Take for instance company X, which has equity of $200. The company borrows $800 and buys an asset worth $1000. In this example we can say that company X has a leverage of 5. That is to say the equity of $200 represents 1/5 of the $1000 asset. If the value of the asset falls by 10%, that is, to $900 given the debt of $800, it implies a fall in equity or net worth to $100 — or 1/9 of the $900 asset, raising the leverage from 5 to 9 and making company X less solvent.
Let us now assume that company X has decided to deleverage and to lower its leverage back to 5 (by doing this the company will become more solvent). To achieve this company X sells assets for $400 and reduces its debt to $400. Consequently company X will now have $500 in assets, $100 in equity and $400 in debt: the leverage is now 5 again. But if many companies try to lower their leverage then there is a risk that the value of assets will fall. If, for instance, the value of company X’s assets falls by 10% to $450, then, given the value of debt of $400, net worth falls to $50 implying that the leverage goes back to 9.
Some commentators are of the view that pressure on European banks to bolster capital could force them to cut assets. However, by cutting lending — trimming their assets — banks are forcing various borrowers to sell off their assets to prevent insolvency. Consequently this sets in motion asset-price deflation. This in turn lowers borrower collateral and causes banks to reduce their lending further, etc.
It follows that if all financial institutions are doing the same thing (trying to fix their balance sheets), they could drive asset prices down, which for a given debt will shrink their net worth and increase their leverage, or make them less solvent. This is the paradox of deleveraging. If this process is not arrested in time it could seriously damage the real economy, so it is held.
So what should be done here? According to popular thinking, the central bank or the government must step in and start buying the assets that banks are trying to get rid of. This, it is held, will stop the asset-price deflation and prevent the nasty dynamics that can ruin the real economy.
Some commentators are of the view that “the paradox of deleveraging,” follows the same principle as “the paradox of thrift,” which was put forward by John Maynard Keynes.
The “paradox of thrift” states that if everyone acts more carefully with his money and saves more, then this will lower aggregate demand, which in turn will lead to a fall in economic growth. As a result, total savings in the economy will actually decline.
In this way of thinking, spending by one individual is income for another individual and it follows that if all individuals were to increase their savings — that is, lower their spending — the overall income in the economy will fall. A smaller income will permit less saving.
From this it is concluded that if people are hesitant to spend, then the government must step in and lift overall monetary spending to prevent the economy from falling into a recession.
In his writings, Keynes relied on the ideas of Bernard Mandeville to provide credence for the “paradox of thrift.” According to Mandeville,
As this prudent economy, which some people call saving, is in private families the most certain method to increase an estate, so some imagine that, whether a country be barren or fruitful, the same method if generally pursued (which they think practicable) will have the same effect upon a whole nation, and that, for example, the English might be much richer than they are, if they would be as frugal as some of their neighbours. This, I think, is an error. 
To reinforce his view that saving is bad for economic growth, Keynes also quotes Malthus:
I distinctly maintain that an attempt to accumulate very rapidly, which necessarily implies a considerable diminution of unproductive consumption, by greatly impairing the usual motives to production must prematurely check the progress of wealth… But if it be true that an attempt to accumulate very rapidly will occasion such a division between labour and profits as almost to destroy both the motive and the power of future accumulation and consequently the power of maintaining and employing an increasing population, must it not be acknowledged that such an attempt to accumulate, or that saving too much, may be really prejudicial to a country?
Monetary Expenditure and Real Savings — What is the Connection?
Observe that spending, saving, and income in this way of thinking is in terms of money. We suggest that what matters as far as real economic growth is concerned is not monetary expenditure, as such, but real savings. It is real savings and not money that funds tools and machinery, i.e., capital goods, which in turn permits the expansion in real wealth.
Monetary expenditure as such does nothing as far as formation of real savings is concerned. By means of money a wealth producer exchanges goods that he has produced for the goods of another wealth producer. So in this sense, payment is always with goods. Money just makes it possible to exchange various goods. Note that without the existence of goods there cannot be any exchange.
Contrary to popular thinking, the heart of credit is not money but saved final goods and services. If John the baker produces ten loaves of bread and consumes one loaf, his saving is nine loaves of bread. The baker’s saving now permits him to secure other goods and services.
