Suppose there were an excess demand for antique furniture. Antique furniture is not part of GDP. By Walras Law, if there were an excess demand for antique furniture, there must be an equal and offsetting excess supply of something else, like newly-produced goods for example. Could an excess demand for antique furniture cause a general glut of everything else, a drop in aggregate demand, unemployment, and a recession? Why not?
Why is an excess demand for money (which is also not part of GDP) so different from an excess demand for antique furniture? Most people own a lot more old stuff than they own money, so it can't just be a quantitative difference.
1. Money is the unit of account, and antique furniture isn't. If there is an excess demand for antique furniture, the price of antique furniture must rise to eliminate it. But if there is an excess demand for money, since the price of money is the reciprocal of the price of everything else, the price of everything else must fall to eliminate it. It is a lot easier for the price of antique furniture to rise than the price of everything to fall.
2. That wasn't the most important reason. This is. Money is the medium of exchange, and antique furniture isn't.
Suppose there is an excess demand for antique furniture, and the price of antique furniture cannot rise to eliminate that excess demand. (Just bear with me, and use your imagination to think up stupid legal price controls on antiques). Everybody plans to sell more newly-produced goods than they buy, and use the savings to buy more antique furniture. Walras Law says that the value of the excess demand for antique furniture equals the value of the excess supply of newly-produced goods. So according to Walras Law, price controls on antiques, and an excess demand for antiques, could cause a recession. Why not?
Walras Law is wrong. Or rather, Walras Law is only right for notional excess demands and supplies. Notional demands and supplies are the quantities people plan to buy and sell under the assumption they will be able to go to the market and actually buy and sell those quantities. But if there is an excess demand for antiques, they will go to the antique market and find there is a shortage of antiques, and they can't buy as many as they want. Not enough willing sellers, because by assumption there's an excess demand for antiques. Notional demands and supplies are irrelevant in disequilibrium, because they are plans made under an assumption that people will learn is false.
Frustrated in the antique market, people revise their plans in the market for newly-produced goods, taking that constraint on antique purchases into account. Unable to satisfy their excess demand for antiques, they will revise their plans to buy less newly-produced goods than they sell. Unless they plan to hold those savings in money, creating an excess demand for money, they must now plan to buy exactly the same value of newly-produced goods as they sell. The excess supply of newly-produced goods disappears.
An excess demand for antiques logically cannot create an excess supply of newly-produced goods (unless it creates an excess demand for money as a side-effect). It would violate the budget constraint if it did.
But an excess demand for the medium of exchange is different. Suppose everybody wants to add to his stock of money, rather than add to his stock of antiques. And suppose the price of money (the inverse of the price of everything else) cannot adjust. They plan to go to the market for newly-produced goods (and the market for antiques) and sell more than they buy in all those other markets. Every market is the market for the medium of exchange. They fail, of course, because it is impossible for everybody to do the same thing. They are unable to sell as much as they wanted, and so revise their plans of how much they will buy, taking that constraint into account. They decide to buy even less, which means (since everyone is doing the same), they find they can sell even less….
The Keynesian multiplier process continues until sales, output, and income drop enough that people decide to stop trying to add to their stocks of money. The recession gets worse until the excess demand for money disappears.
Problems in the market for antique furniture cannot cause a general excess supply of goods, unless they cause an excess demand for money as a side-effect.
Similarly, problems in the market for anything else, like mortgages, asset-backed commercial paper, credit default swaps, other fancy derivatives, or pet rocks, cannot cause a general excess supply of goods, unless they cause an excess demand for money as a side-effect.
This recession is a monetary problem. The cure must be to increase the supply or reduce the demand for money.
(Nothing here is really new, of course. Clower for example knew this stuff 40 years ago. But people forget, or are too young to remember, or never learned it in the first place.)
Worthwhile Canadian Initiative 
I thought it was good Nick. But…
Every thing you said refers to REAL quantities. In the liquidity trap your 1” doesn’t hold. We have investment bought < actual savings, THAT’S WHY PEOPLE END UP HOLDING EXTRA MONEY.
It’s not the money they want but the money is giving them a better return then 1) they require to hold it; and 2) real investments. It’s not an excess demand for money!!!
Thanks, Nick.
That helps.
“Savings demanded” is a phrase I need to get my mind around. Seems abstract in an unusual way. Also takes me back to the stock/flow thing with money and savings.
Your patience with elementary questions is always appreciated.
JKH: Yep, I find it hard to think about “savings demanded” as well. “Desired savings” is perhaps better. “Savings supplied” sounds much more natural, to my ear, because we can think of people supplying loanable funds. But then desired savings is not really the same as the supply of loanable funds, since desired savings also includes putting cash under the mattress. It includes investment as well: if I want to spend part of my income on newly-produced investment goods, that is both desired savings and desired investment, but we can think of this as me lending money to myself, so it appears both on the supply and the demand side of the loanable funds market, and therefore cancels.
The basic problem is that “desired savings” is defined negatively, as income from the sale of newly-produced goods minus demand for newly-produced consumption goods. That’s why it’s so hard to think of it in concrete terms.
Adam: I think 1 and 2 (or 1′ and 2′, or 1″ and 2″) can be defined in either real or nominal terms. Just divide or multiply by some common P.
But I am really worried by your: ” In the liquidity trap your 1” doesn’t hold. We have investment bought < actual savings, THAT’S WHY PEOPLE END UP HOLDING EXTRA MONEY.”
I thought our roles were: I’m the old guy, mumbling weird, heterodox heresies, from ancient economists; you are the smart young guy, trying to keep me on the straight and narrow modern orthodoxy.
Now I’ve suddenly got to switch roles, because you have taken my role! You are echoing a view from 1920s and 1930s British monetary theory: Robertson, Hawtrey, Hayek, etc., that the excess of savings over investment adds to money balances. You only hear that view nowadays from the Austrians, who hold that central banks increasing the stock of money creates some sort of excess of investment over savings — “forced savings” in the older terminology.
It’s not wrong what you said. But you are using “saving” in a non-orthodox way. Actually, the way you are thinking is closer to my way of thinking: you divide income into demand for consumption, supply of loanable funds, and demand for additional money.
So, you can see why I’m flummoxed!
A wonderfully clear explanation Adam, thanks! I was assuming that the rise in income was permanent. I must hang out here more often! Presumably, Krugman is making the assumption that a fiscal expansion can indeed boost income, which gets back to the question raised by Fama or Cochrane about why fiscal expansion boosts income when the government expenditure involved has to be funded somehow (which I guess is the same as JKH’s point here). I don’t expect you to answer that though; that issue was probably discussed here at the time!
One wonders about the wisdom of teaching ISLM “which leaves lots out” in a first macro course, while leaving the intertemporal microfoundations for a graduate course. No doubt the reason for that is that the intertemporal explanation is normally expressed in terms of difficult mathematics, but it surely does not have to be like that. But then the teaching of economics is another debate.
“We have investment bought < actual savings, THAT’S WHY PEOPLE END UP HOLDING EXTRA MONEY.”
This is way about my head, but I would just add this: Does ‘investment bought’ include deleveraging? In the current environment, I don’t see many people holding wads of cash, but their sure seems to be lots of people throwing every dollar they can get their hands on into holes they dug in the 2003-2007 time frame.
Nick,
As a first thought (it’s late here in the UK) on whether it matters who the central bank buys assets from when doing QE, I would say that it does not matter. Assume the central bank could buy, say, government bonds from either a bank or a pension fund. If it buys from the pension fund, the central bank will pay base money into the pension fund’s bank, which will in turn credit the pension fund’s account with them. If the pension fund subsequently wants to draw down their deposit but the bank wants to hang on to the base money, it can raise a similar amount of base money by selling in the market the bonds that it could have sold to the central bank. In other words, because the central bank pays for the bonds in base money, and because base money is used to settle payments, if desired, the banking system can reach the same position regardless of whether the bonds are purchased from banks or non-banks.
Nick, what I’m saying is a bit more basic than that. We both agree that QE is correct the difference is that you, more or less (I don’t want to put words in your mouth) believe that the money supply today is the problem and I think you need to lower real interest rates (in this case by creating expected inflation).
Now the normal way it’s supposed to work is if I’m a saver and you’re an investor I buy financial assets, for example a demand deposit at a bank. The banking system/financial markets channel the funds to you and you invest (either funding yourself through debt or equity). My point here is nothing more than the fact that even in normal times, from my point of view I save by buying financial assets.
Now, today we have a situation where the amount I want to save exceeds the amount you want to invest. I don’t know that though, so I do the usual thing and buy some sort of financial assets. The problem is that the market doesn’t have any place to channel the funds so they just sit as money. But, from my point of view, I don’t want the money as a medium of exchange. Quite the opposite, I’m trying consume less now in return for more later.
But this: “You only hear that view nowadays from the Austrians, who hold that central banks increasing the stock of money creates some sort of excess of investment over savings — “forced savings” in the older terminology.”
This is not at all what I’m saying. I was trying to make the point more stark that my savings level is determined by feeding the relative prices of today and future consumption into my first order conditions for intertemporal utility maximization (consumption Euler equation), you can’t change my behaviour without changing those relative prices (that is, real interest rates). Printing money doesn’t “satisfy my excess demand for medium of exchange” because what I really want is higher future consumption and only more real investment can give me that.
Same statements go for real investors trying to maintain MPK = r, you have to change the real rate to get them to invest more. It comes directly from the condition of maximizing discounted present value of all real profits.
Rebel, I agree. It doesn’t matter who they buy the assets from. My reasoning is simple though, I think it’s all about interest rates.
Adam,
I might be doing you an injustice, but when you say that “the funds….just sit as money” you seem to be missing the details of how base money is supplied and collateralised (and even if you do know, you may be misleading others). Apart from the facts that money can be used as the medium of exchange and the unit of account, there is not much difference between a banknote and a corporate bond, especially in a QE regime. If savers are so risk averse as to favour a zero duration, zero credit risk, statutorily liquid asset, and the central bank believes that this behaviour is undesirable, the central bank can simply supply more money by buying, say, the corporate bonds that the same savers might have bought in normal times. Given the nature of this extra demand for money, the extra money supply poses no inflationary threat. In this case, the central bank is just acting as a financial intermediary. In fact, in theory, the central bank could buy the corporate bond at the normal price so that corporate activity (eg investment) proceeds undisturbed, and the issue reduces to one of whether the savers are irrational to be so risk averse, in which case the central bank (and ultimately the taxpayer generally) ends up richer, or whether the savers are realistic, in which case the central bank (and taxpayers) end up poorer.
Economics textbooks are poor at covering the practical details of monetary policy implementation, especially recent developments, which is why I wrote the blog post referred to above (as I said, I used to be a central banker, working in market operations). If you are a bit hazy about the mechanics, you may find it helpful; if you are fully aware of how it works, I apologise.
Rebel,
You’re not really doing me an injustice since I know nothing of the details of how base money is supplied and collateralised. However, my point had nothing to with the details of how base money is supplied an collateralised.
My statement, “the funds… just sit as money” can be changed to “the funds just sit”, in whatever arbitrary form. The “as money” part was because I had in mind a very simple financial structure where the only financial asset savers buy is demand deposit accounts.
When you say “If savers are so risk averse as to favour a zero duration, zero credit risk, statutorily liquid asset…” you’re missing the details of how investment is determined. You said “the central bank could buy the corporate bond at the normal price so that corporate activity (eg investment) proceeds undisturbed”. That presumes corporate investment activity is not zero.
Suppose that the required return on real, risky investment is 3%, this comes from a risk-free real interest rate of -2% and a 5% risk premium. The central bank intervention you refer to can solve the problem of savers being too risk averse but investors are people too. They also are risk averse, they will only undertake the investment project if the expected return over their financing costs (everything real, not nominal) pays them the risk premium.
Now, suppose that past investment has driven the marginal product of capital, (real, risky capital) to 1%. No investment gets done and the central bank can do nothing about that because these things are all determined on the real side of the economy. MAKING CREDIT MORE AVAILABLE ONLY HELPS IF SOMEBODY, SOMEWHERE WANTS TO BORROW.
Real savings and real investment demands are determined on the real side of the economy by real interest rates and the real risk-premium. Monetary policy only affects these decisions by changing the real interest rate (well in the baseline model). In particular, generating inflation expectations is how you lower the real rate when the nominal rate is zero.
Now, there are many coherent stories in which monetary policy works by changing the real risk premium or relaxing constraints, getting credit to someone who wants to borrow and invest but somehow can’t get credit. For example, you might reasonably assume that corporates are risk neutral, then they may want to invest but can’t because savers want a higher risk premium then current investment opportunities offer.
However, that’s not the story Nick is telling. In the basic model investors and savers, in aggregate, are just people, they’re just us. Furthermore, in none of these alternative stories is there really an excess demand for money per se.
BTW, it should be clear that the investment demand being zero is just a stark example. The point is that the demand for funding may be too small to put all the savings to use.
One more follow up, just to pre-empt a potential objection. We usually do not assume that firms are risk-neutral nor do we assume (in the basic model) that they maximize profits. We assume they maximize firm value which is the risk-adjusted, discounted present value of all expected profits. This is, btw, also the correct things for firms to do from the point of view of maximizing social welfare.
Nick,
“The basic problem is that “desired savings” is defined negatively, as income from the sale of newly-produced goods minus demand for newly-produced consumption goods. That’s why it’s so hard to think of it in concrete terms.”
That for me is the most profound characteristic. I’m always sceptical of arguments that don’t treat savings explicitly as a residual or dependent function. That’s why I have a problem with “demanded”, or “desired” for that matter, as both of those words seem to imply some independent intention.
“But then desired savings is not really the same as the supply of loanable funds, since desired savings also includes putting cash under the mattress.”
This is very interesting. I’m not sure I agree. I think it probably a false comparison, as follows:
Saving equals investment ex post, in aggregate, and in a closed system.
The aggregate “texture” of saving is fundamentally different than investment, because individual economic units can “short” saving. They can’t short real investment. Saving uses finance as a channel through which to connect to investment, and finance allows short positions in saving; i.e. borrowing more than income in order to spend on consumption. There is no such thing as the shorting of real investment flow, however (leaving aside depreciation, which is an entirely separate measurement area).
Therefore, there is a network of long and short positions of financial intermediation that is either “buried” within, or separate from, depending on how you want to define micro level terms, the net aggregate positive saving position of the economy in question.
Nick, this overlaps with our earlier discussion of gross and net debt. Taken to the limit, you can view all of financial intermediation as consisting of this network of long and short positions. You can define all of these longs and shorts as saving and dissaving, or you can define none of them as such. It depends on your paradigm for the definition of savings at the micro level, where shorting is fundamental to financial intermediation.
E.g. I save by purchasing newly issued stock in a company that makes a real investment. In one measurement paradigm, my saving exists in the form of an equity financial claim. This is offset by the company’s investment. In the event that the company makes no real investment, my saving is offset by the company’s dissaving by having issued a financial claim. Perhaps it offsets that with an operating income loss, or perhaps with zero operating income and saving in the form of cash in the bank. Generalizing beyond this example, saving includes all types of financial intermediation in the sense that the record of saving and dissaving at the economic unit level can be added up from all such micro balance sheet positions resulting from such period flows. However, one must broaden the definition of dissaving in this sort of framework to include not only dissaving from income, but dissaving from the pre-existing stock of financial assets. It gets conceptually complicated. The linkage of saving to some ultimate investment is then passed around the system like a hot potato until the necessary ex post accounting equilibrium is evident.
Returning to the example, in an entirely different measurement paradigm, the saving is not my purchase of an equity financial claim, but the addition of net worth (net personal equity) to my household balance sheet. That balance sheet equity position in itself is not securitized or “financialized” or “intermediated” in any sense, when viewed on its own as representing my saving, because it an accounting record on the liability/equity side of my balance sheet that stands independently of any type of financial intermediation that I’ve chosen in order to “execute” my desired/actual saving on the asset side of my balance sheet. So in that sense, defining micro measurement terms in that way, saving includes no type of financial intermediation. This is the measurement limit of separating the substance of the real economy from the financial economy, because all financial intermediation contributes zero to saving, because all financial claims are offset by the dissaving defined through sale or issuance of those same claims.
The hoarding of cash under a mattress then becomes one of those micro economic unit levels of financial intermediation, which may or may not be saving, depending on your micro level measurement paradigm, chosen from the above alternatives.
So I think my point is that the notion of mattress hoarding not being part of loanable funds is a false comparison. Such mattress hoarding must be matched by sale of same from another unit (from either flow or stock) or in the case of newly issued currency, borrowing by the government (unless the central bank is involved in credit easing as well). But such examples of micro level unit intermediation offset by opposite flows from other micro unit intermediation (or disintermediation) occur throughout the economy and financial system. And then the definition of these activities as micro level saving and dissaving or just plain non-saving intermediation activity depends on your measurement definition paradigm. Hoarding of cash under the mattress is another one of these micro level transactions, of which there are numerous other examples that don’t contribute directly to macro level investment, under one definition, and where all such examples contribute zero under the other.
Lots of interesting comments this morning. Not sure my brain is up to it.
Rebel @5.34: Here is my post on Krugman/DeLong vs Fama/Cochrane: http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/01/loanable-funds-and-liquidity-preference-delong-vs-fama.html The key to my way of thinking is that fiscal policy works because the initial excess demand for loanable funds creates an excess supply of money, which creates an excess demand for goods.
Patrick @6.56: No, “investment” does not include de-leveraging.
Consumption + investment (ignore government and net exports) represents purchases of newly-produced goods ONLY. We divide them into the two categories according to whether they goods are durable (investment) or non-durable (consumption). (OK, we often fudge that in practice, and treat consumer durables as consumption expenditure). Again, it’s a very different use of the word “investment” than in ordinary language.
Rebel @7.07: I used to think it doesn’t matter who the central bank buys the asset from in QE. And I still think it doesn’t matter in normal times. But a few days ago I changed my mind, and am now not so sure. If banks have a shortage of capital, and can’t extend more loans, then demand deposits can’t expand if the central bank buys bonds from a bank, because the bank won’t increase loans to individuals, even if it has lots of extra reserves from selling the bond. So the money supply M1 does not expand. But if the central bank buys the bond from an individual, who then deposits the cash in a bank, M1 expands. (But I’m still not fully comfortable with this idea, in my head).
Will post this while mulling over other comments.
“But if the central bank buys the bond from an individual, who then deposits the cash in a bank, M1 expands.”
To repeat my view on this, that’s a critical distinction in the money effect, and is central to understanding the Fed’s QE impact in this environment, and also why they’re quite comfortable with paying interest on reserves.
JKH @8.40: Investment is purchases of newly-produced real goods. Think bulldozers. I see no reason in principle why you can’t sell bulldozers short. The only difficulty in practice is that each bulldozer (especially after it has some wear and tear) is unique. Unlike financial assets, most produced goods are unique, and so not fungible, so it’s practically difficult to short them. (The exceptions are commodities like oil and wheat, which do have futures markets, and you can short them, I think. If all bulldozers were identical, I see no reason why you couldn’t short them.
Now savings may take the form of purchases of financial assets, which are fungible (my TD share is identical to your TD share), and so can be shorted. But not all savings are purchases of financial assets. If I use part of my income to buy a new bulldozer (I am an owner-operator, working construction), that is savings (and investment).
At its simplest, at the individual level, if I save part of my flow of income (i.e. do not spend it on consumption goods) I can do 3 things with it: buy capital goods (investment); lend it to someone else (buy a financial asset which is the liability of someone else); hold more cash (which you could argue is the financial liability of the central bank, though I don’t see it that way, and in any case, money might be something like cowrie shells, which aren’t produced, and aren’t the liability of anyone).
To my mind, it’s that third component of saving that is crucial to understanding recessions (because money is the medium of exchange). Not sure if you are saying the same thing.
Adam @2.08: (Leaving the hardest till last!)
” We both agree that QE is correct the difference is that you, more or less (I don’t want to put words in your mouth) believe that the money supply today is the problem and I think you need to lower real interest rates (in this case by creating expected inflation).”
That’s roughly what I think. Except that I agree with you that changing expected inflation is important too. And I would add that changing expectations about future real income are important too. But we can’t change those expectations (if they are rational) unless (future) monetary policy did actually have some direct leverage on (future) prices and real income. Otherwise, monetary policy would be like sacrificing a goat: it works only because people think it works.
I think I now understand what you are saying about excess savings and the demand for money. You are saying (just to check I have understood you) that the ORIGINAL cause of the problem is an excess demand for future consumption (relative to current consumption), that is not matched by a willingness of firms to reduce output of current consumption, and increase purchases of investment goods and thereby increase output of future consumption goods. People didn’t originally want to hold more money. Their “wanting” to hold more money is a consequence of their failure to find investors willing to borrow their money. They don’t really want to hold more money; they want to lend it.
[This is really helpful for me in getting my head clear(er), BTW]
Now, I disagree with you on that being the original cause, but that’s an argument for another day. Let’s suppose it is the original cause.
My point, in this post, is that even if that is the original cause, it can only create a recession if that original cause in turn causes an excess demand for money.
An excess demand for future consumption relative to current consumption MAY cause an excess demand for money, and if and only if it does so, it will cause a recession.
Similarly, an excess demand for antiques relative to current consumption MAY cause an excess demand for money, and if and only if it does so, it will cause a recession.
In your example, the first best choice is to buy new financial assets from firms wanting to finance new investment. When they are unable to do their first choice, households may want to hold more money as a second best choice. But if their second best choice was to buy present consumption instead, there would not be a recession.
In my example, the first best choice is to buy antiques. When they are unable to do their first choice, households may want to hold more money as a second best choice. But if their second best choice was to buy present consumption instead, there would not be a recession.
Gonna stop there, and mull it over some more.
Nick,
Consider an example where I use my saving to buy a newly issued bond in a company that invests in a newly produced bulldozer.
My saving is reflected in the increase in the equity of my household balance sheet.
My new financial asset is a bond. The bond is the liability of the company. From a net saving perspective, the two offset. If viewed as my saving, my bond is offset by the company’s dissaving by its “short” bond position. (The company has effectively shorted bonds by issuing them.) This automatic offset of financial asset and liability is why a “purer” form of saving measurement is to identify the actual saving at the level of new equity in my household balance sheet, rather than according to the particular financial instrument I choose to intermediate my saving to real investment. That’s one of my points.
My second point is on the inapplicability of bulldozer shorting in this same context. You say you can short bulldozers. I suppose you can. This means borrowing a bulldozer, selling it, and then buying it later and returning it. I’m not sure that’s done a whole lot as the physical consequence of real investment. But more fundamentally, bulldozer shorting or any other type of real investment shorting is absolutely inapplicable in the context of GDP saving and investment:
Shorting is necessary in order for financial intermediation to occur, by definition. Shorting in this intermediation sense is the incurrence of any monetary liability, such as the issuance of household mortgages, credit card debt, business commercial paper, bonds, and government debt of all types. By extension and for consistency, it includes the issuance of equity claims, given that equity claims are in respect of a firm’s obligation to record residual income for the account of shareholder’s equity, after taking into account all other financial obligations. It includes retained earnings for the same reason. All of these things are on the liability/equity side of the balance sheet as financial claims issued, and therefore are short positions at a fundamental level, as opposed to long positions on the asset side. In my example above, the company’s issuance of a bond was a short position in saving, at the margin, or by assumption. The company shorted money or income that it didn’t have, at the margin, or by assumption. Conversely, my need to use financial intermediation in order to utilize my new equity saving position in a real investment somewhere means that a short position must be created somewhere as a result of using that route.
Conversely, shorting is not an inherent component of real investment at all. That’s the point. Not that it’s impossible to short a bulldozer, but that the act of producing a bulldozer investment as part of GDP does not itself involve shorting. In that sense, shorting a bulldozer is a discretionary balance sheet transaction occurring post the bulldozer GDP production event. Conversely, shorting in the case of issuing a financial liability is a necessary part of the GDP process in order to connect my balance sheet saving from income to the investment expenditure in the bulldozer. Bulldozer shorting, if you want to do it, occurs post that GDP event. It is a balance sheet transaction, but not an income statement transaction.
Nick,
I wasn’t making a point about hoarding cash and recessions. I’m not sure I’d disagree with you on that, and I wouldn’t disagree with most of your most recent response. My observation was on the technical relationship between hoarding cash and loanable funds. If the private sector increases its cash holdings as a result of saving, then the government must have issued new cash. But this is true of any government liability. So I’m not sure that cash hoarding deserves the special attention it seems to get in this regard. And the relationship between the private sector and the government sector insofar as saving is concerned is just an extension of a similar relationship that exists within the private sector. The aggregate saving of the private sector consists of the sum of its surplus saving units and its deficit saving units. The aggregate saving of the entire economy is the sum of the aggregate saving of the private sector (include the current account as appropriate) less the (typically) deficit saving of the government sector – in other words, the sum of its surplus saving units and its deficit saving units, the same as is the case for the private sector considered on its own.
Almost there but not quite. You said: “excess demand for future consumption (relative to current consumption), that is not matched by a willingness of firms to reduce output of current consumption”
Since the firm borrows the capital it doesn’t have to reduce current output. In aggregate we reduce current consuption output to fund the real investment but that’s not in the single firm’s problem.
Take the example I gave where the risk premium is 5%, the real rate -2% for a required return on real investment of 3%. Now, remember that we assume the firm maximizes it’s market value, this basically gives it the same risk prefrences as the market. If there if the firms real investment opportunities only yield a risk-adjusted expected return of 1% neither the firm nor I want that investment.
Thus, I end up holding money as the best investment I can find. Because the risk-free real rate is -2% it actually seems a good investment. But I also am perfectly willing to hold government bonds at zero return.
Here’s the money quote:
It’s not money I want, it’s more future consumption. Printing more money doesn’t help me get that (and doesn’t satisfy my demand unless I suffer money illusion). In fact, since there are no investment opportunities that I want it’s not physically possible to satisfy this demand. The only way to get me to spend the money is to lower real rates so much that future consumption has become so expensive (in terms of current consumption) that I no longer want to substitute more future for less current consumption.
(An analogy might help but don’t get hung up on it. It’s basically like equilibrium in an endowment economy. We can’t increase tomorrow’s aggregate endowment so equilibrium only occurs when real rates adjust so, in aggregate, we are just happy to consume the edowment, that is we no longer want to do any intertemporal substitution.)
Adam,
I think see what you mean – your concern is the creation of new assets by investment, rather than the transfer of existing assets (eg bank deposits) in trade generally. I dare say that that has become more of a problem after the inital shock of the financial crisis. In such a case, the central bank could buy debt direct from the government, which could do the investment itself, eg in bridges, railways etc. That is printing money, as I define it.
Actually I should be more clear:
At current prices I want more future consumption. Since there are no real investment opportunities that anybody wants to undertake this is not physically possible. However, at current relative prices (ie, current real rates) I’m willing to substitute less current for more future consumption.
Thus the problem,
1)at current relative prices (real rate) I don’t want any more current consumption.
2)at current required returns (real rate + risk premium) there are no investment opportunities anyone wants.
Result is I don’t demand anything. This has nothing to do with money, it might happen in a world without money (but prices are still sticky). The problem is that prices won’t adjust.
Suppying more money does nothing if relative prices don’t adjust.
JKH: “Shorting is necessary in order for financial intermediation to occur, by definition.” Wow! That’s a novel (to me) way of looking at borrowing and lending. But I see what you mean. (But I would take out the word “intermediation”, since what you say applies to direct lending from ultimate lenders to ultimate borrowers, as well as if they go through a financial intermediary, though you presumably weren’t trying to say anything to the contrary).
I would say what you are saying a bit differently. Yes it does come right back to earlier posts about net debt. Investment is the increase per year of the stock of real capital goods, and real capital goods (like bulldozers) are net wealth. Part of savings (but only part) is the rate of increase in financial assets. Financial assets are not net wealth. (One minor exception to that: outside non-interest bearing fiduciary money, along with other Ponzi schemes, is net wealth).
“If the private sector increases its cash holdings as a result of saving, then the government must have issued new cash.” Agreed. But that’s like saying the actual quantity of cash “bought” must equal the actual quantity of cash “sold”. That’s true by definition, in or out of equilibrium. I’m talking about an excess demand for cash, which means that the quantity of cash people want to add to their existing holdings exceeds the quantity the government actually issues.
If we have an excess demand for X, we are not in equilibrium in the market for X. That’s true for any good. But if X is cash (the medium of exchange) it’s different, because there is no one market for cash. The market for cash is every market in the whole economy. We “buy” more cash by not buying other goods.
Rebel, yes exactly. As Nick keeps saying, it’s all about demand for new stuff.
Agreed that the credit crunch was a problem of capital not getting from savers to investors because the intermediaries were messed up. I suppose you could certainly make the case that that was what made people first start valuing future consumption so much (they started to worry that without enough investment now consumption would be scarce tomorrow). So perhaps fixing the financial system is key.
However, this not fundamentally an excess demand for money and increasing the money supply, all by itself, does nothing.
Let me correct that very last sentence. It should read: “We TRY to “buy” more cash by buying less other goods”. Collectively we fail to buy more cash, but our attempts to buy more cash cause a recession.
Of course, we could also TRY to buy more cash by selling more other goods, as well. But we can ALWAYS succeed in buying less, but cannot succeed in selling more, if other people want to buy less. The short side of the market determines the actual quantity bought and sold.
Nick,
Yes, I was using the term “financial intermediation” in the broader sense of finance per se (i.e. the creation of financial claims), including direct borrower/lender connections as well as the use of financial institutions as institutional intermediaries – amounts to the same as what you said.
“the quantity of cash people want to add to their existing holdings exceeds the quantity the government actually issues”
An example of where you lose me when invoking theory. This is not possible in my world, which I like to think correlates with the real world. Cash is a portfolio choice. The bank and the government can’t deny you the amount of cash you want unless you don’t have the money on deposit with the bank to begin with, which is a contradiction of sorts.
Adam: agreed, we are talking about firms in aggregate.
I find it interesting to compare two variants of the endowment economy example. In variant 1, people can consume their own endowment. In variant 2, there is a taboo against consuming your own endowment (to motivate trade), and also a taboo against consuming the endowment of anyone who consumes your endowment (to motivate monetary exchange, rather than direct barter).
In variant 1, there is an excess demand for money, because money is the only durable good, and everyone want to sell future consumption to buy future consumption, but current consumption is not affected by this excess demand for money. Everybody consumes his endowment. A social planner could not improve the outcome.
In variant 2, consumption falls, as everybody stops buying other people’s endowment in an attempt to get more money, so they can spend it next year on future consumption. A social planner could improve the outcome, by forcing everyone to buy more current consumption.
The first economy has no recession, and that’s because “money” does not serve as a medium of intratemporal exchange. It’s really antique furniture. The second economy has a recession, because money really does serve as a medium of exchange in that economy.
That was the main point of my post.
Now, suppose that endowment model (variant 2) were the truth about the current recession. A helicopter increase in money (outside money), and permanent money, would increase net wealth, and increase current consumption demand, even if prices and expected future prices were assumed fixed. (Though it would need to be a very large increase).
Let’s consider a third variant, which is exactly like variant 2, but also has antique furniture as a second durable good. (It’s not produced, and you get pleasure from owning it). My hunch is that the larger the stock of antique furniture (or the more pleasure it gave), the less would be the likelihood of a recession, other things equal. Now the real world has lots of assets like antique furniture. Land, old houses, gold, etc. They all ought to rise in price if we got into a recession caused by excess savings, and would be expected to fall in price thereafter.
In other words, even though it is theoretically possible that the excess demand for money that caused a recession could originally be caused by an excess demand for future consumption, how empirically plausible is that theory?
The increased risk premium theory is more plausible as an original cause, I think. Risky interest rates went up, and safe ones went down.
JKH @1.35: This may resolve the issue:
At full employment there is an excess demand for money. So people buy less stuff, because each individual can always get more money just by buying less stuff. But in aggregate they can’t get more money this way. But their buying less stuff causes a recession, so income drops. As income drops, people want to hold less money. Income drops until the excess demand for money, at that new lower level of income, disappears. So at the new unemployment equilibrium there is no longer an excess demand for money.
Rereading the post, as well as your last comment, it sounds like the increased demand for money is a portfolio effect of the usual paradox of thrift in terms of saving. Everybody tries to save more income, or everybody tries to hold more money – same result, one in flow terms, the other in stock terms. If correct, this seems perfectly understandable to me.
(I think the specification of cash (i.e. government issued currency) as distinct from bank deposits really complicates the point unnecessarily).
I’m probably coming in late, but I thought I’d make a point about Cochrane’s article. If you read the whole thing it is obvious that he understands fiscal stimulus can affect inflation (and hence other nominal aggregates.) The portion cited by Krugman was very poorly worded. In context, it was clear Cochrane was starting with an assumption of fixed velocity, but he didn’t say that. In addition, I don’t like the sort “real” argument he used. If you are debating AD, you need to focus on what affects NGDP. So I have some problems with both Krugman and Cochrane in that dispute.
Adam, I pretty much agree with your response to me (surprisingly!)
Nick: “Now, suppose that endowment model (variant 2) were the truth about the current recession. A helicopter increase in money (outside money), and permanent money, would increase net wealth”
No, that’s money illusion. We all know that the aggregate endowment next period can’t increase, if a helicopter gives us all more money and I spend my share but you save yours then YOU will succeed in consuming more next period at my expense! You’ll get a greater share of tomorrow’s endowment at the expense of giving me a larger share of today’s endowment. That’s exactly the deal you would like BUT it’s not the deal I want. We both want more tomorrow in return for less today. Hence, we both save the helicopter money just to keep up with each other.
the permanence of the helicopter money works, but only by increasing inflation and thus lowering the real rate. Prices are sticky but not fixed forever.
JKH,
In your comment of May 6 at 10.18, you seem to be saying that you think it matters whether the central bank buys assets from banks or non-banks, because the latter boosts M1. I would be interested to read more about why you think boosting M1, as opposed to base money only, matters.
(I made the point at May 5 at 19.07 that, if the banks held suitable assets that they would have willingly sold to the central bank if invited, they can reach the same position by selling those asset into the space in the market opened up by the central bank’s purchase from non-banks)
RebelEconomist,
Not unusually, you raise a critically important question. Here is my quick and dirty response; may not be precisely as right as I would like.
In a normal non-QE environment, the banking system as a whole does not undertake risk lending on the basis of reserves held on deposit at the Fed. They are not reserve constrained. They are capital constrained. Normal non-QE levels of excess reserves are insignificant, and used at the margin to guide the fed funds rate, not to guide bank lending, other than very short term money market manoeuvring that is essentially a response to fed fund rate guidance. Such short term money market operations are not terribly capital constrained. Commercial paper etc. requires some capital, but the duration weighting is very short. The Fed is not stupid. They have designed the normal non-QE system to work this way.
The question then is what does the Fed expect from the way in which it has designed its QE system, at least so far? Part of the answer is evident in the fact that they are paying interest on reserves. This is nearly moot at the zero bound, but they’re doing it beginning now to establish a precedent in case a QE residual is still in place when they decide they want to start increasing the fed funds rate. Anyway, this is evidence of the fact that the Fed isn’t expecting banks to fundamentally change their lending behaviour because of QE. They still expect banks to be capital constrained for risk lending. And as far as short term money market manoeuvring is concerned, the payment of interest on reserves establishes the functional requirement of a lower bound on short term rates. It is a neutral feature of the system in that sense. Money market spreads against fed funds might be narrower under QE than under normal non-QE conditions, but this is no big deal.
The important point is that the Fed isn’t expecting banks to suddenly invoke the classic and classically erroneous “reserve multiplier” in respect of risky assets, just because of QE. That’s not the reaction they’re expecting from the banks because that would be a stupid reaction. Banks need to be capital constrained in the accumulation of risky assets, QE or non-QE.
So, the payment of interest on reserves is an unambiguous signal that the Fed expects that banks will conduct their risky lending operations according to capital constraints; not according to massive reserves provided by QE. As far as “risk-free” asset such as treasuries are concerned, banks need to be wary of interest rate risk, which in fact requires capital allocation as well. In short, the Fed does not want QE to provide a signal to the banks to suddenly engage in reckless asset expansion. This is also obviously why the Fed and Treasury have taken such extraordinary steps to assist the banking system with its transition back to capital adequacy. The Fed’s payment of interest on reserves is completely consistent with the overarching emphasis on capital adequacy.
So why do QE at all?
(Here I am defining QE in my own way, which I’ll address also when I post my comment on your blog later.)
QE in my definition IN THIS CASE is a by product of credit easing together with balance sheet expansion.
Therefore, an essential function of QE in this case is to provide funding for the Fed’s balance sheet expansion.
The other function relates to your question. QE when transmitted through the non-bank sector results in an M1 increase.
The two most important stimuli provided by CRE/QE, IN THIS CASE, are the assumption of collateralized credit risk by the Fed, and the expansion of M1 Q when the Fed’s activities are transmitted through the non-bank sector and back to the banks as M1.
The Q and V effect of the additional M1 is what provides the Fed’s intended monetary stimulus, not the expansion of the base.
This is all consistent with the Fed’s longer run structural use of bank reserves for its own purpose, which is to provide a leveraged mechanism of policy interest rate control.
The fact that QE has resulted in expanded banks reserves isn’t a signal for the Fed to suddenly use bank reserves in a different way. Again, this is why they pay interest on reserves.
My rough thesis here is diametrically opposed to that presented on such illustrious blogs as Money Illusions, and perhaps obliquely askance to even more illustrious blogs such as WCI.
JKH, great comment. Care to comment on how one comes to know quite that much on the inner workings of the banking system?
If not then perhaps you’ll answer this: what’s your view on the stimulative effects of the fed’s balance sheet expansion and assumption of credit risk? (That is, I’m asking your view on how it works, the transimission mechanism).
JKH: I think I need to do a post on this. But first, can you clarify this bit: “The two most important stimuli provided by CRE/QE, IN THIS CASE, are the assumption of collateralized credit risk by the Fed, and the expansion of M1 Q when the Fed’s activities are transmitted through the non-bank sector and back to the banks as M1.
The Q and V effect of the additional M1 is what provides the Fed’s intended monetary stimulus, not the expansion of the base.”
What did you mean by “M1 Q”? Why the “Q”? Typo? Quantity of M1?
What are “The Q and V effect”?
Nick,
M1 quantity
quantity and velocity
e.g. additional M1 quantity of X, times velocity of quantity X, gives incremental monetary impact of incremental X
although I know X gets commingled with pre-existing M1, but still …
I’m guessing quantity and velocity?
Rebel @ 7.44: My tentative answer to your question is that M1 is the medium of exchange (or close to it), because people use cheques and debit cards to pay for stuff using their demand deposits. If an expansion of the monetary base stays as “excess” reserves (can’t really talk about “excess” reserves in Canada), but does not increase either demand deposits or currency in public hands, it does not increase the medium of exchange.
Thanks JKH. Is what you are saying here roughly compatible with what I was saying at the end of this post http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/04/bad-banks-and-the-effectiveness-of-fiscal-and-monetary-policies-on-ad.html about QE circumventing the banks’ capital constraints?
Adam P @ 3.06 and 3.07:
We don’t disagree. It’s just my way of explaining things. Take a simpler example: I want to explain why an increase in income increases the price of apples. If I were doing it very slowly I would say: “Suppose the price of apples stays fixed, quantity supplied stays the same. But quantity demanded would increase if income rises and price stays the same. So there is excess demand for apples. That excess demand puts upward pressure on price. So price won’t stay the same; it will rise (unless of course we really do assume it’s fixed)”.
I’m doing the same thing here. This is what I should have said: “Suppose the present and future price level stay the same. The new money would increase real wealth. With greater wealth there will be increased demand for consumption. So there would be upward pressure on either output or prices. So prices would rise (either today or tomorrow), unless of course we assumed that one or both of them were fixed.”
If prices were fully flexible, the rise in prices would eliminate the increase in wealth at the old price level. But the rise in wealth is what caused the rise in prices which eliminated it.
(This is how Patinkin used to explain the wealth effect.)
Nick,
Yes, that’s the core point.
