...against fictions and other tall tales

Saturday, 19 December 2015

Loanable Funds Theories: Classical vs Keynesian

It's been an embarrassingly long time since my last post. It's not due to a shortage of good topics to write about. Rather, I've been caught up in a number of projects at work and been busy on the home front. Hopefully, this post, which was inspired by ongoing conversations I've been having offline with friends and colleagues will partly make up for the radio silence.

One of the biggest fallacies in macroeconomics and macro policy is the idea that the interest rate is determined by the intersection of the upward sloping supply curve of (desired) savings and downward sloping curve of (desired) investment.

According to this traditional, credit-based approach to interest rate determination (as opposed to a "money-based" one à la Keynes's Liquidity Preference Theory in which the interest rate is determined by the supply of and demand for money), savings consist of the supply of loanable funds (i.e., funds that are not spent on consumption), assumed to be positively related to the interest rate, while investment is the demand for loanable funds and assumed to be negatively related to the interest rate.

As far as simple theories go, the old (classical) loanable funds theory is of very little or no use for making sense of the real world and in terms of providing prescriptive insight to policymakers. For one, there's very little evidence that the real rate of interest has significant impact on business investment.

But more importantly, the problem with the traditional loanable funds theory is its public policy implications: it assigns no role to government as a stabilizing feature of the economy during a recession, which is a ludicrous proposition given what we know now about the Great Depression, the Japanese Lost Decade(s) and the Great Recession (the lesson being that government has a role to support recovery, as market mechanisms won't be sufficient, or at the very least, will take too long).

Consistent with the classical origins of this approach, the traditional loanable funds model holds that government intervention to stabilize the economy is not needed because, as the economy falls into recession, there is an automatic stabilizing force clearing the loans market, enabling the supply of and demand for funds to adjust, shifting to the left, reaching a new equilibrium, as in the diagram below.

The self-adjusting mechanism works as follows: first, as the shift in demand for funds is assumed to be greater than the shift in the supply of loans during a downturn, the result is a decrease in the (real) rate of interest, which subsequently causes investment to recover, thus helping to restore economic activity and growth.

In his General Theory, J.M. Keynes illustrated how widespread the belief in traditional loanable funds theory was during his time:
Certainly the ordinary man — banker, civil servant or politician — brought up on the traditional theory, and the trained economist also, has carried away with him the idea that whenever an individual performs an act of saving he has done something which automatically brings down the rate of interest, that this automatically stimulates the output of capital, and that the fall in the rate of interest is just so much as is necessary to stimulate the output of capital to an extent which is equal to the increment of saving; and, further, that this is a self-regulatory process of adjustment which takes place without the necessity for any special intervention or grandmotherly care on the part of the monetary authority. Similarly — and this is an even more general belief, even today — each additional act of investment will necessarily raise the rate of interest, if it is not offset by a change in the readiness to save.  
In a recent post, Paul Krugman dismissed the relevance of the traditional loanable funds model for the real world, pointing out the basic Keynesian insight that the theory is only relevant if the level of income in the economy is fixed. In reality, as Krugman correctly argues, given that income is not fixed, all the traditional loanable funds theory does is "define a relationship between interest rates and income, the IS curve of the conventional Keynesian IS-LM model".

Also, in his post Krugman showed how misleading the model can be for explaining the determination of the rate of interest in a world where the central bank sets the short term interest rate as a way to achieve its policy objective (i.e., hit its inflation target):
The Fed sets interest rates, whether it wants to or not — even a supposed hands-off policy has to involve choosing the level of the monetary base somehow, which means that it’s a monetary policy choice.

Keynesian Credit-based Loanable Funds Theory (credit view) vs Classic Loanable Funds Theory (money view)

So it needs to be repeated: the old loanable funds theory is irrelevant for understanding how the economic activity resumes after a downturn. That said, the basic insight that "credit matters" and that credit fluctuations have significant effects on the real economy should not be rejected out of hand.

