« Excess supply and monopolistic competition, once again | Main | Out-sourcing your own job »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Some negative coverage is better than being ignored.

The actual *motion* itself that the Greens passed doesn't seem too crazy. (Actually, I confess I can't really see the real difference between what that motion says and the current practices of the Bank of Canada under the Bank of Canada Act.)

But the rest looks like a mishmash of old Social Credit and Narrow Banking (100% reserves) theories. It's rather hard to disentangle.

I highly recommend the comments to Stephen Gordon's foreign ownership post, including those by Duncan Cameron. There are reasons to be concerned about foreign ownership undermining economic democracy.

Herb Wiseman has been riding this hobbyhorse for a while now and it pops up in small-city newspaper letters to the editor from time to time.

It catches people through the following:

1) Debt is bad, and therefore should be minimized.
2) Banks are bad.
2) Debt to banks and other commercial entities is especially bad.
3) As we do need a debt, better to borrow it from the Bank of Canada and thus borrow it from ourselves, not that greedy lot on Bay Street.
4) Who said anything about inflation?

I've seen both good and bad behaviour from banks. Among small businesses banks have a dubious reputation for calling loans at the most inconvenient times and thus shutting down a viable business.

The best explanation I have heard about this was from a podcast on small business banking: "The purpose of the Chartered Banks is to provide senior, secured financing to Canadian businesses." The fact that businesses need other forms of financing that could and probably should be met through other institutions or structures is lost on most people, including politicians.

I have often thought that community or regionally-based "risk funds" which invested in startups, new ventures, extended more flexible financing terms and helped with equity would be a worthwhile initiative as a way to complement the Chartered Banks. A second tier of institution that focused on getting businesses ready for the big leagues of the Chartered Banks by engaging in financing that is just too risky for the Chartered Banks to comfortably service.

Seems like it would make more sense to raise taxes as necessary (and run a surplus) in order to offset the money supply increase. Once we work through the $600 billion or so in debt outstanding, they can work on a new plan.

Am I missing something or does 78% and 12% add up to 90%? Forget monetary policy, we should be concerned that the Green party's candidates can't do simple math.

The previous GPC resolution was to convert all numeracy to base 9.

With any other group, innumeracy would be a more plausible explanation. But with the greens...

Lets imagine that bank lending were significantly subsidized. Then the resulting reduced loan pricing would cause a distortion of capital allocation which will reduce overall efficiency. Measures that reduce lending may then improve efficiency. Since banks fund their balance sheet in large part for free (M1+M0) thanks to bank charters and deposit insurance, this is exactly the situation in which we find ourselves. The proposal is not at all the best way to deal with the problem since it doesn't deal with the core issues. But it's not harmful from the perspective of efficiency.

Bob, the traditional autocratic parties play "Simon Says". The Canadian Greens are different. It's "Mother, May I?"

The Greens' finance critic told me his plan is to raise investment taxes in order to eventually eliminate all payroll taxes. This seems like the strangest tax shift I can imagine. His argument was that payroll taxes are "job killing" but I don't understand how sharp rises in capital gains and corporate taxes are exactly the efficient strategy here. It can't be that they think investment taxes are more efficient than consumption because they plan on raising the GST as well. They also plan on expanding EI program spending quite a bit so it isn't because they think payroll-funded programs are too large either. Weird.

May also said that she wants to see an 80 cent Canadian dollar because it would help manufacturing. Someone should press her to explain what sorts of policies would get us there over what time horizon.

Their platform also contains a $5,000 cheque (with no clawback) to each welfare recipient as a down payment on a guaranteed annual income but doesn't cost this in their fiscal plan.

The other thing that I think is underreported is May's theory on growth economics. She has compared economic growth to cancer on a number of occasions.

I support a number of the Greens' touchstone policies - carbon tax, guaranteed annual income and proportional representation - but I don't want to touch some of the weirder ideas with a ten foot pole.

The other thing that I think is underreported is May's theory on growth economics. She has compared economic growth to cancer on a number of occasions.