For instance, the baker can now exchange his saved bread for a pair of shoes with a shoemaker. Observe that the baker’s saving is his real means of payments — he pays for the shoes with saved bread. Likewise the shoemaker pays for the nine loaves of bread with the shoes that are his real saving.
The baker may also engage in a different transaction with the shoemaker. He could lend him nine loaves of bread in return for ten loaves of bread in one week’s time. Note that the loaned bread sustains the shoemaker and allows him to continue making shoes. After one week, hopefully, the shoemaker has produced enough shoes to be able to secure the ten loaves of bread needed to repay the baker the loan of nine loaves of bread plus the interest of one loaf.
The introduction of money in our story doesn’t alter what has been said so far. Without the medium of exchange — money — no market economy could take place. By means of money people can channel real savings, which in turn permits the widening of the process of real wealth formation.
Whenever an individual lends some of his money, he transfers the medium of exchange to a borrower. By means of money, the borrower can access the existing pool of final goods and services. By means of money, the borrower can now secure real savings (final goods and services) that will support him while he is engaged in the production of other goods and services. Note however that the final goods must already be in existence for the exchange to take place.
In the same way that real savings sustain the producers of final consumer goods, such as the shoemaker in our example, savings also fund the production of tools and machinery, which in turn permits the expansion of final goods and services. This increase in turn permits a further increase in savings that can now support the buildup of a more sophisticated production structure, which in turn permits a further expansion in the production of final goods and services. In this sense real savings are the key to economic expansion. This is contrary to popular thinking, which argues that savings can be bad for economic growth.
Observe that popular thinking reaches erroneous conclusions because it is only concerned with monetary flows without paying attention to real stuff. Again, for mainstream thinking what matters is monetary spending, since the more people spend, the greater the monetary income is going to be. From this it is concluded that an increase in saving, which is less monetary spending, must be bad news.
But does it make sense to suggest that people save money?
Demand for Money and Savings
People don’t save money as popular thinking suggests but rather exercise a demand for money. Once real savings are exchanged for money the recipient of money can exercise his demand for money in a variety of ways. This however will not have any effect on the existent pool of real savings.
An individual can exercise his demand for money either by holding it in his pocket or in his house or by placing it in the custody of a bank in a demand-deposit or even in a safe-deposit box. Whether the individual lends out his money or puts it under the mattress, it does not alter the given pool of real savings.
By putting the money under the mattress, an individual doesn’t engage in an act of saving – he is just exercising a demand for money. What individuals do with money cannot alter the fact that real savings are already funding a particular activity. (Whether individuals decide to hold onto the money, or lend it out alters their demand for money, but has nothing to do with savings).
Whenever an individual lends some of his money to a borrower this means that he transfers a medium of exchange that the borrower could activate in order to secure real goods and services. By lending money, the individual has lowered his demand for the services of the medium of exchange. Conversely the borrower has increased his demand for the services of the medium of exchange.
Note that the act of lending money doesn’t alter stock; he simply transfers the services of the medium of exchange to the seller of financial assets — no present real savings are affected as a result of these transactions.
Also note again that people don’t demand money to hold it, as such, but rather to use it in an exchange. Even if prices are going down, it doesn’t follow that people will start hoarding money. They will still continue to use it to maintain their life and well-being.
The greater the expansion in the production of goods and services, the greater the demand for the medium of exchange is going to be. Again the increase in the demand means an increase in the demand for the services of the medium of exchange to enable a greater amount of goods and services to be exchanged.
Once it is realised that saving is real stuff and has nothing to do with money as such the so-called paradox of thrift turns out to be a logical impossibility. If we have two bakers and each of them has increased savings from five to nine loaves of bread, collectively we cannot have less then eighteen loaves of bread saved as the paradox of thrift implies. We can also conclude that, contrary to the “paradox of thrift,” the increase in savings is the key for economic prosperity.
Is Deleveraging Really Bad for the Economy?
The existence of banks enhances the use of real savings. By fulfilling the role of middleman, banks make it easier for a lender to find a borrower. When a bank lends money, it in fact provides the borrower with the medium of exchange that can be employed to secure goods and services.