Such was the thinking in the 1970s and 1980s when a few Keynesian economists, including Andrew Weiss, Joseph Stiglitz, Bruce Greenwald, Benjamin Friedman, Alan Blinder and others (including, to some extent, Ben Bernanke) set out to devise Keynesian-inspired credit-based models as alternatives to the conventional and popular Keynesian and monetarist "money-based" models that dominated macroeconomics at the time (all of which assumed a special role for money in the determination of aggregate demand).

The result was a set of economic models highlighting the importance of credit in the economy and the critical role of commercial banks in affecting real output. Though they were labelled "new and improved" versions of loanable funds theory because they emphasized credit rather than money, these models were nothing like their classical predecessor, as the new models assigned no special role to the supply of savings and recognized the critical role of government regulation and macroeconomic stabilization to improve economic outcomes.

At the heart of these new models is the idea that, unlike the traditional loanable funds model, credit is not allocated in an auction process, with the loan going to whoever is willing to pay the highest interest rate. Rather, in these models banks understand that increasing interest rates can in some instances (especially when economic activity is weak or slowing) increase the probability of borrowers to default, as increased lending rates can lead to adverse effects on the incentives of borrowers to undertake activities that are increasingly risky.

Default and bankruptcy are therefore possible in these models -- unlike in the traditional loanable funds model -- because lenders are often unable to properly assess the risk profile of potential borrowers due to a lack of information or the high cost of adequately assessing the default risk of potential borrowers.

So, rather than being determined by the forces of supply and demand, the interest rate in these Keynesian credit-based models is determined by the maximum expected return to banks, that is, the rate with which profits are maximized and risks (e.g., probability of loans not being repaid that can lead to an increase risk of bankruptcy) to the bank are minimized.

In other words, the interest rate is a variable determined by banks themselves as a way to remain profitable. In these models, there is no presumption that increases in the interest rate will boost bank profits given that higher interest rates can increase the probability that borrowers will not pay back their loan, which could result in profit losses for the bank and, in some cases, could lead to bankruptcy.

Nor is there is any presumption in these models that the market for loans is perfect and clears. The market mechanism in these models does not lead credit supply to equal demand because, in their attempt to maximize profits and minimize risks in a context where information about borrower risk is scarce and/or costly to uncover, banks will not supply the amount of credit necessary to meet the demand at the lending rate. In other words, the result is an excess demand for funds (i.e., credit rationing).

Perhaps the simplest and most revealing of these Keynesian loanable funds models is the model by Joseph Stiglitz and Bruce Greenwald (2003)*. In this model, unlike in the traditional loanable funds model, the supply of loans (regardless of whether those loans are supplied via traditional financial intermediation or credit creation**) is not depicted as an upward sloping curve, as in traditional credit-based (loanable funds) models. Rather, banks' supply of loans is represented by a backwards bending curve (see chart below). The reason the curve bends backward in this model is because a rise in the interest rate increases average borrower risk. Also, the model assumes banks scale back the amount of loans supplied as the rate of interest increases to avoid borrower default and, by consequence, profit losses. (In a recent talk, Stiglitz referred to this model as a modern, Keynesian version of Irvin Fisher's debt-deflation theory. The similarities aren't obvious, but they are there.)

When combining the backwards sloping supply curve with a conventional demand curve for loans (where the demand for loans increases as the interest rate falls), the result is a credit market that is not perfect in the sense that the market mechanism (supply and demand for credit) does not provide a market-clearing interest rate, where the demand exceeds the supply for credit.

Ideally, in this model banks should be lending at point L* and setting the lending rate at r* where expected returns are maximized and where credit rationing (CR in the diagram) occurs. However, screening loan applications and paying interest on deposits imply costs to banks, therefore, typically the lending rate will be set a little lower, at point e. At point e, however, there is even more credit rationing because there is less lending.

From a macroeconomic perspective, credit rationing can lead to reduced economic activity by limiting aggregate demand, investment, employment and output. When credit is not available, firms involved in production cut employment and investment, thus lowering national income, output and the general level of employment. In some instances, as Alan Blinder argues, credit rationing can even lead to a Keynesian shortage of effective supply, that is, a shortage of produced goods and services to meet current demand, which in some situations could have the effect of increasing prices.