That's standard in modern environmentalism, if you assume that economic growth = growth in use of natural resources.

It's an odd party, given that it draws a broad swath of the political spectrum and clings to its "grassroots" image and that leads to some...interesting proposals. I was there when this motion was debated and came to the conclusion that the motion itself was harmless, as it would be interpreted by the Finance Critic. The original motion, on the other hand - which did talk about 0% interest on GOC debt borrowed from the BOC - made me worry a bit that the party would continue to be condemned to fringe status. My take is that they're gradually approaching balance but could use a little more economic ballast to keep from swinging too far one way or another when a couple of these crazier ideas roll through.

They have one cash cow - the carbon tax - and need to decide whether they want to use it to cut income and business taxes (as currently claimed) or revamp the income security system to replace the welfare model with a basic income / GAI one (as currently dreamt). There's not enough money to do both. Still, the fact that they're on this path at all with possible equity and efficiency wins makes them top of my list - I was at the convention, after all.

Thank you for your comments everyone!

Alice Funke brought to my attention a Twitter comment from the author which linked to a YouTube video which clarifies her thinking on this issue. I re-post without comment:

http://www.youtube.com/watch?v=mUjBLLzYPGg

In fairness to the Greens, every new party has it's share of wingnuts until it becomes mainstream (i.e., has a chance of winning) to attract people who aren't loons. Within the last decade I can recall delegates (a decided minority, to be fair) at a federal NDP convention commemorating the glorious revolution of 1917.

Of course, it's a catch-22 situation, since it's hard to attract people who aren't loons when you're promoting policies that are loony.

I've read some of the monetary reform things, and quite a lot of it actually makes sense. For instance, you say that getting the Bank of Canada to 'create new money' to buy gov bonds interest free would just add more money to the money supply - but not necessarily so. When you or I buy a bond, or a pension fund, of course there is just a transfer of money and no new money is created - but when a commercial bank buys T-bills or other gov bonds - is it not true that they do exactly the same thing - just create the money out of thin air, as the popular expression goes? Again, increasing the money supply - but when commercial banks create the money, they also demand interest on this money they created out of thin air - and borrowing money to pay for interest has been a big part of the current national debt. (and I believe commercial banks currently hold ~20% of the national debt) But the point is - there would be no difference in how much money is created if the gov uses the B of Can instead of commercial banks. Yes or no?
(I have other points to follow, if anyone cares to talk and answer this first point ... )

Dave,

Well, it's been a long time since I thought about monetary economics (and I never much cared for it in the first place), but here goes. If the government sells bonds on the open market (i.e., borrows from the public or banks) those bonds have to be paid with real cash (so you have to get rid of some other asset to pay for it). So that's just a transfer of money from the banks to the government and no new money is created. If the government borrows money by selling bonds to the bank of Canada, on the other hand, the bank of Canada doesn't pay for that by getting rid of a real asset, it can just credit the government's account with the proceeds, in the process creating new money new money. (Indeed, in substance that's what quantitative easing is, the bank of Canada is buying up government bonds that have already been issued - it's economically no different than having the central bank lend the money directly).

While you're right that banks can also create money by making loans, that money supply doesn't decrease if the government borrows money be selling bonds from the bank of Canada. So, all else being equal, having the government borrow money from the central bank will increase the money supply.

" is it not true that they do exactly the same thing - just create the money out of thin air, as the popular expression goes? "

If that were true, then what is stopping, say, the Bank of Montreal from buying the entire TSX?

"... when a commercial bank buys T-bills or other gov bonds - is it not true that they do exactly the same thing - just create the money out of thin air,"

No. At least not according to what Nick would call the fiction they teach to undergrads. They'd have to deliver 'real' money to get the bonds. So-called money creation happens when commercial banks make loans with less than 100% reserves. I suppose they could lend someone money and that person could buy a gov't bond, though that would almost always be a silly thing for someone to do.

"When you or I buy a newly issued $100 bond from the government, there is no net increase in the money supply - you have $100 less, the government has $100 more."

