank­ing reg­u­la­tion is need­ed.


The endoge­nous mon­ey per­spec­tive on macro­eco­nom­ics leads to very dif­fer­ent advice on how to man­age the bank­ing sys­tem than main­stream Neo­clas­si­cal the­o­ry. First­ly, banks can­not be relied upon to choose a pru­dent lev­el of debt: they will always want to cre­ate as much debt as they can per­suade the pub­lic to take on. There is noth­ing irra­tional in this behav­iour: it is mere­ly a ratio­nal response to the per­verse incen­tives they face ema­nat­ing from their capac­i­ty to pro­duce debt and mon­ey.

The objec­tives of post-GFC bank­ing reg­u­la­tion should there­fore include, in addi­tion to the cur­rent objec­tives of price sta­bil­i­ty and full employ­ment, qual­i­ta­tive and quan­ti­ta­tive mon­i­tor­ing of pri­vate debt, and the devel­op­ment of mech­a­nisms which lim­it or ter­mi­nate the dan­ger­ous pos­i­tive feed­back links between growth in pri­vate debt and asset prices.

Monitoring of private debt

The RBA should mon­i­tor the lev­el, rate of change and rate of accel­er­a­tion of pri­vate debt, and react with both qual­i­ta­tive and quan­ti­ta­tive con­trols when these indi­ca­tors gen­er­ate warn­ing sig­nals.

There is as yet no guid­ance as to what is an appro­pri­ate ratio of pri­vate debt to GDP. Some pri­vate debt is absolute­ly essen­tial, as not­ed ear­li­er, and the lev­el of debt will nec­es­sar­i­ly fluc­tu­ate in response to waves of inno­va­tion, as Schum­peter elo­quent­ly explains (Schum­peter 1934, Chap­ter 6). There­fore some vari­a­tion in the lev­el, rate of change and accel­er­a­tion of debt are to be expect­ed even in a healthy econ­o­my. Judg­ment is required, and this is some­thing that politi­cians should expect their advi­sors to be able to devel­op and deliv­er.

Clear­ly, the RBA has failed in this task in the past half century—for the sim­ple rea­son that, by fol­low­ing Neo­clas­si­cal eco­nom­ic the­o­ry, they ignored the role of pri­vate debt com­plete­ly. Had they tak­en account of it, they might well have react­ed to the trend for debt to rise that began in 1965, after 20 years of rel­a­tive sta­bil­i­ty at between 20 and 30 per­cent of GDP.

Fig­ure 20: Aus­trali­a’s aggre­gate pri­vate debt to GDP ratio since 1945

It is thus lit­tle won­der that today’s RBA staff—such as Ms Ellis—deny the rel­e­vance of mea­sures like the debt to GDP ratio, since to do oth­er­wise would be to admit that the RBA has been derelict of its cus­to­di­al respon­si­bil­i­ties to date.

This under­stand­able defence of past fail­ures can­not be allowed to let fail­ure con­tin­ue into the future. The RBA should do the hard work need­ed to devel­op a sen­si­ble mon­i­tor­ing of debt to GDP lev­els. This should include not only the lev­el of debt, but the pur­pos­es to which that debt is put. Lend­ing for invest­ment should be both encour­aged and large­ly left to its own devices. Lend­ing for asset purchases—some of which will always be nec­es­sary since assets like hous­es can­not be pur­chased from income alone—should be close­ly mon­i­tored for signs that it may be fuelling an asset price bub­ble. A key fac­tor here must be mon­i­tor­ing the accel­er­a­tion of cred­it. In addi­tion to my argu­ments on this top­ic here, the Com­mit­tee should con­sult Schu­lar­ick & Tay­lor 2009: they con­clud­ed that the accel­er­a­tion of debt was the most impor­tant warn­ing of an approach­ing finan­cial cri­sis (Schu­lar­ick and Tay­lor 2009).

Of course for the fore­see­able future, the Trea­sury and RBA will have to man­age the ten­den­cy for the lev­el of debt to decline, with the neg­a­tive impli­ca­tions that has for eco­nom­ic growth. Giv­en that delever­ag­ing after the Great Depres­sion took 15 years (includ­ing the Sec­ond World War), this process will dom­i­nate the RBA’s and the Trea­sury’s activ­i­ties for as much as a generation—and if you think that is an extreme state­ment that is eas­i­ly dis­count­ed, ask your­self: what would you have said if, in 1985, I had pre­dict­ed that eco­nom­ic growth in Japan would cease in the near future, and remain stag­nant for two decades?