What determines the flow of lending is the flow of real savings. If the baker were to consume his entire production of ten loaves of bread then there will be nothing left for lending.
It is therefore futile to urge banks to lend more if real savings are not there. Likewise it doesn’t make much sense to suggest that a central bank can somehow replace nonexistent real savings — in this case nine loaves of bread — by printing more money. (It is also an exercise in futility to raise government spending to fix the problem. After all if a government spends more it means that somebody else will have less resources left.) All that adding more money to the economy will do is to weaken wealth generators and thereby reduce the future supply of real savings and weaken future real economic growth.
So what then are we to make of the “paradox of deleveraging”?
Is it true that if every bank were to attempt to “fix” its balance sheet, the collective outcome would be disastrous for the real economy? On the contrary, by adjusting their balance sheet to reflect true conditions, banks would lay the foundation for a sustained economic recovery. After all, by trimming their lending, banks are likely also to curtail the expansion of credit “out of thin air.” It is this type of credit that weakens wealth generators and hence leads to economic impoverishment. Contrary to the proponents of the “paradox of deleveraging” we can only conclude that if every bank were to aim at fixing its balance sheet, and in the process curtailing the expansion of credit “out of thin air,” this would lay the foundation for a healthy economic recovery.
For most commentators, it is a major threat to an economy if banks are curtailing their expansion of credit in order to improve their net worth and hence solvency. This, it is argued, sets in motion a vicious process that leads to asset-price deflation, which for a given value of liabilities actually weakens banks’ net worth and makes them less solvent.
When all banks are trying to “fix” their balance sheets, the outcome could be the exact opposite of what they intended to achieve, so it is held. So what should be done to arrest this vicious process? According to popular thinking, the central bank and the government must step in and start buying assets that banks are trying to get rid of.
Note that this conclusion is in line with the writings of Keynes. Contrary to this way of thinking we have concluded that, by adjusting their balance sheets to the facts of reality, banks actually set up a process that permits sustained economic growth.
On this Ludwig von Mises had the following to say:
The unprecedented success of Keynesianism is due to the fact that it provides an apparent justification for the “deficit spending” policies of contemporary governments. It is the pseudo-philosophy of those who can think of nothing else than to dissipate the capital accumulated by previous generations. Yet no effusions of authors however brilliant and sophisticated can alter the perennial economic laws. They are and work and take care of themselves. Notwithstanding all the passionate fulminations of the spokesmen of governments, the inevitable consequences of inflationism and expansionism as depicted by the “orthodox” economists are coming to pass. And then, very late indeed, even simple people will discover that Keynes did not teach us how to perform the “miracle … of turning a stone into bread,” but the not at all miraculous procedure of eating the seed corn.
A version of this article was previously published at BrookesNews.com on 25 August 2008. The same ideas were presented in article for Mises.org yesterday: Will the Latest Plan “Fix” the Eurozone?.
 Paul McCulley The Paradox of Deleveraging in PIMCO Market Commentary July 2008
 John Maynard Keynes, The General Theory of Employment, London Macmillan &Co ltd 1964 p 361 (Notes on Mercantilism,etc.).
 Ibid, p362-363.
 Ludwig von Mises, Lord Keynes and Say’s Law – in The Critics of Keynesian Economics edited by Henry Hazlitt D.Van Nostrand Company, Inc p 320-321.
The analysis presented here on the paradox of deleveraging is flawed. Only the effect on asset prices is considered, and not the effect on the money supply and incomes. I will illustrate this with an example that I posted on this site a few weeks back.
In a normally functioning economy, households spend most of their income, and save a proportion of it which is then lent out to businesses and invested. Thus total income is equal to total spending. But if, for example, businesses all begin to pay down debt at once, there is no one to borrow the funds that households have saved.
Say a household has £2000 in income and saves 10 percent of this, or £200. If all companies are paying down debt, there is no one to borrow and spend this £200, which therefore stays in the banking system. So only £1800 of the £2000 can become income for someone else. Assuming the next household also saves 10 percent, they will spend £1620 of the £1800 and save £180. In this way, the initial £2000 keeps reducing, and money piles up in the banks, which depresses the economy.
So in this scenario, aggregate demand decreases by the sum of net household savings plus net debt repayment by firms. The money supply will also plunge as debt is repaid and not lent out. This is clearly a fallacy of composition – a course of action which makes sense for an individual firm, but not for all collectively.
This decrease in aggregate demand would be bad news for profits of many businesses – households are getting poorer and poorer and will have less money to spend than they did – many small businesses (e.g. Restaurants) would be unable to adjust to this sudden drop in demand and will have to go bust.
This shows that mass deleveraging is clearly bad news for the economy and is harldly the key to economic prosperity. It also shows that the statement “It is also an exercise in futility to raise government spending to fix the problem. After all if a government spends more it means that somebody else will have less resources left” is false; mass deleveraging means money sitting in the banking system not being used, so government spending isn’t going to take resources away from anyone.
If money were inelastic (based on a true gold standard, say) would you still feel that the government would need to intervene in the face of deleveraging? Or is it only necessary because of our system of fractional reserve banking?
That’s an interesting question. If by a true gold standard you effectively mean full reserve banking, then I think deleveraging would still be a problem.
A full reserve system would not really change the example that I outlined. There would still be the problem of total spending being less than total income, if all businesses were paying down debt and not borrowing. This would still mean household savings piling up unused in the banking system and not being lent out, depressing the economy.
If the government borrowed this unused money and spent it, this would ensure that total spending remained equal to total income, preventing a crash in the economy.
The only difference would be that the money supply would not decrease in this case as we are talking full reserve.
You’re becoming worse than me for recycling your old material ;)
In that short-term that’s all quite correct.
And that is the difficult bit :)
Be wary of first-order models where all debts are money substitutes. Banks have far more assets than they have outstanding money substitutes. That’s because they have a great many other liabilities, such as fixed-interest bonds. If there is a great demand for money then bank account holders will put up with fewer free services from their bank, they may even pay fees. Therefore banks will serve them ahead of bondholders who demand interest. As I’ve pointed out before you can easily see this in the Bankstats data from the Bank of England, there are periods where the size of bank balance sheets and the money supply move in opposite directions.
Money deflation is caused by monetary policy or by the banking industry being in disarray because of insolvencies and fear of insolvencies.
Money isn’t a resource is it? Money is a form of debt. If gold coins are lying idle then it’s quite true that that is a resource lying idle. But, a bank balance is quite different it’s a debt that a bank owes to it’s customers. The bank supports that by it’s assets, which are loans and collateral. Whether they’re “lying idle” is a different matter.
What you’ve described here is what Krugman often calls “Part 1 Keynesianism”. When I first started off learning about economics I thought it was right for a while, then after that I thought it was wrong. My view now is that it’s too incomplete to really say it’s right or wrong.
Normally I talk a lot about monetary theory and how monetary calculation can cause systematic errors. I think that’s important, but it’s not really the same thing as the Part 1 argument you’re giving here, so in this post I’ll act as though I’m not a monetary equilibrium guy.
One obvious problem is that households may not reduce their spending by the same amount as their reduction in income. Friedman made this point in his PIH, but I think he relied too much on it, households may still reduce their spending even if they don’t reduce it in proportion to income. So, all this line of thinking shows us is that the most extreme Keynesian views are unrealistic.
A while ago I posted a reply I gave to Greg Hill on a similiar topic on the thinkmarkets blog. The problem I concentrated on their was the idea that GDP shouldn’t fall. There are situations where it is appropriate that businesses cut back on output because they are not producing goods that consumers demand. To put it another way, we view output positively because we presume the utility of the output good exceeds the utility of the inputs in other uses. If that condition isn’t met then output should not be view positively. It’s entrepreneurs and customers who jointly judge this, it doesn’t happen automatically. The same sort of thing is always true of assets whether they are held by businesses or private individuals.
This is my problem with the argument that Keynesian stimulus could have prevented the housing bubble causing a recession in 2008. If there was misallocated capital then production plans needed changing. There must be “recalculation” as Arnold Kling has put it, where new uses must be found for capital and labour. Assets that are stored may be “idle” in the simple sense because no use has been found for them in a production process that would yield a profit.
At this stake Keynesians always presume that this must be caused by a fallacy of composition, the explanation you give here. You guys presume that the initial shock must be irrational. If it is irrational then no course of correct is needed by the wider economy. But, if it’s a considered response to circumstances then that sort of thinking doesn’t necessarily apply. If I find my house is worth £50K less than I thought because of overbuilding of housing then I may consider that I need to increase the quantity of assets I own to provide for my future. That means I need to save to repair my intertemporal plan. Keynesian stimulus can’t compensate for this fall in real wealth. At best it can trick me and those like me into thinking my assets are worth more than they really are for a while.
Something quite similar applies to debt. You talk about “households getting poorer and poorer” while debt is being paid down. This presumes that the short-run aggregate demand effect is so overwhelmingly large that the fact that debt is being paid off isn’t significant. This is one place where expectations come in…. Suppose I owe £3000 and I have the option of paying it down now or spending the money. If I pay it down now then I will be able to spend more in future periods, and be free of the frictional costs associated with debt. But, in the current period my actions will reduce aggregate demand, if there is no compensating change. If the return on investments is viewed rationally then an investor should prefer the scenario where I pay down my debt, if it increases the net present value of my spending stream on products related to his investment. Or to put it another way, consumers who have heavy burdens of debt will spend less in future periods depressing expenditure then, rational investors will realise that beforehand. I was reading about the levels of private and state debt in the UK today, as an small investor I’d be very happy if those levels fell and it would encourage me to make long-run investments.
It’s quite true that some businesses (such as the restaurant you mention) may be cashflow limited, that’s similar to cashflow limitation in the PIH argument. Even if such businesses do exist it isn’t clear that the overall effect of paying down debt will be negative over a large number of periods.
I could bring up the liquidity trap issue here too. What you’re assuming here is that extra funds supplied to the savings-investment market will not change the way it operates. We’ve argued many times about that and I don’t want to repeat that all over again. What I will say though is that I don’t think changes of that sort are immediate or perfectly elastic. There are many markets that can’t respond quickly to a change in conditions. One way to think of this is that only a small area of the supply-and-demand schedules are well known. If a change happens that pushes things outside those boundaries then supply or demand schedules can have strange shapes until businesses have had a chance to compensate for the situation. So perhaps our different views on the liquidity trap aren’t so important. That said, over a longer period of time the substitutional relationship between direct investments, bonds and short-term debt relationships comes into play. Although short-run rates of interest may fall close to zero it would be difficult to argue that these other rates could. As I may have mentioned before, all the way through this crisis I’ve been able to buy (in the UK) 1 to 2 year fixed-rate bonds with interest rates of ~3% AER.
It’s not so much that I don’t think that a “fallacy of composition” can occur, it’s more that we can’t in any way clearly see if it is occurring. We need to disambiguate the complicating factors I mention above and I don’t see how that’s possible.
I don’t think the situation with money is so dire. It’s possible for private banks to tell when the demand for money is changing and change the supply in rough correspondence. But I don’t see how it’s possible to divine what the demand for output should be.
Rob Murphy has written about this too, in the context of Krugman’s blog posts…
I don’t entirely agree with him about this, but here it is anyway.
“You’re becoming worse than me for recycling your old material ;)”
I know, bad form. But not quite as bad as TCC, which posted 3 articles making exactly the same argument within days!
I think the key part of your argument is the following:” If the return on investments is viewed rationally then an investor should prefer the scenario where I pay down my debt, if it increases the net present value of my spending stream on products related to his investment. ”
Quite right – an investor would prefer the scenario where you pay down your debt. I would stress that the repayment of debt in my example is perfectly rational, and the right course of action, for individual companies. They have to repair their balance sheets to avoid investors taking fright. But the problem comes when all companies start repaying debt – that is the fallacy of composition.
I’m a bit undecided on whether money policy could counter this kind of problem at the moment. If no one wants to borrow money, it will just pile up in vaults. But then this is a bit of a simplification and there is also the influence on creditors to consider.
Comments are closed.