There is no increase in the supply of hard money, but you don't have $100 less. You have a $100 bond. Before that, you had $100 deposit.

You can sell the bond whenever you want and spend the proceeds, particularly if it is a government bond, which is risk-free, and can be converted into a deposit at any time -- almost instantly. You are not deferring consumption for the duration of the bond, anymore than if you buy an antique as a store of value. The only thing you can say for certain about deferring consumption is that you need to defer it by however long it would take you to sell the antique, or the bond.

"However, when the Bank of Canada purchases the bond, there is a net increase in the money supply (since the BoC is essentially 'creating' the money)."

The BoC is creating reserves, but not necessarily deposits. If, before the intervention, A was short a deposit and had one too many bonds, he could sell the bond to B, who had one too few bonds and one too many deposits. They both swap so that they hold what they want.

After the intervention, A sells a bond to the BoC, obtaining the deposit he wanted, but B still has an extra deposit that he would like to convert into a bond. Is B forced to hold the deposit, or is he forced to rush out and spend? No, by withdrawing the deposit, he can force the financial sector to sell a bond to him as it responds on the margin to a deposit outflow.

So at the end of the day, the household sector has one fewer government bond (which the BoC has) and one more bank bond -- but the same number of deposits. The banking system has $100 more in reserves on the asset side, and one more bond liability. We are back to a no spending equilibrium, with the banking system stuck with some excess reserves. That's all that's happening here. Not a whole lot, but nothing to particularly advocate for.

The fallacy here is in not including a financial sector, that intermediates between households and high powered money. Therefore households are not forced to hot-potato high powered money via spending as they struggle to rid themselves of excess money balances. They can force those deposits back on the intermediation sector -- the financial sector, and rotate from holding deposit claims on banks to holding bond claims on banks (or holding assets that banks previously held).

As an example, we can verify that neither in Japan's QE, nor in the U.S. QE did household deposit levels change, even though the amount of reserves did change, and in the case of Japan, they changed substantially, so you would expect deposits to increase if households were really unable to eliminate them.

RSJ:

"You are not deferring consumption for the duration of the bond"

Ah, the old story of bonds vs. cash ;)

"you need to defer it by however long it would take you to sell the antique, or the bond
"

The same can be said about virtually any asset, like a house for instance. So, one can claim that the house == cash or any arbitrary asset == cash. The point is one cannot consume or spend as long as one holds any asset other than cash/deposits and not enough cash to satisfy the consumption.

But we've been through this.

Further, you say "he can force the financial sector to sell a bond to him".
What kind of forcing do you have in mind ? Does "force" mean "making it profitable for the bank to maybe sell some bank's asset" ?

Also, what kind of bond/bank asset the bank would sell to the household ?


"you would expect deposits to increase if households were really unable to eliminate them
"
Perhaps, households paid down their debts to banks after selling their treasuries, thus household deposits did not change. Are you suggesting that banks manufactured for example MBS/CDO stuff the households are eager to swap for new cash thus leaving their deposits unchanged?


Is there any data on asset flows during QE1 or the Japanese QE ? Who specifically got QE1's $1T in cash in exchange for agencies MBS ?

Without data of this kind, households' behavior during various QEs is nothing but scholastic disputation.

P.S. I am not much of a believer in equilibria -- Minsky's "no equilibrium" ideas seem arguably more realistic.

"I suppose they could lend someone money and that person could buy a gov't bond, though that would almost always be a silly thing for someone to do.
"

If the bank lends someone money, this someone deposits money in the same bank and this someone buys a T-bond, the very same bank has to come up with the amount of M0(reserve money, "base money") needed to settle the bond acquisition transaction.

VJK,

" So, one can claim that the house == cash or any arbitrary asset == cash."

No, that is not what I am claiming. I am claiming that when assets can be resold, then there is no longer any requirement that it be held to term, and that there becomes a disconnect between: 1) how long the person holds the security 2) when they consume and 3) how long the borrower's repayment time is. The disconnect between 1) and 2) is because they can rotate into something else without consuming. The disconnect between 2) and 3) is that they can sell to someone else and consume before the bond matures.