The impact of credit rationing on output will depend on whether the firms are dependent on bank loans. Firms that rely on bank financing will cut spending while businesses using bank loans for working capital will stop operating.

The table below provides a comparison between the traditional (classical) loanable funds theory and the modern Keynesian version.

Monetary Policy...

So what are the implications of this model for monetary policy? The conventional story holds that central banks reduce interest rates and investment increases as a result. In the Keynesian loanable funds models, the central bank may succeed in driving down the rate of interest on government securities (Treasury bills), however, it may not get banks to reduce their lending rate if banks perceive an increased risk of default on the part of households and firms due to a worsening economy, or if banks believe economic conditions are not likely to improve. Instead of increasing their loan portfolio, banks could simply choose to purchase safe government securities, as was done during the Great Depression, an outcome that does nothing to support recovery unless it prompts government to implement a fiscal stimulus by making public sector borrowing more attractive.

One doesn't have to think too much to see the relevance of these models to the period since the onset of the Great Recession.

As for contractionary monetary policy, the outcome is similar to the mainstream story in that investment can be curtailed as a result of the increased rate on government securities. However, as Stiglitz and Weiss, argue, "banks will often be unwilling to raise interest rates because of a fear that higher rates will have the adverse effect of chasing away credit-worthy borrowers and adverse incentive effect [of] inducing them to undertake greater risks". Instead, banks may opt to restrict the supply of loans, as in the diagram below.

So the point here is...

A basic principle in Keynesian economics is that no matter how dedicated unemployed workers are in their search for employment or how low the unemployed are willing to bid wages down, there are times when these actions are futile because jobs just are not available to meet the demand. The key insight of Keynesian credit models is similar, except that the crucial element is the insufficient supply of credit. In other words, this 'credit view' can be summarized as follows: often times the amount of loans is not sufficient to meet the demand for credit, regardless of the rate of interest.

To conclude, the main take away from this post is that the influence of interest rates (including the natural rate of interest) is often oversold, as the rate of interest may not be as important as it's often made out to be in the determination of aggregate demand.

The innovative aspect of credit-based Keynesian models was to shift the focus from money (as emphasized in monetarist models) and interest rates (as emphasized in traditional, old Keynesian models) towards elements such as the general degree of risk perceived by banks, both with regard to the default risk of potential borrowers and banks' expectations about future economic conditions. These are the key factors that influence the amount of credit supplied in the economy in credit-based Keynesian models.

* This blog post is dedicated to Joseph Stiglitz and Bruce Greenwald.

** Joseph Stiglitz & Bruce Greenwald (2003): "When a bank extends a loan, it creates a deposit account, increasing the supply of money."


Blinder, Alan. "Credit Rationing and Effective Supply Failures" in Macroecomics Under Debate, Ann Arbor, University o Michigan Press, 1992

Stiglitz, Joseph. and Bruce Greenwald, Toward a New Paradigm in Monetary Economics, 2003.


  1. Right you are.
    An excellent presentation of the issue. The distinction was worth emphasizing, and Stiglitz/ Greenwald deserve to be highlighted!

  2. Keynes was able to write the General Theory because he was a mathematician by training and a supreme logician. Most of the logic in the General Theory is a brute force application of logic to accounting as Keynes developed it at the macro level. He developed national income and wealth accounting before it was formally invented. And he did this in the case of loanable funds.

    It is utter nonsense to claim that the equivalence of saving and investment determines anything when that equivalence is a tautological accounting identity based on the only coherent relationship that can prevail between saving and investment as core concepts in economics.

    That was his core message about loanable funds. Anything discussed about it beyond that is bunk. Any idea good or bad slapped on top of it will just be weighed down by such foundational gibberish. Enough already.

    So it is with Krugman now as he involves his great new insight about how central banks set interest rates while still not understanding this basic message of Keynes.

    If Keynes were alive today, he'd be writing about how banks manage their capital. I reckon that would answer the question about the utility of QE reserves in about a one page update to the GT.  But just as the profession today remains dumb to the basic accounting instincts of Keynes, it hasn't got a clue about how banks actually manage their capital, preferring the band aid evolution of its own historic litany of "models" rather than taking the obviously required initiative for some updated real world research.

