Sunday, 1 December 2013

Malinvestment and the endogeneity of money

So much has been written about the endogeneity of money that I thought it was now widely accepted. But recent exchanges have shown me that people STILL aren't getting it. Most recently, there have been two themes doing the rounds that bother me:

- malinvestment is caused by a growing money supply
- the presence of excess reserves in the system indicates a growing money supply (and therefore malinvestment)

Both are wrong. They are wrong for slightly different reasons, but they both boil down to the same thing - that money exists independently of the actions of banks. It does not. If there is malinvestment in the system, it is caused by an excess of bank lending, not by a growth in the money supply: very fast broad money growth is a consequence (or better, an indicator) of excessive bank lending. And if there are excess reserves in the system, they are caused by the desperate attempts of central banks to stop the money supply falling as banks deleverage. You could say they are caused by LACK of bank lending. 

Firstly, let me run through how bank lending works. The example I shall give is simple unsecured bank lending, but it applies equally to all forms of bank lending, including retail and wholesale secured lending (mortgages, repo).

Banks do not "lend out" existing money. They agree a loan, then obtain the funds to enable the borrower to pay the money. Let's use a car loan as an example. The bank lends say £5,000 as an advance against a car that the borrower intends to purchase. The accounting entries are DR loan account (asset)*, CR customer current account (liability). At that point the bank's balance sheet has expanded by £5,000 and sterling M2 has also increased by £5,000. But because the entries are balanced, there is no need for additional reserves at this point.

At this point also, no payment has been made. All the bank has done is put money in the borrower's account. And it has not "moved" this money from anywhere. The bank's cash balance does not change, and neither do the balances on any other customer deposit accounts. Banks DO NOT LEND OUT DEPOSITS.

The customer may leave that money in his account for several days or even weeks after the accounting entries have been made for the loan. The money supply therefore increases as an inevitable consequence of loan approval, not the consequent payment. Unfortunately traditional "money multiplier" explanations of lending fail to separate loan approval from its subsequent drawdown. Consequently we get absurdities such as "banks need reserves in order to lend". No they don't. They need reserves in order to make payments. And it doesn't matter whether the money they are paying out is money they have lent to the customer, or money that the customer himself has deposited. The same reserves are used to make payments in both cases.

When the customer actually pays for the car, he withdraws £5,000 from his current account. The entries are DR customer current account (liability), CR reserve account (asset). At the central bank, the entries are DR reserve account (liability), CR receiving bank's reserve account (liability). Note there is no impact on the asset side of the central bank's balance sheet. The amount of reserves in the system does not change as a consequence of payments being made: only the distribution of those reserves changes. Or, putting it another way, banks don't lend out reserves.

The sending bank has to obtain funds if funding that payment would take its central bank reserve account below the required reserve limit. It does so by borrowing from other banks, or as a last resort by borrowing from the central bank. In these days of instantaneous funds transfer, you could even say that the bank funds the payment by borrowing back the £5,000 it has created from the bank it has just paid it to, since it can run a daylight overdraft in its reserve account and clear it at the end of the day. Making payments has no effect on broad money, but if a bank has to obtain additional reserves from the central bank then it can result in the monetary base increasing. Monetary base expansion therefore tends to lag broad money expansion.

When the loan is repaid (let's assume bullet repayment to keep it simple) the entries are DR customer current account (liability), CR loan account (asset), reducing the loan account to zero. The bank's balance sheet has shrunk by £5,000 and so has M2 by the same amount. Repaying a loan actually destroys money. Yes, the bank may make another loan of £5,000 to someone else, in which case the money supply will be restored, but it doesn't have to. It will only do so if the risk versus return profile at that point in time is in its favour. And that is not a constant factor.

Prior to 2008, banks were lending exorbitantly, and borrowers were lapping it up. Money supply consequently grew very fast. Then came the subprime crisis, followed by the collapse of Lehman and the financial crisis. The swathe of mortgage and other loan defaults at that time not only caused bank failures, it caused a sudden catastrophic fall in the rate of growth of the broad money supply. As an example, this chart from the ECB clearly shows the overheating growth of M3 followed by sudden fall back when the crisis hit:

Click here to see the full series

Post-2008, banks are more risk-averse than they were, partly because of regulatory changes that force them to include more equity in their funding mix, and partly because - as the Large report noted - bank staff responsible for lending decisions have had their fingers very badly burned and are terrified of getting it wrong again. (Sometimes I think we forget that lending decisions are made by people.) And banks are still cleaning up their balance sheets. Therefore they are not lending as much as they were prior to 2008. They discourage demand in two ways - firstly by charging higher interest rates on loans, particularly to poorer risks (that's why interest rates to SMEs have gone up considerably since 2008), and secondly by refusing to lend at all. Furthermore, households and corporates are not borrowing to the extent that they were before the crisis. They are paying off debt and not taking out new loans.