As much as mon­e­tary and fis­cal pol­i­cy must now pre­pare to deal with the con­se­quences of past fail­ures, we should also con­sid­er how to redesign the finan­cial sys­tem so that, once this cri­sis is tru­ly behind us, a future cri­sis can be pre­vent­ed.

I do not believe that reg­u­la­tion alone will achieve this aim, for two rea­sons.

First­ly, Min­sky’s propo­si­tion that “sta­bil­i­ty is desta­bi­liz­ing” applies to reg­u­la­tors as well as to mar­kets. If reg­u­la­tions actu­al­ly suc­ceed in enforc­ing respon­si­ble finance, the rel­a­tive tran­quil­li­ty that results from that will lead to the belief that such tran­quil­li­ty is the norm, and the reg­u­la­tions will ulti­mate­ly be abol­ished. After all, this is what hap­pened after the last Great Depres­sion.

Sec­ond­ly, banks prof­it by cre­at­ing debt, and they are always going to want to cre­ate more debt. This is sim­ply the nature of bank­ing. Reg­u­la­tions are always going to be attempt­ing to restrain this ten­den­cy, and in this strug­gle between an “immov­ably object” and an “irre­sistible force”, I have no doubt that the force will ulti­mate­ly win.

If we rely on reg­u­la­tion alone to tame the finan­cial sec­tor, then it will be tamed while the mem­o­ry of the cri­sis it caused per­sists, only to be over­thrown by a resur­gent finan­cial sec­tor some decades hence (scep­tics on this point should take a close look at Fig­ure 3, which shows the debt to GDP data for Aus­tralia from 1860 till today).

There are thus only two options to lim­it cap­i­tal­is­m’s ten­den­cies to finan­cial crises: to change the nature of either lenders or bor­row­ers in a fun­da­men­tal way. My pref­er­ence is to address the lat­ter by reduc­ing the appeal of lever­aged spec­u­la­tion on asset prices.

Redefining financial assets

There are, I believe, no prospects for fun­da­men­tal­ly alter­ing the behav­iour of the finan­cial sec­tor because, as already not­ed, the key deter­mi­nant of prof­its in the finance sec­tor is the lev­el of debt it can gen­er­ate. How­ev­er it is organ­ised and what­ev­er lim­its are put upon its behav­iour, it will want to cre­ate more debt.

There are prospects for alter­ing the behav­iour of the non-finan­cial sec­tor towards debt because, fun­da­men­tal­ly, debt is a bad thing for the bor­row­er: the spend­ing pow­er of debt now is an entice­ment, but with it comes the draw­back of ser­vic­ing debt in the future. For that rea­son, when either invest­ment or con­sump­tion is the rea­son for tak­ing on debt, bor­row­ers will be restrained in how much they will accept. Only when they suc­cumb to the entice­ment of lever­aged spec­u­la­tion will bor­row­ers take on a lev­el of debt that can become sys­tem­i­cal­ly dan­ger­ous.

Jubilee Shares

The key fac­tor that allows Ponzi Schemes to work in asset mar­kets is the “Greater Fool” promise that a share bought today for $1 can be sold tomor­row for $10. No inter­est rate, no reg­u­la­tion, can hold against the charge to insan­i­ty that such a fea­si­ble promise fer­ments, and on such a foun­da­tion the now almost for­got­ten fol­ly of the Dot­Com Bub­ble was built. Both the promise and the fol­ly are well illus­trat­ed in Yahoo’s share price

Fig­ure 21: Yahoo Share Price

I pro­pose the rede­f­i­n­i­tion of shares in such a way that the entice­ment of lim­it­less price appre­ci­a­tion can be removed, and the pri­ma­ry mar­ket can take prece­dence over the sec­ondary mar­ket. A share bought in an IPO or rights offer would last for­ev­er (for as long as the com­pa­ny exists) as now with all the rights it cur­rent­ly con­fers. It could be sold sev­er­al times (say, sev­en times) onto the sec­ondary mar­ket with all the same priv­i­leges. But on its next sale it would have a life span of 50 years, at which point it would ter­mi­nate.