Is the above really controversial? Seriously, I thought I was echoing more or less conventional wisdom.

How long the buyer holds onto the bond is most likely a function of their own investment goals and time horizons. If the bond rises in price, they may sell it right away to capture the gain and rotate into something else. Or they may roll over 1 month bills for a long time. The point is that the motivations of the buyer are determined by their own expenditure plans and asset allocation needs, and these do not need to be the same as the duration of the instrument. The buyer may be a pension fund that will hold the bond to term. It may be a software program that will hold the bond until some criteria is reached. It may be a Japanese day trader. Who knows? All I'm saying is that the bond is now viewed as an asset, rather than as a promise to not consume. It becomes a store of wealth, with the puzzles inherent in deciding how long you should own it. Clearly 30 year bonds are likely to be held for more than a day -- I'm not saying that there is zero correlation between maturity and hold time. But given that most people do not spend all their wealth in one day anyways, I don't see this correlation as a meaningful constraint on spending flows.

Stocks, after all, have a long duration, but I don't think that the average hold period is very long, or that you can infer a lot about consumption plans by looking at the average hold period.

"Also, what kind of bond/bank asset the bank would sell to the household ?"

It would sell them whatever households want to buy, provided they offer an attractive enough yield.

That's how things work -- how does the Fed "force" the private sector to sell bonds to it? It merely offers an attractive price and sellers sell at that price. The same mechanism is at work when anyone wants to buy a security. Banks are more than happy to sell assets to households at a profit, particularly as households will tend to think in terms of expected utility which is biased in many ways, and whereas banks generally do not have these biases. AFAICT, my impression of the financial sector is that they are more than happy to sell assets to households. Btw, when I say "banks", I mean the whole financial sector, including MMFs, for example, who may simply be forced to sell assets when there are withdrawals.

"Perhaps, households paid down their debts to banks after selling their treasuries, thus household deposits did not change. "

Yes, there are many ways that households can rid themselves of deposits. Paying down debt would be equivalent to covering a short position, whereas I was talking about taking a long position, but the two would have the same effect on deposits.

"Are you suggesting that banks manufactured for example MBS/CDO stuff the households are eager to swap for new cash thus leaving their deposits unchanged?"

No, I don't think it works like that. I think all these decisions are made by price, not *who* you sell to. The bank is not thinking "Mr. Smith is about to withdraw since he wants a bond, so let's send him a prospectus". Banks see a deposit outflow and so they sell some paper, but that is short term. Longer term they are not going to be rolling over paper against long duration assets. At some point, securitization becomes attractive due to an increased demand for bonds. I do believe that the trade deficit caused a massive safe asset drain from the U.S. and that there were many people who were willing to pay a premium for safe assets, so that the financial sector manufactured them in order to meet this demand. I think that was Frank Portnoy's claim as well.

re: equilibrium, it's not a question of "beleiving" it. I am not talking about an equilibrium across *time*, but across actors. That equilribium can be dynamic -- e.g. changing in time -- or it can be static. It just means that people who want to change their positions change to their new positions.

Here is a fun example, from //www.hussmanfunds.com/wmc/wmc060130.htm in which Hussman is mocking the idea that there is "money on the sidelines".

"The example I used to give to my former students is this. Suppose Mickey wants to buy stocks, and sells some money market funds out of his portfolio to pay for them. Well, in order to get the cash to give to Mickey, the money fund has to sell some of its commercial paper holdings to Nicky, who buys the commercial paper with her cash, which then goes to Mickey, who uses the cash to buy stocks from Ricky. In the end, the cash that Nicky used to hold “on the sidelines” is now held by Ricky, while the commercial paper held by Mickey (via the money market fund) is now held by Nicky, and the stocks held by Ricky are now held by Mickey. Ownership claims have changed, but there is exactly the same amount of cash, money market securities, and stock shares in existence after these transactions as before them. Yes, prices may have changed depending on who was most eager to do those transactions, but there was no net flow of money into or out of the stock market."