M1 – to be, or not to be. That is the question.
It’s come up a couple of times since in discussion, where I think you said you were still thinking it through.
Nick, I think JKH needs to do a post on this.
Adam P,
I’m a big fan of Bernanke and what he’s doing, as unpopular as that position might be.
With “credit easing”, the Fed has temporarily taken over part of the commercial banking system function. QE is largely an enabling facility for the required Fed balance sheet expansion in order to do this. It also has M1 easing benefits. The Fed has actually shied away from using the term “quantitative easing” because it is fundamentally ambiguous, implying many different things to different people. “Credit easing” is in fact a term that Bernanke has embraced.
The larger US strategy includes the Fed, Treasury and FDIC functions. It targets banking system assets (PPIP), liabilities (Fed), and equity (Treasury), as well as fiscal stimulus. The strategy is to try and defeat a disease by completely enveloping it from all angles. I think it’s the Powell doctrine (overwhelming force) applied to the banking system problem, where force is defined in the sense of surrounding the problem. I prefer it to nationalization, which I think of as a cop out.
I think all of this stuff will work over time. The key is that they tackle it incrementally and remain flexible, which is actually the point that a lot of people criticize.
BTW, Timothy Geithner wrote an important prelude to the stress test results (to be released at 5 p.m. today) in the NYT this morning:
“They applied exacting estimates of potential losses over two years, along with conservative estimates of potential earnings over the same period, and compared them with existing reserves and capital. The results were then evaluated against strict minimum capital standards, in terms of both overall capital and tangible common equity.”
http://www.nytimes.com/2009/05/07/opinion/07geithner.html?_r=1
This is quite consistent with what I wrote here:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/04/canadian-vs-us-bank-regulation.html
at April 26, 2009 at 06:40 PM
That was in a discussion with Nick relating to the definition of bank solvency. Treasury will do a dynamic projection of solvency over a two year period. Among other things, such a projection will include estimated operating earnings along with all of the potentially negative inputs relating to asset write downs and losses. Relative to my discussion with Nick, it means that the definition of solvency is anchored in the balance sheet, but it includes allowance for a bank to earn its way out of its problems over a reasonable time period. So the interpretation is the combination of a balance sheet approach and an income statement or cash flow approach.
This is absolutely critical in assessing the state of the US banking system. I’m guessing that the system is capable of generating upwards of $ 250 billion or more per year in operating earnings, pre-tax. That money, $ 500 billion or more, can be used, pre-tax, to offset against additional asset write downs and loan losses over that period of time. This is a fundamentally different view of bank solvency than the Armageddon interpretation favoured by people like Roubini, who mark losses to market and declare the entire system insolvent, and then scale the walls, demanding comprehensive nationalization. It is also linked inextricably to the idea of marked to market accounting and its effect on the interpretation of capital and solvency. MTM accelerates the present value of estimated losses into current capital assessment. Projecting operating earnings levels the playing field somewhat, in the sense that it is equivalent to present valuing the future benefit of the ongoing franchise banking spread along with the bad asset stuff.
Re my own analytical approach, it’s balance sheet centric. I’d love to see monetary economics or economics for that matter embrace more of a balance sheet approach. Then I might start to comprehend some of the theory. As it is, I’m still shaky when it comes to understanding supply and demand curves.
Thanks JKH, I see. I am interested to note that you mention the Q and V effect of M1 rather than something about bank lending. I wrote something similar in a reply to a comment on my blog this morning. It seems to me that piling up M1 is not necessarily a sign that QE is not working.
However, the point is moot if banks are not happy to have a pile of callable liabilities and can reduce them by selling assets. Do you have any view about whether it matters if the central bank buys its assets from the banks or non-banks?
By the way, I had a scan over some of the past posts at Money Illusions, and I think I agree with most of the arguments you made there.
Adam P,
“Nick, I think …”
Thanks, Adam, but Nick did highlight earlier the potential for a unique M1 effect, and this was also the point that RebelEconomist questioned.
I’ll be interested to see how Nick views this going forward; i.e. as something that is important to the overall interpretation of QE, or as merely one of two scenarios that on balance doesn’t really change the big picture for QE.
RebelEconomist,
“Do you have any view about whether it matters if the central bank buys its assets from the banks or non-banks?”
You could always say that what goes around comes around, and that maybe it doesn’t matter in the end. But my instinct is to say that the purchase of assets from non-banks is an immediate expansion of financial intermediation via the banking system, including an increase in M1, which is expansionary from a monetary perspective. The other way around, banks trade assets for excess reserves. In theory, that might cause banks to open up lending channels, not because of more excess reserves, but because capital is freed up. And such lending will increase M1. But it’s roundabout, so I prefer the non-bank purchase route.
“It seems to me that piling up M1 is not necessarily a sign that QE is not working. However, the point is moot if banks are not happy to have a pile of callable liabilities and can reduce them by selling assets.”
Maybe I’m just punch drunk on the subject, but I’m not sure I understand this. Can you elaborate a bit?
Adam P: that’s what I was thinking too. But I need to post on it as well.
JKH: I see the M1/capital constrained banks as one of many potential mechanisms through which QE might work. But in practice it might be the most important one.
I have delayed posting on it, because I wanted to get my head round it from as many angles as possible. I’m still not fully at ease with the idea. But the fact that JKH can see it working, as one who knows how actual real life systems work, (and he and I normally have a very different perspective on things) gives me more confidence. And I wanted to get my head a bit clearer on the M1/base/medium of exchange issue first (hence this post).
I will post on it. It would be great if JKH could post on it too. I will leave it up to you how to do that, JKH. Separate post, addition to my post (I can “hoist from the comments”), or whatever you like.
Too many things to think about. I should probably have been thinking about this last night, rather than the slightly silly post I made at 2.00 a.m. this morning.
And then there’s JKH’s comment about stress tests, which seems equally important.
Brain overload. Too much to think about.