The problem prior to 2008 was that banks mispriced risk and therefore lent far more than they should have done at too low a price. The result was malinvestment (sub-prime loans and their derivatives, dodgy corporate lending and so forth). And the result was also a sharply rising money supply AS A CONSEQUENCE of that malinvestment. In Austrian economics, this is inflation - but it showed itself as rising prices in certain asset classes, not a general rise in the price level. I'm not going to discuss the reasons for this here: inflation-targeting central banks claim credit for it, but in my view it might also have something to do with the imbalances in international trade and the general stagnation in wage levels in developed countries.

We now have the opposite problem: banks don't want to lend, and households and corporates don't want to borrow. The result is a broad money supply that stubbornly refuses to grow. Central banks are attempting to counter that by increasing the monetary base - with some success, though mainly through portfolio effects as bank lending, particularly in Europe, is still very low. Except for the ECB, that is, which has been happily allowing broad money supply to stagnate for the whole of 2013, mainly due to banks paying off the LTROs early and not lending to periphery businesses.

This is not to say, however, that there is no malinvestment at the moment. There is, and it comes in two forms: residual malinvestments from prior to the crisis that banks, households and businesses are still trying to unload, and new malinvestments. But the new malinvestments are not caused by central banks increasing the supply of the monetary base. They are caused, just like the older malinvestments, by the mispricing of risk.

Good investment requires accurate pricing of risk. Too cautious a view is as damaging as too reckless a view. So when banks and investors take too cautious a view, they stuff their balance sheets with safe assets while denying viable businesses the investment they need. When businesses take too cautious a view, instead of investing in risky projects for the future they hoard cash on their balance sheets and buy back their own stock. When households take too cautious a view, the result is a stagnant housing market and depressed consumer demand. And when governments take too cautious a view, the result is lack of public sector investment.

Prior to the crisis, malinvestment was caused by too optimistic a view of risk: post-crisis, it is caused by too pessimistic a view. Malinvestment doesn't end when there is a crash: a crisis does not necessarily clear out all malinvestment. Denying finance to viable businesses because of too-tight credit criteria, or killing off developing businesses by raising interest rates to "clear out zombies", is just as much malinvestment as providing credit to businesses whose business plans have serious weaknesses or whose market is in terminal decline.

We can reasonably ask what is being done with all the money that central banks are producing. I have no doubt that there are all manner of stupid investments being made. But I am not convinced that investment would be any better directed without central bank support. Unless we become much, much better at pricing risk, malinvestment is an inevitable consequence of doing business.

Related reading:
Understanding the Modern Monetary System - Cullen Roche
Malinvestment, not overinvestment, causes booms - Ludwig von Mises (excerpt)
The Trading Dead - Tom Papworth, Adam Smith Institute
Reply to Pedro al Sombrero Negro - Smiling Dave's Blog

* Footnote for non-accountants. DR to an asset increases it, CR decreases it. DR to a liability decreases it, CR increases it. So a DR to an asset is balanced by a CR to a liability, and together they increase the size of the balance sheet. Conversely, a CR to an asset is balanced by a DR to a liability, and together they decrease the size of the balance sheet. I hope that is clear.


  1. There's a typo in the footnote, where it says "DR to a liability decreases it, DR increases it." should say "DR to a liability decreases it, _CR_ increases it." thanks for the post

  2. Good piece Frances... just have a problem with you calling natural market clearing forces mal-investment. Mal-investment occurs in the real sector. This is characterized by investing in areas where the demand doesn't justify the investment (over-investment) or where there are not enough resources to complete the investment such that, when the crash comes, you have half finished projects which will act as capital consumed.
    Mal-investment, according to Austrians, arises from an increase in money supply that suggest there is more capital than there actually is. (notice the difference between capital and financing). The increased investment as a result of low 'price' of money and the increased consumption, also as a result of the low price of money, leads to over-consumption and over-investment. The mal-investment is revealed when real resources get bid up by excess money over capital which leads to a collapse in demand and a collapse in resources available to invest... usually manifested with higher real interest rates. This collapse in debt driven consumption and debt driven investment is what requires clearing in the real economy or what we call markets (govt intervention will only reallocate resources more destructively). This process is really a recession and is a cure... not a form of mal-investment. Otherwise, we might have to re-define what investment is...

    1. I fundamentally disagree, I'm afraid. Indeed that was partly the point of the post. Too high a price of money causes what we might call "mal-investment" (in the sense that investment is unnecessarily denied) just as much as too low a price. By focusing only on mispricing in boom times and ignoring mispricing in the bust, Austrians tell only half the story.