That is an example of transition from one equilibrium to another. This has nothing to do with Minsky, or dynamics, but with market clearing.

Btw, this "Mickey/Nicky/Ricky" example is what I tried to argue with Nick about elsewhere on this blog. It was my attempt to say that interest rates were not set by demand for net borrowers intersecting with the supply of net savers, because there was this big pool of portfolio shifters who were neither savers nor investors, but were merely doing balance sheet operations, not income operations (such as the one described above). And they were the ones who ended up controlling the price due to their greater numbers.

Bob Smith | November 18, 2010 at 09:15 said: "... If the government sells bonds on the open market (i.e., borrows from the public or banks) those bonds have to be paid with real cash (so you have to get rid of some other asset to pay for it). So that's just a transfer of money from the banks to the government and no new money is created. .."

I'm not at all sure about this 'no new money being created' if a bank buys gov bonds. It's my understanding, from many sources after doing quite a lot of reading on this, that commercial banks create most of our money as debt, by making loans - as the loans are paid back the money is 'uncreated', making a continual supply of new loans necessary to maintain the money supply, which has some serious problems resulting we might get into later. The thing is - banks do not, as far as I can make out, as you say, 'transfer' (already existing) money in one form or another from themselves to the 'loanee' when they 'make a loan', as you say, they simply punch some numbers on the computer, and essentially new money is created on the spot and deposited in the loanee's account.
And I see no reason this would not Include government bonds, which is just a 'special' kind of loan. (and to respond to a different poster briefly, the reason banks do not buy up the TSX or other things by creating X trillion dollars at any given time is that they do have certain 'capital' requirements they must follow limiting their overall loan amounts, requirements established by the BIS in Switzerland, the 'boss' of all worldwide, or at least western, banks, which requirements were shattered by the major American banks in the recent world meltdown as they had vastly overvalued the book value of these assets and created a huge ponzi-style pyramid which was finally exposed and collapsed - but that's a different story. But here in Canada the major banks are, obviously, well within whatever these capital limits are in creating a trillion+ dollars for the Cdn money supply. But the loan limits are not based on 'reserves', which seems to be a misconception based on older rules - since 1994 in Canada, it seems, banks have had to have no reserves at all ( http://gilliganscorner.wordpress.com/2008/04/06/canadas-private-banks-have-no-reserve-requirements/ )

Some things that tend to support this idea that banks create new money when they buy government bonds:

* there is about $55 billion of 'hard cash' ('legal tender' bank notes) available in Canada (http://www.bankofcanada.ca/en/about/pdf/boc_balancesheet0810.pdf ) - but there is somewhere in the neighborhood of $1.3-1.5 Trillion in Cdn bank accounts (http://www.statcan.gc.ca/pub/13-022-x/2010002/t/tab01-eng.htm )
* and against those deposits, we have ~ $1 trillion in government debt in Canada (including provincial and municipal), and ~$2 trillion in consumer debt, and maybe somewhere close to another trillion in business debt. (figures can always be quibbled with, but I am sure these are the ballpark at least, which is all I really want to help me understand the process here; you can verify as you like through your own sources or google,) - certainly some of these loans will be from existing funds such as the large pension or mutual funds, but equally certainly a lot of them are bank debts ...
* the Bank of Canada itself says that '..the bank notes issued by the Bank represent only a small portion of all the money circulating in the economy at any one time. The bulk of the money supply consists of deposits that the public holds at financial institutions..' (http://www.bankofcanada.ca/en/backgrounders/bg-m2.html )
* it would be of some importance too, I think, to realise that of that $55 billion in 'cash', you can only think that most of it is not in a bank somewhere (thus creating an even greater discrepancy between what is ostensibly available as 'cash withdrawal' money from that $1.5 trillion in deposits based on a max of $55 billion in Cdn bank notes, and what is really available) - there's 30 million Cdns out there, most of whom have cash in their pockets, not to mention tens of thousands of businesses which have cash on hand, and god knows how many overseas money changers with Cdn dollars available - so whatever is actually available in bank vaults is some small fraction of that 55 billion.
* minor point, but it gets the point home - nobody, ever, gets a friendly phone call from the bank saying the deposit they thought they had is no longer available, as their money has been lent out - because nobody's money ever gets lent out, new money is simply created for the new loan. This is very easy to do these days, as by far the great majority of all transactions are done electronically these days, with no cash ever changing hands.