The objec­tive of this pro­pos­al is to elim­i­nate the appeal of using debt to buy exist­ing shares, while still mak­ing it attrac­tive to fund inno­v­a­tive firms or star­tups via the pri­ma­ry mar­ket, and still mak­ing pur­chase of the share of an estab­lished com­pa­ny on the sec­ondary mar­ket attrac­tive to those seek­ing an annu­ity income.

I can envis­age ways in which this basic pro­pos­al might be refined, while still main­tain­ing the pri­ma­ry objec­tive of mak­ing lever­aged spec­u­la­tion on the price of exist­ing share unat­trac­tive. The ter­mi­na­tion date could be made a func­tion of how long a share was held; the num­ber of sales on the sec­ondary mar­ket before the Jubilee effect applied could be more oth­er than sev­en. But the basic idea has to be to make bor­row­ing mon­ey to gam­ble on the prices of exist­ing shares a very unat­trac­tive propo­si­tion.

The Pill”

At present, if two indi­vid­u­als with the same sav­ings and income are com­pet­ing for a prop­er­ty, then the one who can secure a larg­er loan wins. This real­i­ty gives bor­row­ers an incen­tive to want to have the loan to val­u­a­tion ratio increased, which under­pins the finance sec­tor’s abil­i­ty to expand debt for prop­er­ty pur­chas­es.

Since the accel­er­a­tion of debt dri­ves the rise in house prices, we get both the bub­ble and the bust. But since hous­es turn over much more slow­ly than do shares, this process can go on for a lot longer.

Lim­its on bank lend­ing for mort­gage finance are obvi­ous­ly nec­es­sary, but while those con­trols focus on the income of the bor­row­er, both the lender and the bor­row­er have an incen­tive to relax those lim­its over time. This relax­ation is in turn the fac­tor that enables a house price bub­ble to form while dri­ving up the lev­el of mort­gage debt per head.

I instead pro­pose bas­ing the max­i­mum debt that can be used to pur­chase a prop­er­ty on the income (actu­al or imput­ed) of the prop­er­ty itself. Lenders would only be able to lend up to a fixed mul­ti­ple of the income-earn­ing capac­i­ty of the prop­er­ty being purchased—regardless of the income of the bor­row­er. A use­ful mul­ti­ple would be 10, so that if a prop­er­ty rent­ed for $30,000 p.a., the max­i­mum amount of mon­ey that could be bor­rowed to pur­chase it would be $300,000.

Under this regime, if two par­ties were vying for the same prop­er­ty, the one that raised more mon­ey via sav­ings would win. There would there­fore be a neg­a­tive feed­back rela­tion­ship between lever­age and house prices: an gen­er­al increase in house prices would mean a gen­er­al fall in lever­age.

I call this pro­pos­al The Pill, for “Prop­er­ty Income Lim­it­ed Lever­age”. This pro­pos­al is a lot sim­pler than Jubilee Shares, and I think less in need of tin­ker­ing before it could be final­ized. Its real prob­lem is in the imple­men­ta­tion phase, since if it were intro­duced in a coun­try where the prop­er­ty bub­ble had not ful­ly burst, it could cause a sharp fall in prices. It would there­fore need to be phased in slow­ly over time—except in a coun­try like Japan where the house price bub­ble is well and tru­ly over (even though house prices are still falling).


This is a well-timed inquiry. Recent events in Europe and the USA have con­firmed that the GFC is still very much with us, and Aus­trali­a’s only mod­er­ate eco­nom­ic per­for­mance even with the ben­e­fit of the Chi­na boom has empha­sised that we are, like the rest of the OECD, affect­ed by the dilem­ma of too much debt. I look for­ward to dis­cussing these issues with the Com­mit­tee if and when a pub­lic hear­ing is arranged.

Bernanke, B. (2010). On the Impli­ca­tions of the Finan­cial Cri­sis for Eco­nom­ics. Con­fer­ence Co-spon­sored by the Cen­ter for Eco­nom­ic Pol­i­cy Stud­ies and the Bend­heim Cen­ter for Finance, Prince­ton Uni­ver­si­ty. Prince­ton, New Jer­sey, US Fed­er­al Reserve.

Bernanke, B., M. Gertler, et al. (1996). “The Finan­cial Accel­er­a­tor and the Flight to Qual­i­ty.” Review of Eco­nom­ics and Sta­tis­tics
78(1): 1–15.

Bernanke, B. S. (2000). Essays on the Great Depres­sion. Prince­ton, Prince­ton Uni­ver­si­ty Press.

Bernanke, B. S. (2004). Pan­el dis­cus­sion: What Have We Learned Since Octo­ber 1979? Con­fer­ence on Reflec­tions on Mon­e­tary Pol­i­cy 25 Years after Octo­ber 1979, St. Louis, Mis­souri, Fed­er­al Reserve Bank of St. Louis.

Blan­chard, O. (2009). “The State of Macro.” Annu­al Review of Eco­nom­ics
1(1): 209–228.

Cotis, J.-P. (2007). Edi­to­r­i­al: Achiev­ing Fur­ther Rebal­anc­ing. OECD Eco­nom­ic Out­look. OECD. Paris, OECD. 2007/1: 7–10.

ECB (2012). Mon­e­tary and Finan­cial Devel­op­ments. Month­ly Bul­letin May 2012. Brus­sels, Euro­pean Cen­tral Bank.

Ellis, L. (2012). Pru­dent Mort­gage Lend­ing Stan­dards Help Ensure Finan­cial Sta­bil­i­ty. Aus­tralian Mort­gage Con­fer­ence 2012. Syd­ney.

Fama, E. F. and K. R. French (1999). “The Cor­po­rate Cost of Cap­i­tal and the Return on Cor­po­rate Invest­ment.” Jour­nal of Finance
54(6): 1939–1967.

Fish­er, I. (1933). “The Debt-Defla­tion The­o­ry of Great Depres­sions.” Econo­met­ri­ca
1(4): 337–357.

Greenspan, A. (2005). Tes­ti­mo­ny of Chair­man Alan Greenspan: The eco­nom­ic out­look. Wash­ing­ton, Joint Eco­nom­ic Com­mit­tee, U.S. Con­gress.

Holmes, A. R. (1969). Oper­a­tional Con­straints on the Sta­bi­liza­tion of Mon­ey Sup­ply Growth. Con­trol­ling Mon­e­tary Aggre­gates. F. E. Mor­ris. Nan­tuck­et Island, The Fed­er­al Reserve Bank of Boston: 65–77.

Krug­man, P. (2012). “Bank­ing Mys­ti­cism, Con­tin­ued.” The Con­science of a Lib­er­al

Krug­man, P. (2012). End this Depres­sion Now! New York, W.W. Nor­ton.

Krug­man, P. (2012). “Min­sky and Method­ol­o­gy (Wonk­ish).” The Con­science of a Lib­er­al

Lucas, R. E., Jr. (2003). “Macro­eco­nom­ic Pri­or­i­ties.” Amer­i­can Eco­nom­ic Review
93(1): 1–14.

O’Brien, Y.-Y. J. C. (2007). “Reserve Require­ment Sys­tems in OECD Coun­tries.” SSRN eLi­brary.

Schu­lar­ick, M. and A. M. Tay­lor (2009). Cred­it Booms Gone Bust: Mon­e­tary Pol­i­cy, Lever­age Cycles and Finan­cial Crises, 1870–2008, C.E.P.R. Dis­cus­sion Papers, CEPR Dis­cus­sion Papers: 7570.

Schum­peter, J. A. (1934). The the­o­ry of eco­nom­ic devel­op­ment : an inquiry into prof­its, cap­i­tal, cred­it, inter­est and the busi­ness cycle. Cam­bridge, Mass­a­chu­setts, Har­vard Uni­ver­si­ty Press.

Sta­ple­don, N. D. (2007). Long Term Hous­ing Prices in Aus­tralia and Some Eco­nom­ic Per­spec­tives. Eco­nom­ics PhD, Uni­ver­si­ty of New South Wales.

Stevens, G. (2008). “Inter­est­ing Times.” Reserve Bank of Aus­tralia Bul­letin
Decem­ber 2008: 7–12.



Bookmark the permalink.

About Steve Keen

I am Professor of Economics and Head of Economics, History and Politics at Kingston University London, 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 private debts accumulated globally, and our very low rate of inflation.