    1. Obviously, I agree with you on the genius of Keynes and the irrelevance of the saving/investment identity for the determination of output. As for the use of 'loanable funds' label, this seems to have been an easy identifier for credit-based models devised in the 70s when the mainstream focus at the time was on money-based models in which the interest rate is determined by the supply of and demand for money. Personally, I like the moniker 'credit view' but in much of the literature the expression 'loanable funds theory' stuck (possibly to the chagrin of the developers).

      As for the last part, Stiglitz and Greenwald would probably agree with you. For one, Stiglitz has been all but sympathetic about QE and recent central bank obsession in seeking to engineer tiny changes in interest rates (with little impact).

  3. Thank you Circuit. I'm agreeing with JKH here. However, if what you intend to say is "NEK models of credit aren't as bad as Classical LF models," then ok, yes, and this is a very good post on that particular issue. However, there's nothing there that wasn't in Minsky, or some others in heterodoxy, so those models aren't original, and they don't cite predecessors. Also, as your PK quote suggests, the NEK view that "CBs set interest rates" is superficially right--it's from the belief that velocity became too variable to target M. There's nothing in the PK quote that suggests the CB couldn't directly target M or MB--and we know from his debate on banking with Keen and myself that he does very much see it this way--so the NEK are still wrong on basic operations of central banks and banks, even if these models are "better."

    1. Thanks STF for the comment. The point was simply to show that there are some sensible models known under the label of 'loanable funds theory' that are Keynesian, where borrower risk is critical and default occurs, where credit and banks matter, and where money and interest rates are mostly irrelevant variables. In my opinion, the models provide a tight, useful way of understanding interaction of policy, bank behavior and their impact on output fluctuations.

      I don't always agree with Krugman but I thought on his comment on the old, classical loanable funds theory was pretty good.

    2. Yes, from that narrower perspective, that's fine and it's a very good post (as usual!). As JKH points out, there are other problems, of course.

      Related question--do you know of any NEK growth models that don't have saving determining capital stock?

    3. I'm somewhat surprised by the comments because my take on circuit's piece was that certain models and perspectives can offer insights beyond their canonical formulae. That models have shortcomings is a trivial statement (they are so by definition)...so to jkh and stf I can only offer "nothing new there". That the profession (exc STF of course) doesn't know how banks manage their capital, I will opt for the profession rather than pretension. So I am not quite certain as to the vehemence of JHK's comment with specific regard to Keynes...in fact unless my reading of Keynes is blurred, the man espoused the spectrum of money-creation theories, depending on his texts, tastes and takes.

      Also, I doubt very much that any MIT economists, including Stiglitz, Blinder, Akerlof and others associated with this research program, all of whom are associates of Robert Solow, one of Hy Minsky's earliest friends, would dare to use ideas from Minsky without citing the man. Just a hunch...That's a big accusation. Do you have any evidence of Minsky's ideas being used in this fashion? Also, the originality of Stiglitz and Greenwald's work in asymmetric information, which was foundational to the research program described in this article, is well known. In fact, Joseph Stiglitz was awarded the Nobel prize for it.

    4. STF: Don't know but I sort of object to this idea that any Keynesian trained in mainstream schools are typical "new keynesian" and that any of their work is somehow faulty. Here's two economists, Joe Stiglitz and Bruce Greenwald who are highly critical of mainstream money-based and market clearing models, who have developed models that rely to a large extent on a skepticism of orthodox notions of central bank independence and central bank 'credibility', yet their work is bunched together with the work of others who hold none of these views. You would think the heterodoxy would be a tad more open and accepting of disciplinary pluralism.

    5. STF: That came out strong but it wasn't intended to be. Will work on a post next on why 'new keynesian' is a bad label because it covers too wide a range of views.

  4. May I respectfully suggest that no amount of work by a single individual will lead to changes in economists' views...It takes several of them working in the same direction...

  5. that have very essential and useful Loanable Funds Theories information to provide with us. thanks for bring useful info with us.

    loan facilities in uk | financial consultant in uk