But the point is - if there's only some 55 billion of 'real cash' in the economy as printed by the government - but the money supply is somewhere between 1 and 3 or even more trillion - where is all the rest of the money coming from, if it's not being created as debt by banks?

(actually, to finish the response to the first post - so, if indeed the banks create new money when loaning to government through buying bonds - there would be no net increase in the money supply if the gov used interest-free loans from the Bank of Canada rather than interest-bearing debt from commercial banks - but the gov accounts would be better off because there would be no effective interest to account for in the future ... - interest which has contributed a significant amount to the current national debt - meaning the Green Party proposal may not be as 'nutty' as initially presented here ....)

RSJ:

"The bank is not thinking "Mr. Smith is about to withdraw since he wants a bond, so let's send him a prospectus".
"

It actually does ;), only after analyzing Mr.Smith portfolio positions before a possible withdrawal. There are whole departments "mining" Mr. Smith's financial data.


"many people who were willing to pay a premium for safe assets, so that the financial sector manufactured them in order to meet this demand. I think that was Frank Portnoy's claim as well.
"

Right. I think it is pretty obvious that if no safe assets are available in nature, the financial sector cannot manufacture them at will.

They did manufacture something recently, but calling that something safe assets would be strange after so much has become widely known to anyone mildly curious about those triple A monstrosities.

So, it is not entirely clear what safe substitutes the financial sector, by itself, can offer in exchange for treasuries, for example. There is no doubt that they can manufacture anything and package it as "safe" and be eager, not forced, to sell to anyone willing to buy.

So, for me, it remains a mystery what "long position" the household might want to take in preference to T-paper, other than speculative trading, of course.

"I'm not at all sure about this 'no new money being created' if a bank buys gov bonds."

In order to buy government bonds, the commercial bank has to have the required amount on its reserve/settlement account, or borrow from another commercial bank, or from the central bank. It cannot, legally and operationally, increase its reserve account by the needed amount.

The banks settle their obligations amongst themselves, as well as with the treasury, through that account (simplifying a bit), hence they need "base money"/M0 to do that.

RSJ:

"Yes, prices may have changed depending on who was most eager to do those transactions, but there was no net flow of money into or out of the stock market."
Well, that's a trivial point, regarding the stock market inflows -- no inflow happens without issuing new shares.

However, he is too easily glosses over price changes. Trading is a zero-sum game, however, its effect on prices and therefore capital reallocation and the consequences of such reallocation are obvious and well-known, ranging from ease of raising additional capital, "wealth" effect, to a bubbly market.

VJK,

The price changes are not glossed over -- the whole argument here is that interest rates are the determining factor, not the quantity of assets.

The central bank can certainly set the marginal cost of reserves, as the demand for reserves is inelastic. But it does this by changing the quantity of reserves in such an insignificant way that it cannot even be detected in the aggregated data. The Volcker rate hikes, for example, did not cut the monetary base in half -- changes to the base were so insignificant as to not be measurable. If I were to supply you with a time series of what the CB liabilities were, you would not be able to detect when an interest rate hike or cut occurred. Therefore you cannot argue that by increasing the CB's balance sheet, you are having an economic effect. You can only argue that the effect is by changing the short term interest rate.

This is basically a type of Ricardian Equivalence argument. When the government deficit spends, it must sell bonds. By definition, the private sector must save bonds (otherwise the deficit spending would not occur). But this quantity adjustment tells you nothing about the economic effects of the deficit spending.

The economically meaningful question is to what degree interest rates are changed when the government sells more bonds. There is zero quantity effect independent of interest rates.

