There have been a lot of posts about debt on this blog and the chart comparing government and household debt, which appeared in two of those debt posts, has proved particularly popular in discussion forums focusing on Australian property prices. Since producing the chart, the Australian government stimulus spending has continued to work its way through the economy and has been pushing up the levels of government debt. While I would still argue, as I have done many times before, that we should not follow the likes of Barnaby Joyce in getting agitated about public debt, it does seem worth updating the chart to illustrate recent developments. The regions shaded red denote Labor party governments in power.
Australian Government and Household Debt (1976-2010)
As expected, government debt levels exhibit a marked up-swing (note that the government data includes Treasury projections to the end of the current financial year). What is striking, however, is that the levels of household debt have not yet fallen. While some of the weakness in the economies of countries like the US and the UK is attributed to consumers “deleveraging” (a fancy term for paying down debt rather than buying flat-screen televisions), Australian households are showing no signs yet of reducing their debt. And 90% of that debt is for housing.
While it may not be evident here, there is in fact a tight relationship between debt levels in different sectors of the economy. If I spend money then either I reduce my financial assets (drawing on my savings) or I increase my liabilities (borrow on my credit card or some other form of debt). Exactly the reverse is true of whomever I give my money to (let’s call them Joe for argument’s sake): Joe’s assets go up or his liabilities go down. Spending money is an example of a “zero sum game”. If I add the change to my net worth (assets minus liabilities) to the change of Joe’s net worth it adds to zero. My negative change offsets Joe’s positive change. Aggregating over the whole economy, the sum is still zero.
Now consider what happens if we divide the economy’s net financial worth into that of the government sector, the private sector and the foreign sector (which includes overseas governments). Any changes in net worth across all three have to add to zero. As a result, the change in the government position is the opposite of the change in the private sector and international positions combined. If the government debt is going up, debt must be going down somewhere else. Now we know the household sector is not reducing debt, but what if we look at the private sector overall, including businesses? A different picture emerges.
Australian Government and Private Sector Debt (1976-2010)
Taken as a whole, over the 12 months to the end of 2009, private sector debt fell by about 2.5% of GDP. This was almost as much as government sector debt rose (about 3% of GDP). The difference can be explained both by the role of the foreign sector as well as slight differences in data collection methods across different sectors. Keep in mind that chart includes the government debt projections out to June 2010, while the private sector debt data only extends to the end of January 2010.
Since household debt has continued to increase, what this means is that Australian businesses have in fact been reducing debt significantly. The reduction in non-household private sector debt over 2009 was almost 7% of GDP. Businesses appear far more concerned about their debt levels than home-buyers do. It will be very interesting to see what happens once the first time home buyers scheme is fully unwound.
Government debt to 2008: A history of public debt in Australia
Government debt for 2009: Reserve Bank of Australia – Series E10
Government debt for 2010: Australian Treasury – Budget Estimate
Private sector debt: Reserve Bank of Australia – Series D2
Gross Domestic Product: Australian Bureau of Statistics – Series 5206.0
Possibly Related Posts (automatically generated):
- Park the Debt Truck! (16 July 2009)
- Unfounded liability (20 May 2013)
- Infrastructure Bonds (17 August 2010)
- Pinching Debt Data (22 May 2009)
The reduction in non-household private sector debt is certainly spectacular, but it isn’t driven only by the rise in Government debt. Businesses were forced by market conditions to cut debt (not least because the freezing of credit markets meant they could not roll debt forward).
Also, the household sector could well be just as concerned about their debt levels, but selling a house to pay off a mortgage is harder than raising equity or selling off brands to refinance loans. The household sector is likely to take longer to show a response than businesses. Perhaps focussing on faster elements of household debt (say credit cards) would provide more convincing evidence than overall debt does.
Exactly how does the household sector reduce it’s debt en masse when 90% of it is mortgage debt?
Wouldn’t house prices have to go on a sustained downslide first?
Or a wave of mass bankruptcies?
90% of the household sector’s debt is of the kind that you can’t simply pay down or reduce or just walk away from as borrowers could in the US.
It makes sense to me that unless house prices take a significant and sustained fall or we experience a a recession that creates an unemployment rate similar to what we see overseas, the household sector is going to have trouble making any great reduction in it’s debt.
Mark: you are certainly right that large businesses have far more capacity to reduce debt by raising equity and/or reducing dividend payments. It’s much harder for small businesses, but the size of large companies is such that they can have a big impact on the figures. It will be interesting to watch the household debt figures over the next one or two years.
Lefty: I certainly agree that it’s hard for mortgage debt to reduce rapidly, but I wouldn’t agree that the only way it can go down it for a sustained decline in prices. For a start, that in itself will not reduce debt and in fact may encourage bargain-hunters to come into the market, taking out new debt. Of course, if people then run into problem servicing their debt the bank may foreclose on them and sell their house and write off any shortfall. But it’s not house price falls themselves that lead to foreclosures, it’s borrowers’ inability to service their debt. Common causes of this are rising interest rates (which we are seeing) and rising unemployment (which we are not). The one way falling prices can trigger significant foreclosures, as we say in the US from 2007, is if there are a lot of borrowers drastically over-extending themselves on their debt in the hope that they will “flip” the property (sell it at a higher price) before having to service higher interest rates. My sense is that there has been far less of this sort of speculative buying in Australia than there was in the US up to 2006 and certainly the levels of “sub-prime” borrowing have always been lower here and have fallen further still over the last couple of years.
The ABS Housing Finance figures show that, although down somewhat from mid-2009, new more debt is still being added at quite a rate. A big reduction in aggregate mortgage debt would require a significant slowdown in new debt as well as reduction of existing debt.
“A big reduction in aggregate mortgage debt would require a significant slowdown in new debt as well as reduction of existing debt.”
This is what I was alluding to. If there is a significant and sustained fall in house prices then new entrants into the market will not require the same amount of debt to purchase. Might this not lead to a (steady) deleveraging of the overall sector over time as the value of new debt commitments falls (hypothetically – obviously there are many other possible factors that could be considered)?
Please, understand that I am writing as a Marxist. As I see things, and against popular perception (even among economists!), one of the central ideas of neoclassical economics is that there are such things as free lunches.
That’s why I find it refreshing when someone mentions “zero-sum games” in economics. As you know, but other readers might ignore, in a zero sum game one player’s loss is the other player’s prize.
Now, I would like to make people aware that the “household” curve in your Australian Government and Household Debt (1976-2010) chart, could also be labeled “financial sector assets (part of)”. This follows from your statements: when you buy your home, you’re borrowing money (which is a liability for you) from Joe’s bank (as assets).
And, to gain some perspective, the 4 Pillars issued some 90% of the mortgages included in that curve. They have a lot of people grabbed by their “cojones”. Is that power or what?
But it is also a weakness: imagine mortgagors started to default… The 4 Pillars may end up with a lot of loose (and bloodied) “cojones” in their hands.
Is it a zero-sum game? If I define “debt” to mean “liabilities” sure but is it? If not and I am a super-fund I can buy an asset e.g. a bank bill. My debt has not decreased as I use up my cash reserves. The cash reserves of a fund are a liability (?) but are they classed as “debt”. The bank debt has increased of course. The bank might lend this money to a corporate which will increase the corporate’s debt but not decreases the bank’s debt. I am a bit confused on how this works though it looks like it should be straightforward.
In any event it would be interesting to break the non-household debt figure down into government, bank and non-bank commercial.
On reflection it is a zero-sum game if net debt = debt – cash reserves (savings). Intracompany if I reduce my debt by paying down I have to use up cash reserves or raise equity, effectively swapping debt for equity. That equity comes from reducing someone else’s cash reserves so increases their net debt by same amount or they can raise the cash by increasing their debt which has same effect ad infinitum. It was the debt-equity exchange I wasn’t sure of.
I’m not sure I follow your reasoning, for which I apologize: home buyers with a mortgage have a debt with the banks that issued the mortgage. For home buyers, the debt is a liability. The aggregate amount of debt is shown in the “household” curve of the chart.
For the banks that issued the mortgages, these mortgages become assets. Thus, the “household” curve denotes the aggregate amount of assets.
Clearly, banks don’t need to hold those assets: they can sell mortgage backed securities. But this is another story.
If I am missing your point (as I am ready to admit that is likely) please, let me know.
On second thoughts, there is something inaccurate in my first post:
“And, to gain some perspective, the 4 Pillars issued some 90% of the mortgages included in that curve”.
That’s not so: the 4 Pillars were responsible for about 90% of the mortgages issued LAST year. Before that, they issued smaller percentages. Overall, I would guesstimate their holdings somewhere between 60% and 90% of the total.
Pingback: Are Australia's banks about to collapse? | Stubborn Mule
Pingback: Infrastructure Bonds
Pingback: Recognise this?
Pingback: Stilgherrian · Return of the Hallucinating Goldfish: Help!
So globally there seems to be more and more debt. What would happen if Governments ganged together and said “sorry” you can all go to hell to the “Puppet Masters”, I mean where the hell does all this money come from apart from China? Is there a “money crop” somewhere that only the “Masters” have access to, sorry for my ignorance, but this simply does not add up if the answer is zero.
You say that “If I add the change to my net worth (assets minus liabilities) to the change of Joe’s net worth it adds to zero. My negative change offsets Joe’s positive change. Aggregating over the whole economy, the sum is still zero”
This is not necessarily true because of the Fractional Reserve Banking system. Banks are empowered to lend up to a certain percentage of their liquid assets. Let’s say that percentage is 90%. As the money moves through the banking system $1000 deposited in a bank can result in TEN times that amount as debt. They create the money out of nothing. If I spend money (say in buying a house) that could result in the buyer borrowing more money from banks (created money) to buy an even more expensive house, thereby increasing the total debt.
I think this is the point that Steve Keen is trying to make on his Debtwatch website(http://www.debtdeflation.com/blogs/). It is not a zero-sum game, unfortunately.
Jeff. The whole question of fractional reserves and multipliers is a different issue. What I have written here is completely consistent with the arguments Keen and others make about money creation through bank lending.
The key is to understand what is referred to as “money”. If a bank lends me $100, which it credits to my transactional account, then the bank has a new $100 asset (their loan to me) and a new $100 liability (the deposit in my transactional account). I might then spend the money at a shop which uses the same bank as me. The $100 moves from my account to the shop’s bank account, but it is still a $100 liability for the bank. Money is generally taken to include deposit balances in transactional accounts, so the initial loan to me “created” a new $100. This is the sort of credit-driven money growth Keen talks about. However, in my post I was talking about “net worth” not “money”. This is the net of assets less liabilities, and includes both assets and liabilities. The bank’s liabilities net to zero (they have a new $100 asset and a new $100 liability), while I have a new $100 liability (my loan) and the shop has a new $100 asset (the deposit in their account) and if I aggregate over the whole economy, the $100 increase in my liabilities is offset by the $100 increase in the shop’s assets.
So, “money” may not be a zero sum game, but financial “net worth” is.
Thanks for your quick reply, Stubborn Mule.
With respect, I don’t think you and Steve Keen are on the same page. I base this comment on his article in BRW (http://www.brw.com.au/p/sections/featur/keen_to_be_heard_ibhMdopX0E8Soh00ql3mPO)
which I quote:
“Traditional economists do not take any account of debt in their modelling,” he says. “In their world, debt simply doesn’t matter.”
Keen quotes Nobel Prize winning economist Paul Krugman to illustrate the point. In his 2010 paper entitled Debt, Deleveraging and the Liquidity Trap, Krugman said: “. . . looking at the world as a whole, the overall level of debt makes no difference to aggregate net worth – one person’s liability is another person’s asset.”
Yet Keen argues that, while this is empirically true, it is also irrelevant.
“That view is fine if you have money being lent from one person to another, because the lender has less money and the borrower has more and everything ends up equal,” he says. “But this completely ignores the role of banks.
“If somebody goes to a bank and wants to borrow money, the bank effectively creates that money out of nothing. It doesn’t have to take cash out of somebody’s savings to get it. And by doing so it injects extra cash and potential demand into the economy without subtracting spending power from anywhere else. In that way, overall demand is boosted and things can get out of kilter.”
I would add that total net worth in the economy thereby increases.
I think I follow your reasoning, up to this point:
“I would add that total net worth in the economy thereby increases.”
Why would net worth increase?
@Jeff: I think you have missed the point of this post. I am not saying that an increase in private sector debt is unimportant. In fact, I think it is very significant and the current environment of the “cautious consumer” is only a relatively mild consequence of the beginnings of deleveraging. What I am arguing is that net financial assets are a zero sum game so we should not expect public and private sector debt to reduce at the same time (strictly speaking, this would be possible if we ran a current account surplus with the rest of the world, which we do not).
To understand the point, you must more clearly distinguish the following concepts: debt, money and net (financial) worth.
Thank you for your patience. I guess I am putting a lot of weight on what Steve Keen says here:
“If somebody goes to a bank and wants to borrow money, the bank effectively creates that money out of nothing. It doesn’t have to take cash out of somebody’s savings to get it. And by doing so it injects extra cash and potential demand into the economy without subtracting spending power from anywhere else.” That is, the assets are not balanced by liabilities in this instance.
I am thinking that when banks lend ‘money out of nothing’ and do not take it out of somebody’s savings that cannot be a zero sum game. Or am I wrong? If I am right, then wouldn’t aggregate net worth increase across the economy because more ‘assets’ are purchased without counter balance on liabilities?
Consequently, when private debt goes up debt does not have to go down anywhere else in the system, neither government nor international. They change independently of each other. Which, I think, is the point you are making so perhaps we are not very far apart in our conclusion.
Apologies if I have still missed the point.
Okay, Stubby, for what’s worth this is my take on Prof. Keen’s views.
Bailey’s Home Loans
@Magpie: I’d more or less agree with your post. I’ll post my one small quibble over on your site.
@Jeff: if you read Magpie’s post, you will see that (financial) assets and liabilities do always balance. However, in saying that, I do not think that there is no significance to increasing levels of debt. Although bank lending does not create net financial assets, it can shift around the patterns of demand (the borrower may want to buy things but without the loan would not, while others with excess financial assets may only want to save). So, I would agree with Keen that rapidly growing debt levels are something we should worry about. You only have to look at the financial crisis to see that! I think the point that Keen makes is that some people say that lending does not create net financial assets (X) and we should therefore not worry about it (Y). The response to this should be that Y does not follow from X rather than trying to refute X.
I am afraid I am not sure what the quibble is. Is it about the “deposits created out of thin air” thing or about the interbank operations?
It’s about “loans create deposits, so banks don’t need existing deposits to lend”. That can be the case (when the deposit stays with the same bank), but it isn’t necessarily the case and the bank itself cannot control whether it will be the case. This means that banks do have to worry about sourcing deposits (or wholesale funding) to fund their lending.
Somewhat related, I’m not keen on the terminology sometimes used “loans create deposits, not the other way around”. Even when they are created “out of thin air”, I see the deposit and loan (liability and asset) as both coming into existence simultaneously.
Let’s start with the message from 7:49am.
You seem to say that, if both parties in a transaction (Old Sven and Bo, say) don’t bank in the same bank, individual banks in an economy cannot create deposits through loans.
Is that it?
@Magpie: the issue I have is not with the simultaneous creation of loans and deposits (I still prefer to think of the two coming into existence simultaneously!) but rather the implication that this means that Bailey’s bank does not have to worry about deposits.
Consider the case where Old Sven banks with Mule bank. Bailey will lend to Bo and to settle the transaction, will advance funds to Mule bank, instructing Mule bank to credit Old Sven’s account. The question is, how does Bailey provide funds to Mule bank? Mule bank doesn’t just want a deposit with Bailey! One avenue would be for Bailey to move funds between settlement accounts with the central bank. The result would appear as follows:
assets +$900 (Bo loan)
assets -$900 (reduced balance with central bank)
net change $0
assets +$900 (increased balance with the central bank)
liabilities +900 (increase in Sven’s deposit balance)
net change $0
Note that there has still been creation of a loan and a deposit from thin air, even though it’s spread across both banks. So far so good. But what if Bailey doesn’t have $900 in its central bank account? Then Bailey will have to borrow the money, either by getting a deposit from someone else or even from borrowing from Mule bank. The latter would look like this:
assets +$900 (Bo loan)
liabilities +$900 (inter-bank loan owed to Mule bank)
net change $0
assets +$900 (inter-bank loan)
liabilities +900 (increase in Sven’s deposit balance)
net change $0
Loan and deposit still created, but if Bailey had not found anyone to source the funds from, the whole transaction could not have happened.
One last point: it is also possible for no deposit to be created. Let’s say Sven had already taken out a loan from Mule bank to buy the house in the first place and used the proceeds from Bo to reduce the loan:
assets +$900 (Bo loan)
assets -$900 (reduced balance with central bank)
net change $0
assets +$900 (increased balance with the central bank)
assets -$900 (decrease in Sven’s loan balance)
net change $0
This time, no deposits have been created and there has also been no net change in loan balances across the system: Bo’s increased, Sven’s decreased.
Thanks for the reply, Stubby.
Let me think about the issues raised, because I believe (but I am not guaranteeing it, yet) there is a solution to the problem you posed here.
But, from your last message, I think we can conclude that the multiple-banks issue does not change the basic conclusions one gets when considering only one bank. Are we in agreement here?
In fact, if I am right, it could be a rather obvious solution, considering the nature of the problem Keen and Krugman debated.
But I don’t want to get ahead of myself. Let me get things clear and I can convince myself.
@Magpie while I’m waiting for your further thoughts, I’ll make a couple of summary observations:
* Bank lending does typically simultaneously increase the total lending and deposit balances across the system (the exception arises when payments are used to repay other debts).
* This fact does not mean that any individual bank can happily lend to its heart’s content without worrying about how to fund that lending.
To repeat my question:
“But, from your last message, I think we can conclude that the multiple-banks issue does not change the basic conclusions one gets when considering only one bank. Are we in agreement here?”
@Magpie: that depends on what the basic conclusions are! If they include “”Bailey’s did not need to have the deposits before making loans. It could have happened, but it wasn’t necessary” then I don’t agree (if Sven banks elsewhere and Bailey’s doesn’t have the funds and cannot find anyone to provide them, it cannot lend).
If the basic conclusions are simply that, whether in a single bank or spread across the system, lending is typically associated with deposit growth “from thin air” then I do agree.
This is an aptly named blog! :-)
@Magpie: truth in advertising! But I do think we agree on the core points. Also, linking back to the original thread from Jeff, I also agree with Keen that unfettered growth in (private sector) debt is dangerous.
Yes, I agree. It seems that Banks cannot fund all the lending they would like from Australian deposits alone so they go to the international money market and borrow. In Australia they lend long (mortgages and long-term contracts) but they borrow short overseas. Constant re-financing of overseas loans is needed at interest rates that keep changing. They are now so reliant on those funds that they pay more attention to those rates than the Reserve Bank cash rate. Banks sometimes lend to people who borrow their max and when rates rise they cannot meet repayments. They may also default if they lose their job. When this happens on a large scale the financial system collapses, asset values fall. Derivatives accelerate and accentuate the process. Anyway, this is my simple minded and maybe flawed understanding of what can happen. I interprete Steve Keen as warning against this unbridled lending to the private sector.
@Jeff: there is no doubt that deposits are a bigger issue for banks here in Australia than they have been for a very long time. ANZ’s Phil Chronican recently spoke about the link or lack thereof to the RBA cash rate. One of the points he makes there is that the cost of funds for banks at the moment is being driven at least as much by intense competition for customer deposits as by the cost of offshore funding. A few years ago, deposit rates for customers were often below the RBA cash rate and now they are well above.
Excuse my ignorance, but if the RBA cash rate is below deposit rates for customers why don’t banks borrow from the RBA?
@Jeff: the workings of the RBA are a bit arcane, so no need to apologise!
There is information here and here about the RBA’s “open market operations”. When the cash rate is set by the RBA, it is really a target they set. They will pay 0.25% less than the target rate on balances with the RBA and will charge 0.25% more than the cash rate for loans. If banks do borrow from the RBA, it must be secured by collateral (high-quality bonds, such as Government bonds). In practice, the idea is that most of the lending would take place directly between banks rather than via the RBA. By setting the target rate, the idea is that banks will borrow and lend to one another at a rate close to the target. In practice, this interbank lending takes place at rates within 0.01% of the target.
You might still wonder why banks would pay for expensive deposits rather than borrow from other banks at the overnight rate (essentially the RBA rate). This is where the term of the deposit becomes important. The high deposit rates banks are paying are for term deposits: particularly 6 months or longer. It would be very risky for a bank to have all its liabilities due and payable the very next day. This risk is “liquidity risk” and, in addition to a bank’s own desire to manage its risk, there are regulations (and more coming) that restrict how many of a bank’s liabilities can have short term maturities. Bank’s manage this risk by issuing long-dated bond (usually 3 to 5 years) and paying higher rates on term deposits than on at-call deposit balances. The fact that the term of borrowing is important and influences the rates pays, underscores the fact that it is a bit of a historical oddity that mortgage rates (typically 25 year loans) should be linked to the RBA cash rate (an overnight borrowing rate).
Ouch! Some people think that rocket science is bewilderingly complex, but it is a model of order and simplicity compared with this financial and economic stuff.
Bank finances are not rocket science………they are more complex!
@Jeff: I think one of the reasons it can get so complex is that most people only ever deal with a small part of the process: very few people ever try to understand the whole.
I am thinking that Steve Keen’s warning about high private debt is a serious problem because things never stay the same.
If interest rates and employment levels do not change much and all the borrowers are able to easily service their debts then the level of total debt doesn’t matter. Unfortunately that is not the real world. At low interest rates banks fall over themselves lending to companies and people. The do cursory valuations on collateral and lend to the limit of borrowers capacity to service the loans. When things change – interest rates go up, GFC, or unemployment rises then defaults can occur at a massive level and the value of collateral does not cover the value of loans.
Steve Keen got the housing price collapse wrong because people were able to service their loans due to the GFC being less severe in Australia in 2008. However he may be right next time, when Greece and Spain default.
Do you agree?
There’s no doubt that a combination of higher interest rates and higher unemployment would do some damage to the housing market. Unlike Steve, I never thought that the Australian property market was very likely to crash dramatically. I was in the camp of a more gradual decline. It has taken some time, but that is what it seems to be doing now.
As for Europe, with “austerity programs” everywhere, and a very weak private sector unable to replace the Government spending that has been removed, the economic prospects for the region seem grim indeed. Spain in particular already has devastatingly high youth unemployment. I really struggle to see how it can end well. There could be significant social unrest coming to Europe.
“@Magpie: that depends on what the basic conclusions are! If they include ‘Bailey’s did not need to have the deposits before making loans. It could have happened, but it wasn’t necessary’ then I don’t agree (if Sven banks elsewhere and Bailey’s doesn’t have the funds and cannot find anyone to provide them, it cannot lend).”
I know this will make no difference, but, what the hell: if (1) Sven banks elsewhere and (2) Bailey’s doesn’t have the funds to transfer to whatever bank Sven banks with, what makes you believe Bailey’s wouldn’t normally be able to find anyone who could lend them?
Further, let’s assume banks normally don’t find anyone to cover their debts, what do you think happens then? Do transfer just “bounce”?
@Magpie: apparently I missed this reply all those months ago!
Transfers can bounce: that’s essentially what happened to Lehman!
I am surprised by this observation: “Transfer’s can bounce: that’s essentially what happened to Lehman!”
Of course that’s what happened to Lehman. But I don’t think what I said could be interpreted as saying that this is a logical impossibility.
From the comment you are presently replying to:
“What makes you believe Bailey’s wouldn’t NORMALLY be able to find anyone who could lend them?”
This is how The New York Times (September 12, 2008) reported the event:
“One observer briefed on the situation described the session as a ‘game of chicken’ between the government and the heads of the major banks”.
This is the game of chicken they were playing: Geithner wanted the banks to put money to save Lehman; the banks wanted the Fed to put money to save Lehman. If no one veered, they both would crash head-on. No one veered.
Firstly: cases like this don’t NORMALLY occur. That’s what made the crisis so severe: Lehman was extremely leveraged.
Secondly, when cases like this do occur, NORMALLY one side veers (in the previous ten years, the article cites only a similar case where the banks veered: Long Term Capital Management; I can cite another: Bear Sterns, where they also veered).
This was not a normal case and I am sure who have heard the same argument before.
One can create a system where no one could normally jump in front of an incoming train; but a person determined enough could still find a way. They just found it.
This is the NYT report:
U.S. Gives Banks Urgent Warning to Solve Crisis (September 12, 2008)
@Magpie: you are right that this doesn’t normally occur, but lenders will still worry about it and charge accordingly. That is part of the reason for why a bank has to pay more interest when it issues a 5 year bond than it does when it borrows money overnight. During the thick of the financial crisis, it became almost impossible for banks in Europe to borrow money even for a few months, which is why the ECB had to step in and provide “liquidity” (i.e. lend to banks). This also means that it would be very risky for banks to have all of their liabilities in a very short term form: if they were all falling due around the same time and this happened to be a tough time for the bank to borrow, the bank could collapse. For this reason, regulators impose strict rules limiting the amount of short-term liabilities a bank could have (even if they didn’t it would be prudent risk management on the part of the bank to pay up for a reasonable amount of longer-dated liabilities). Broker-dealers like Lehman Brothers were not subject to these rules and had a very significant amount of short-term funding (much of it overnight only).
At times longer-term bond markets can be difficult to access and if a bank already has as much in the way of short-term borrowing as it is allowed to have, even in relatively normal times (by that I mean that the bank itself is not about to collapse, but bond markets might be shaky as they have been on and off this year), a bank might decide to constrain its lending because it is concern about raising the liabilities.
I think I am missing your point and I am sort of confused.
“You are right that this doesn’t normally occur, but lenders will still worry about it and charge accordingly.”
Okay, it cost higher interests. Yes, sure. So? Weren’t we discussing liquidity? At least I thought that the objection was liquidity. Are we now discussing the bank’s profitability, too?
“This (lack of liquidity for interbank loans) also means that it would be very risky for banks to have all of their liabilities in a very short term form.” (…)
“If a bank already has as much in the way of short-term borrowing as it is allowed to have (…) a bank might decide to constrain its lending because it is concern about raising the liabilities”.
I have no problem with that. Clearly, apart from reserve, banks have other legal requirements (for instance, relating capital to loans base). Incidentally, all these are legal/statutory constrains, not consequence of the way the reserve system works; but they are real, nonetheless.
So, if a bank has borrowed short term as much as it is allowed to, well, it shouldn’t be able to borrow anymore, by definition. But this means it may be forced to stop lending, even if it can create money “out of thin air” (sorry, I couldn’t help it!).
More seriously: yes, banks need to keep an eye on the term structure of assets and liabilities. Apart from bank regulations, interbank short-term loans cost money and an adequate management of assets/liabilities should avoid these costs.
But even if they weren’t constrained, they might still prefer not to lend and speculate directly in financial markets, instead (as I understand is happening in many places now).
Having said all that, how does this invalidate the general postulate that banks create money? (I am assuming this is your contention; is it? If not, then what is it?)
@Magpie: I do still agree with the general postulate that bank lending creates money. What I am less comfortable with is the broader statement which is often made (and I sense, perhaps incorrectly, that you are saying something similar) that this implies that as long as there are credit-worthy borrowers this means that banks can always lend because they create the deposits to do the lending. To lend to a credit-worthy borrower and individual bank may have to pay a rate of interest on their liabilities which is higher than the borrower is prepared to pay (or that other banks are prepared to charge).
Note that the relevance of profitability is that liquidity is always available, but it may come at a price that is too high to pay.
I should add that if they do not lend, banks have options other than speculating: they can pay down/buy back expensive liabilities.