    2. I agree with you that just as underpricing is undesirable, overpricing is... however, I wouldn't call the result of overpricing mal-investment. I would call it under-investment. Here's the catch... if you proceed down this road, then the free market (i.e. free of government intervention including the Monetary Authorities' intervention) is desirable as a means to set the price of money. We are talking loanable funds here so that we don't have over-pricing neither do we have underpricing.
      Austrians concentrate on underpricing for a reason... most if not all governments, are usually net borrowers and therefore usually tilt the market to their side. This is only natural and that's why the work of Austrians focuses more on underpricing of money. Like I've intimated earlier, Austrians prefer that markets set prices as opposed to one set of participants who are likely to be driven by certain biases or lack of market information or dynamics so that the market is allowed to correct itself every time it has excesses on whichever side...

    3. Umm. Johnson, in a recession there is no less capital, but it is not productively deployed because investors are hoarding safe assets. That is malinvestment, not under-investment.

    4. That really depends on your definition of capital. Sometimes economists misuse the term capital and financing. By virtue of fractional central banking one could argue that financing can be deemed infinite but capital can't. Capital can be destroyed... when factories burn down or are closed because of under-utilization, that is capital destroyed. Of course that requires one to have a theory of capital... something most schools of economics do not have.
      However, what you call purchase of safe assets... Keynes would may call that mal-investment. We call that savings. Especially when QE is destroying the value of money.
      Savings are important to society. As government destroys value, it is those who save i.e. buy oil, gold or currency that is beyond the reach of the home government that serve their economy because they retain value which could come in handy in future.
      To summarize, what you are calling mal-investment, I would call savings.

    5. Johnson,

      Investment is the use that is made of savings, not the savings themselves. The use that is made of savings can be productive or unproductive. "Maliinvestment" is unproductive investment.

      The essence of Keynes's argument is that when saving exceeds productive investment opportunities, malinvestment is inevitable. I don't think you would fundamentally disagree with this, would you?

      You argue that savings invested in commodities are productive in the long run. I would agree with that under certain circumstances - for example, if there is a real risk of political collapse that seriously raises the risk of investing in government debt or cash. But those are extreme circumstances. I am not convinced that political risk is generally so severe as to warrant sizeable holdings of commodities.

    6. "Mal-investment" is unproductive investment.

      True... but buying commodities is not mal-investment, it is flight to safety. Ideally, buying commodities with a better store of value than money means the 'investor' has little faith in the ability of money to maintain value. If cash we had commodity money, the 'investor' would have preferred to hold that. That's why its savings... the investor looks for an asset that will store value until when he can find an investment worth his while.

      'The essence of Keynes's argument is that when saving exceeds productive investment opportunities, mal-investment is inevitable.'

      When savings exceeds investment opportunities, interest rates come down in a sustainable manner and society is able to make more risky investments that pay off longer thus improving the long term welfare of society. This is the reason that places like Germany are able to make capital goods and third world countries are unable to. No such thing as a glut in savings is my argument.

      Investment is the use that is made of savings, not the savings themselves... You argue that savings invested in commodities are productive in the long run...

      Again, my point is, purchase of commodities should rightly be seen as savings. e.g. If I have savings but I can't find an investment worth my while, what do I do? I buy oil and store it overseas. When I find an investment I like, I sell my oil and buy my assets. I have therefore used oil as a store of value. If the currency were a better store of value I wouldn't have bought the oil. Ergo, purchase of oil was actually savings. I cannot term that as mal-investment because the alternative is to choose from the investments I have available which would yield returns below my required rate of return hence a mal-investment.
      The above would happen during a period when Central Banks are printing money aka QE.
      If it were a true glut in savings and interest rates were coming down significantly and the local currency and there is no inflation or mild deflation, then we can argue that interest rates are sustainably low and the investor, as pointed out earlier, will look for long cycle investments such as mining or production of capital goods so that he can boost yield and in that way, society is richer for it... again, a la Germany/China

  3. Frances,

    Re your point that “bank staff responsible for lending decisions have had their fingers very badly burned…” what’s your basis for saying that? There seems to be a widespread impression that a significant contributor to the crisis was PERVERSE incentives for bank staff and managers: e.g. that bank staff were rewarded in proportion to the VOLUME of loans they made, rather than the long term viability of those loans.

    If in fact those staff are being rewarded or punished in line with that long term viability, then that’s good news in a way, though obviously it will be discouraging lending just at the moment, as you point out.

    1. Those incentives have now been dismantled in every bank, Ralph. But I really suggest you read the Large report (link is in the post). He identifies risk aversion by bank lending staff who were scared of making more mistakes as a key factor in the collapse of SME lending at RBS since the crisis.

  4. Excellent until the last section. To err at investing is to plough means into assets to where they can no longer generate sufficient net cash flows. But - and here is my Austrian ‘but’ – ‘malinvestment’ is a term usually (and I think usefully) reserved for the subset of laying down too much capital in physical form, especially a durable and functionally specific one, thus turning a monetary crisis into a real-side one and so drastically reducing its tractability and ease of remedy.

    Underinvestment ‘shortens’, rather than preferentially ‘lengthens’ the interwoven skein of production, adding too few vertically specialized, end-consumer remote, path-dependent, long-amortizing stages to the web. It must eventually give rise to its own market by mixing too many liquid assets with too few satisfied needs in much the same way as was seen at the end of WWII in the USA. Overinvestment is a very different animal.

    Much of the ongoing lethargy of the world economy is due to the uncorrected Boomtime ‘lengthening’ (plus residential real-estate ‘broadening’). Hence, the anchor-chain of past obligations continues to drag down both income-fearful debtors and balance sheet-falsifying lenders. This toxic combination is what is impeding central bank efforts to reinvigorate that growth credit which they suppose to be the sine qua non of renewed economic expansion.

    Ironically, just as the central banks’ prior criminal laxity gave rise to the Boom, so their unimaginative appeal to artificially depressed interest rates is the what has limited the scale and pace of liquidation, recalculation, and title transfer from loss-deniers to gain-chasers (outside the gambling halls of the Global QE-sino, that is). This has both extinguished the incentives and crowded out many opportunities to rebuild our irrevocably lost wealth.

    It is an Austrian tenet that unless there is some gross distortion at work, individual mistakes tend to cancel out. Moreover, while entrepreneurs are all too fallible, we would almost define such folk as being better attuned to shifts in business conditions than are the rest of us poor salarymen.

    Thus, under a functioning market system operating within a framework of stable property rights and benign institutions, the dynamic is that of a non-zero sum, wealth-creating, self-stabilizing arbitrage. Here is an evolutionary process in which success in serving the consumer is rewarded with the transfer of resources to the better practitioners as part of a relentless selection which slowly raises material standards for us all. Under such conditions, malinvestment is not at all an ‘inevitable consequence of doing business’.

    Mass delusion during the boom is only possible if the signals by which the market marks out pricing gaps between supply and demand are swamped in intervention-derived noise. Entrepreneurs can mutate from equilibrium-seeking agents of negative feedback and positive gains to a viral plague of self-aggravating, loss-breeding speculation only when the cost of financial capital is pushed too low for too long – i.e., with that very ‘central bank support’ whose absence you seem to doubt would make little difference.

    Their supposedly depressed ‘spirits’ are not so much the result of some irrational group hysteria but rather of the justified apprehension that since both policy and politics have aimed at delaying a full reckoning of past follies, their evergreened, zombie counterparts will continue to depress their prospective returns through an exercise in state-sanctioned, bank-enacted, unfair competition. Worse, they darkly suspect that any profits they do make will be subject both to confiscatory, often retrospective taxation and to blame-deflecting, populist opprobrium. It is to this malaise that we can attribute the lowered ‘marginal efficiency of capital’, not to some technological hiatus or ‘secular stagnation’. And it is at this door, too, that we can lay the blame for the anaemic recovery, not at that of some fetishistic desire to hold money for its own sake in a supposed ‘liquidity trap’.

    1. The fundamental error you make is in assuming that in the bust phase capital reallocates itself to productive enterprises. No it does not. In a recession, investors are over-cautious and capital flies to safe havens. Safe havens by their very nature do not use capital productively - in order for capital to be used productively it must be placed at risk. Therefore there is as much malinvestment going on in the bust phase as the boom phase, just in the opposite direction.

    2. Sometimes I despair at the partisan nature of economics. While I struggle to buy in to the full Austrian story about the varying length of production processes, it seems common sense to me that a misleading boom leads to a flow and hence a remaining stock of investment that is unwise in hindsight - eg in the UK, estate agencies. It is surely wrong to say that capital is not reallocated in the bust phase - that is what bankruptcy does (did you see "The man who buys anything" this week on Channel 4?), and is probably not being allowed to do enough of. Also Frances, I am sceptical about your idea that "in order for capital to be used productively it must be placed at risk" - the textbook story about loanable funds demand abstracts from risk altogether, but it is still reasonable.

    3. Tim,

      I think you need to be extremely careful when abstracting from macro to micro. There may be individual examples of "good" capital realloction, but that does not mean that overall capital is reallocated in a more efficient or effective way.

      On loanable funds - no, that is a complete misunderstanding. "Loanable funds" by definition are placed at risk: when you make a loan there is always the possibility that you won't get your money back. It is not possible to generate a positive return without placing capital at risk.

    4. ""Loanable funds" by definition are placed at risk" Where is this defined? You are just making this up, Frances.

      I dare say that in normal times (ie when the central bank is not doing unconventional easing), an overnight reverse repo loan, which is about as risk-free as you can get, would normally command a positive real rate of interest.

    5. No, I am not making this up. Think it through. There is no such thing as risk-free lending. There is market risk in overnight reverse repo (collateral values can drop very fast). The yield on the 3-month UST is usually used as the proxy for the risk-free rate in WACC calculations, but even that is not completely free of risk. Therefore "loanable funds" (to the extent that they exist at all) are "at risk".

      I will block you if you are rude to me again.

    6. I am thinking it through - I appreciate that there are residual risks in, say, overnight repo, but they are so tiny as to be unable to account for anything but a miniscule positive real rate.

      I am trying not to be rude to you, Frances, but when you tell me that I am making a "complete misunderstanding" and justify that by asserting a "definition" without reference, I get a little irritated.

    7. There is no such thing as risk-free lending, Tim. It is merely a question of the degree of risk. In general, higher risks require higher returns. Your overnight repo doesn't carry much risk, so the the return is tiny. But it is not zero. Collateralising a loan does reduce its risk, particularly if there is a substantial haircut. But it doesn't eliminate the risk completely. Collateral values can fall, as I said. Indeed fast falling collateral values were one of the principal drivers of the financial crisis.

      I really don't need to give a reference for this, surely. It's common sense.

    8. "I really don't need to give a reference for this, surely. It's common sense."

      The issue here is whether a positive return on lending can (or even should) be expected without taking risk is vital to several of your posts, so yes, I do think that you should provide evidence to support that idea, or give it up. As I say, it is not common sense - the textbook explanation of capital markets and interest includes a production possibility frontier and consumer preferences and abstracts from risk (I presume because this could be allowed for by restating the PPF in terms of expected production possibilities), yet does not rule out a positive interest rate.

      You probably think that I am being fussy about this point, but I think that is hard to underestimate its importance at the moment. The reason why is that it is vital to understand where the risk-free interest rate would be in a well-functioning economy in normal times so that we can get an idea of how far away from normal we are. The Keynesian approach is to accept as given that the risk-free "natural rate" of interest is negative at present and try to lower market rates of interest to get negative return projects funded, but I would question whether this is wise. If the normal natural rate of interest is positive, then this approach is going to build up a stock of investment projects that would need to be worked off before we can get back to normal, and if the authorities dare not allow this work-off process to proceed, the economy will remain abnormal indefinitely. In that case, a more sensible approach might be to understand why the natural rate is negative, and try to fix that.

      I would particularly value being able to convince you about this, Frances, because you have been a prominent advocate of the view that savers should not expect positive risk-free returns, but I cannot explain why I think you are wrong if you don't show your working.

    9. Tim,

      I have shown my workings. Repeatedly, actually. Just not in this post. And others have also written on the same thing.

      There is no such thing as a risk free asset. That is a standard criticism of the idea that certain types of assets (such as insured deposits and government debt) carry no risk. Intrinsically, they do. That's why under normal circumstances even a "risk-free" asset carries a positive interest rate.

      However, the widespread belief that debt assets, in particular (bonds and deposits) should be completely risk free is what drives the explicit support given by governments and central banks to asset prices for the last six years. It's so that asset holders (including your savers) don't have to take huge losses. In effect, governments/CBs are reducing the risk of these assets. But the consequence of that is depression of real returns.

      Of course, if savers were prepared to accept that their assets could lose value - haircuts on deposits in failing banks, for example - then they would be able to demand a higher interest rate. But demanding a positive rate of return on investment while expecting others to take the risk is rent-seeking. While asset holders expect to be protected from losses due to commercial realities, and there is no growth to justify positive rents, nothing - or even negative - is the correct rate of return.

      The relationship of risk and return is not something I am making up. It is standard portfolio theory. Go look it up.

      I am not prepared to discuss this any more on this post, as it is off topic. If you wish to discuss further, please do so by email.

    10. "There is no such thing as a risk-free asset." I feared that we would get bogged down in that one! Correct, but, as I explained earlier, you can find assets that are so close to risk-free that it seems implausible to ascribe anything but the tiniest positive return to the existence of these residual risks. And since such assets, like overnight repo, have tended to command significantly positive returns, that suggests that an imputed / theoretical completely risk-free rate is normally non-zero - which the link on the risk-free asset you posted does not deny.

      I am aware of standard portfolio theory as I used to teach it, and it certainly does not say that the risk/return tradeoff runs through the origin - you have probably seen the typical CAPM diagram with the capital market line cutting the return axis above zero? The reference you give is not clear on this point, but it does say (my emphasis): "According to the risk-return tradeoff, invested money can render HIGHER profits only if it is subject to the possibility of being lost."

      I suspect that this discussion is relevant to your next post (on zombie firms), which I think is informative and interesting, so if necessary, we can continue it there. In the meantime, thanks for going this far.

    11. Yes, the standard CAPM diagram does show the risk-free rate as above zero. And near-risk-free assets have in the past had positive interest rates. But these are not normal times. We are in Nick Rowe's "weird world" where the natural interest rate is lower than the rate of growth. The standard CAPM diagram does not anticipate zero or negative risk-free rates. Indeed this will be the subject of another post - negative rates have considerable implications for cost of capital calculations.

      If the risk-free rate of interest is zero or lower, then savers have no right to expect positive returns unless they are prepared to accept losses. Really, therefore, what we are arguing about is whether the risk-free rate of interest is above zero. You are in effect saying that because the risk-free rate has always been above zero in the past, therefore it should be now too, and the only reason it isn't is central bank interference. I can see that you might not want to believe that "this time is different", but that doesn't mean it isn't.

      Much of my writing at the moment is an attempt to explain the reasons for Rowe's "weird world". Central bank interventions certainly play a part, but in the absence of counterfactuals it's impossible to establish whether the risk-free rate would be higher without them. It could even be lower: some people think that the natural rate is well below zero, and there's no doubt that central banks are propping up short rates (that's what positive interest on reserves does).

      This discussion isn't really relevant to the zombie firms post, either, given that the conclusion of the post was that they don't exist and prematurely raising rates would be likely to kill off the next generation of firms. I think email would be a better place for this discussion.

    12. Actually, I think / hope that we are now in agreement - in normal economic times, a positive risk-free rate of interest is reasonable, and quite possibly to be expected. A couple of implications follow about which I am much less sure, especially as inflation falls. One is whether the abnormal negative risk-free rate now prevailing is part of the solution (as Keynesians would have it) or the problem. The other is that, in fact in global terms, economic growth is presently about average rather than low, and capital markets are quite open, so what does a locally negative risk-free rate do for capital flows (eg into London property). Again, thanks for the discussion, Frances.

  5. Brick says

    I mostly agree but would clarify on a few points. While malinvestment might not be a result of money supply increases, central bank actions can be a precursor to malinvestment in other economies via capital flows. Perhaps what I am saying is that there can be a cross over from speculation to investment and they are not always separate. Lack of well priced risk returns (lack of investment opportunities) have skewed the balance toward speculation and away from investment such that malinvestment is more likely (mispriced risk).

    When the difference between long term and short term rates reduces banks have tendency to increase charges to make up profit, especially if the central bank is trying to pull down long term rates. Those charges can make some properly risk weighted returns unattractive to investors and give an incentive towards speculation.

    What is not touched on is whether one of the key malinvestors is the government or more specifically an investor who gets poor returns. Here I am thinking that the opportunities for government to get high economic returns for investment is significantly reduced in a developed economy. Building a new road or school has a higher return than just replacing one.Perhaps if government could create a new type of long term debt with high economic returns it might address the lack of investment opportunities.

    OK so I have gone of at a slight tangent (apologies). I am not totally convinced that banks have tightened up on lending all that much (Concentration limits, skewed targets and Credit skills of Relationship Managers from the RBS report), but have a dysfunctional system of scoring risk which fails miserably in a stagnant economy. For instance I doubt whether a bank credit system will take into account whether every builder in the area has stopped trading except one. If the problem is not clever enough credit criteria rather than too-tight credit criteria then demand and investment may be affected. I guess I agree we need to be much better at pricing risk.

    1. Hi Brick,

      Good points. Yes, I agree that QE can have some very damaging effects in other economies. I'm certainly no fan of QE. And I agree that this is due to investor behaviour not government action. One thing I didn't touch on the article is the search for yield that is leading those investors who are less risk averse to move funds to higher-risk investments. This can be seen as a productive use - indeed it was one of the aims of QE - but there is of course a danger that these flows can suddenly reverse, as we saw when Bernanke suggested tapering. Sudden reversal of capital flows is incredibly damaging. We are playing with fire.

      Personally I would prefer to see governments investing more in infrastructure and long-term projects, rather than central banks creating lots of money with no ability to influence where it is invested. It is my firm belief that when the private sector will not invest - because of what amounts to a collective funk - the public sector must. The slow recovery in my view is because the UK government joined in the funk (heavily influenced by markets and international opinion, admittedly) and foolishly cut public investment to the bone in a misguided attempt to cut the deficit before the private sector was able or willing to increase investment.

  6. There's no doubt that, historically, rapid money supply expansions and large-scale malinvestment aka bubbles have tended to appear together.

    The question is, is QE rapidly expanding the money supply, and is there some reason or reasons why this time such an expansion could appear without large-scale malinvestment?

    I would say QE is expanding money supply rather rapidly, but most of that money supply expansion is happening through central bank lending to government, and not through lending to the private sector. This kind of money supply expansion can spur malinvestment but is relatively less prone to do so than a money supply expansion driven by lending to the private sector.

    Contrary to the popular myth, we are not in a situation where central banks are increasing base money without any corresponding increase in broad money. That could happen if central banks were lending to commercial banks and commercial banks weren’t increasing their lending to the real economy. But central banks haven’t lent much recently to commercial banks, except in the Eurozone. The UK’s mortgage support involves central bank lending to commercial banks on strict condition they re-lend to homebuyers. The US Fed accomplishes essentially the same thing by buying the commercial banks’ mortgage loans through the intermediation of public mortgage guarantors. In both cases the central bank lending most definitely reaches the real economy.

    When central banks lend to the real economy, they expand the broad money supply, just as commercial bank lending to the real economy expands the broad money supply. The only difference is that when a central bank lends to the real economy it also increases the supply of base money. People tend to notice more the increase of base money, because creating an equal amount of base and broad money expands the smaller base money supply proportionally more than it increases the bigger broad money supply. Still, it’s the increase in broad money that matters by far most.

    The mechanics work as follows. When a central bank lends to the private sector, it creates base money by crediting it to the reserve account of the borrower’s commercial bank, which creates broad money by crediting an equal amount to the deposit account of the borrower. When a central bank lends to the central government, it creates base money by crediting it to the central government’s deposit account; when the government spends those funds, the Fed transfers funds from the central government’s deposit account to reserve account of the recipient of the spending’s bank, which creates broad money by crediting it to the recipient’s deposit account. (Such broad money creation from government spending is balanced by an equal amount of broad money destruction when the government funds itself by taxation or borrowing from the public.)

    Because QE lending is to mostly government, not the private sector, the most direct risk of large-scale malinvestment is if it encourages government to go on a “white elephant” investment spree. Among the big three QE central banks only in Japan is this a serious risk.

    There is also however a significant threat that QE could more indirectly cause malinvestment by distorting asset prices. Rising broad money supply means private parties must collectively hold more cash. This drives up asset prices as cash must become relatively less inexpensive in real terms to motivate more holding of it. This is driving overpricing of all kinds of financial and real assets – corporate bonds, equities, farmland, commercial real estate. It’s also suppressing returns for new investors while elevating them for those who owned assets before the price run-up.

    So far, I don’t see a lot of real investment happening in developed markets, so it’s hard to talk about large-scale investment. But we shouldn’t doubt that QE, pushed far and long enough, does have the power to drive malinvestment.

    1. Hi Tom,

      I broadly agree with your general point here. QE does increase broad money when the securities are purchased from investors rather than banks.

      However, you are wrong about the UK's mortgage support. It does not involve central bank lending at all. Help 2 Buy is government (not CB) either lending directly or guaranteeing commercial mortgages. And Funding for Lending is collateral enhancement: it only involves creation of base money if the bank uses the enhanced collateral to obtain funds through operational standing facilities at the Bank of England, but there is no guarantee that banks would do so - they could just as easily use the collateral to obtain cheap funds in the repo market.

      And I think you have the causation wrong regarding the Fed's mortgage support. The Fed is buying loans that have already been made. It is not creating new loans which generate deposits. If the Fed's action increases broad money, therefore, it is not because of lending activity but because it is purchasing securities from private sector holders, just as with government debt.

  7. Nice post. Unlike many others, the accounting on banking and money is correct. Still much discusison on effects of QE, ref. recent debate on QE and deflation. But, even if QE results in broad money increasing, there is no guarantee they will be used. Could be stuck with non-bank financial institutions, who only invest in high-yielding assets and create asset price bubble, and no employment. Better than to have central banks buy stuff from ordinary people, that could generate real demand, ref. Policy Note from Levy Economics Institute on this, see This would be tax relief without concurrent increase in govenrment debt.

    Basic problem is huge profits and cash piles on corporate balance sheets that sits idle due to extreme (but rational) risk aversion. As some have alluded to, better than to tax (part of) it, and have the governement increase public investment and growth. Ref. Keynes who was very clear about this in the GT (pp. 377-78): It seems unlikely that the influence of banking policy on the rate of interest will be sufficient by itself to determine the optimum rate of investment. I conceive, therefore, that a somewhat comprehensive socialization of investment will prove the only means of securing an approximation to full employment

  8. "because the entries are balanced, there is no need for additional reserves at this point."

    I don't think this is correct in general. In a system with reserves requirements, a bank will be required to hold additional reserves as soon as the balance sheet expansion counts towards its reserve requirement. While it is true that this will in most cases lag balance sheet expansion because of features like reserves maintenance periods, the key point is that reserves will be required fairly soon after the loan has been created, whether the loan is drawn down or not.

    I think this is another area of economics that seems to be subject to unproductive dogmatic and non-converging argument between entrenched "schools". I appreciate that loans can drive deposits, but struggle to see why an inflow of bank deposits would not encourage a bank to expand its lending as much as make its deposit terms worse, or why, ceteris paribus, an increased supply of reserves from the central bank (which necessarily goes hand in hand with a marginal fall in reserves-settled inter-bank lending rates) would not lead to an increase in the size of the banking system balance sheet and hence be manifested as a money multiplier.

    1. Tim,

      What I said is correct. Reserves are not required at the time a loan is made. They are needed to fund PAYMENTS, not loans.

      The "reserve requirement" argument is spurious. Many countries don't have have reserve requirements. And even in those that do, reserves are not obtained in advance of loan agreement. They are obtained ex post. Which is what I said.

      Increased reserves do increase the balance sheet of the banking system of a whole. But that does NOT necessarily mean a "money multiplier" effect in the sense of an increase in lending. Banks do not lend out reserves. The constraints on bank lending are capital and risk, not reserves.

      What an increase in reserves does, if it is not matched by a corresponding increase in lending, is raise the proportion of the asset side of the bank balance sheet that is devoted to safe assets. If reserves were increased enough (or QE continued for long enough), we would eventually have all deposits backed with reserves - i.e. full reserve banking. This does nothing whatsoever to increase lending. Indeed, as Peter Stella has argued, the "deadwood" effect of reserves on bank balance sheets may actually reduce it.

      The "money multiplier" theory is a mathematical construct that has no foundation in how lending actually works. It does indicate the relationship between base and broad money, but says absolutely nothing about the direction of causation. An increase in base money does not necessarily result in an increase in broad money. Banks are not passive intermediaries.

  9. "Reserves are not required at the time a loan is made. They are needed to fund PAYMENTS, not loans."

    Rubbish, Frances: reserves are required as a proportion of deposit liabilities - see eg here: If the loan creates a deposit, then reserves will be required within days if not immediately, whether that deposit has been used to make a payment or not.

    If reserves are required for balance sheet expansion, then reserves are clearly a factor of production in loan production (let's assume for simplicity that banks' business is to make loans, rather than, say, to buy securities), as is capital. If the cost of a factor of production falls, production can be expected to increase, and if the cost of reserves is related to the supply, it follows that there will be a positive and causal relationship between the stock of reserves and the stock of deposits (whether it is sufficiently close to being linear to be accurately described as a "multiplier" depends on the form of the relationships involved - eg elasticity). What is wrong with that?

    (of course, when a bank is holding an abnormally large amount of reserves for other reasons - eg as safe asset - their cost as a factor of loan production is effectively zero, so it would be foolish to expect QE to work through the money multiplier in those circumstances)

    1. Tim,

      You are confusing how reserve requirements are determined with how reserves are actually used. Reserve requirements do indeed depend on balance sheet size. But as I've already pointed out, not all countries have reserve requirements. And even in those that do, reserves are obtained EX POST.

      In normal times, creation of reserves lags lending. Therefore reserves cannot in any way be regarded as a "factor of production" of loans: rather, it would be correct to regard loans (or rather their associated deposits) as factors of production of reserves, since no central bank will refuse to create reserves if required to enable banks to settle loans already agreed (or meet reserve requirements if those exist). The stock of reserves DOES NOT determine the volume of lending, though their price does influence lending decisions. You have the causation the wrong way round.

      I really object to my remarks, which are founded on actual experience in banking and loan accounting, being described as "rubbish". I've warned you before about being rude to me. Be polite, or be blocked.

    2. I am not entirely clear what you mean by ex post, Frances; to me the key question is whether or not a bank can wait until a deposit created by a loan is drawn before it needs reserves, and my answer is that it cannot. That said, I think I have given enough explanation of my point of view to be happy for readers to make their own choice between our differing accounts.

      I apologise if my irritation shows when you respond to my initial argument by asserting the opposite, Frances. I am not without practical experience in these matters myself, but I try to give independent evidence like my link to the Fed website (and before the financial crisis, the ECB, BoJ and BoE also had reserve requirements) to back my arguments.

    3. Tim,

      "Ex post" means after the loan is agreed and the accounting entries have been made. Banks do not have to obtain reserves in advance of lending. They may not even have to obtain them in advance of settlement, if the central bank provides a daylight overdraft facility - as the Bank of England does, for example (free intraday repo).

      Your account hangs entirely on the existence of reserve requirements. But as I have already pointed out twice, not all countries have these. The Bank of Canada, Bank of Australia and Bank of New Zealand, for example, all had zero reserve requirements prior to the crisis. They aren't the only ones, either. You say the Bank of England had reserve requirements prior to the crisis. This is not entirely true: it had a voluntary reserve scheme, but to what extent a bank participated in the scheme was a matter for negotiation.

      Reserves do have to be topped up at some point. Exactly when depends on the regulatory requirements in a particular jurisdiction. But in NO jurisdiction do they have to be obtained in advance of lending. As I said before, it is not the stock of reserves that acts as a brake on lending, but their price. If funding costs are high, banks are likely to be cautious about lending.

      You are not the only person who is irritated. I find it irritating that you treat me as ignorant and allege that I am making things up. I assure you I am not. How reserve accounting works, the irrelevance of the money multiplier and the flaws in the loanable funds theory are not simply figments of my imagination. There is a considerable body of evidence regarding all of these - including from the Federal Reserve itself.