By the same token, when the central bank sells bonds, the private sector, by assumption, buys those bonds. Again, there is zero quantity effect, but the effect is only on the possible change in yield. Similarly, when the central bank buys bonds, the private sector must sell bonds. Here, too, there is no "quantity" effect, but only a possible change in yields.

For some reason, the monetary economists think that when Treasury buys or sells bonds, that this has zero economic effect, as the private sector adjusts to reverse the change, but when the central bank buys or sells bonds, that this has profound economic effects, and the private sector does not reverse the change. How can this be?

I think they believe that Ricardian equivalence only applies to Treasury, and not to the CB, because they have no financial sector in their models.

If they would have a financial sector, then when the CB buys a bond, creating an extra deposit (backed 100% by cash), then the household sector is not forced to hot potato the deposit by buying more goods, driving up prices. Instead, the household sector can sell that deposit back to the financial sector, effectively buying a bond from the financial sector with the deposit, just as the CB buys a bond with cash.

This does not require any additional investment to occur. It does not require any additional borrowing or lending to occur. It does not require any additional spending to occur. And therefore it does not force inflation to occur.

All it requires is that households prefer to hold bond claims on banks instead of deposit claims. If the banking sector has, say, $1000 of liabilities matching $1000 of assets, then household preferences will determine how many of those liabilities are deposits, how many are bonds, and how much is equity. Should the CB give households an "extra" deposit by buying a bond from them, they can push that extra deposit back on the financial sector, and buy a bond from it. So the net effect is that the financial sector has one fewer bond (or is short one bond), and has more cash. I.e. excess reserves. But as long as the households don't have any extra cash or extra deposits, then why would they spend more?

The total balance sheet of the banking system is not forced to increase or decrease by this shift. The total assets held by households do not increase or decrease by this shift, either. But for some reason, this type of equivalence is completely ignored, whereas fiscal policy is supposed to be ineffectual because of precautionary savings, undoing the same transactions if Treasury happens to be the one doing the buying and selling.

Why the asymmetry that biases away from fiscal policy and towards monetary policy? Because they don't have a financial sector in their models, and so they believe believe that households hold the level of cash and deposits that the CB tells them to hold, instead of holding the level of cash and deposits that they want to hold, with the financial sector absorbing the difference.

As an aside I dug up more data on Japan's QE. Note that there was a rush out of commercial bank savings accounts and into government deposits, particularly postal deposits, which were viewed as being more safe. But there wasn't an overall increase in either checkable deposits + currency or in total savings deposit vehicles. I made a second graph, using a log-scale and also plotting against the CPI. This time I added the gov. and commercial banking time and savings deposits together, combined currency and checkable deposits, and plotted against CPI.

"So, for me, it remains a mystery what "long position" the household might want to take in preference to T-paper, other than speculative trading, of course."

Right. The real constraint is when households want more riskless assets than are available.

Obviously if households want more deposits, and fewer assets, then it is easy, the financial sector is more than happy to supply them with deposits, using the proceeds to buy higher yielding assets.

It's only a problem if you don't have a financial sector. Then you will believe that the "M/P" in your utility function is equal to the amount of liabilities issued by the central bank as a matter of policy, rather than the amount of liabilities issued by private banks as a matter of responding to household preferences. Only if you don't have private sector deposit providers in your model can you argue that a shortage of deposits is the source of excess demand.

But if households want more riskless debt, then the private sector cannot create it, only government can, and the quantity is determined politically, not due to household desires. That's something that the financial sector can't solve, except via fraud. Then you need other adjustment mechanism.

In which case removing riskless assets via QE is the wrong solution.

Nevertheless, just because QE removes assets that households want, this does not force households to hold more deposits, and it does not force firms to invest more. Households can still dispose of unwanted deposits by buying financial sector bonds.

Nothing forces households to hold a certain level of deposit claims on banks. They can hold other claims on banks if they don't want deposits. No one needs to rush out and spend the money on goods in order to "rid themselves of excess money balances", as there are no excess money balances on the part of households. The banks are the ones with excess reserves, and this is where the excess money balances show up.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad