Temporary vs permanent money multipliers

"Otherwise what I was mostly trying to suggest was that the banks anticipate the fact that the central bank won't let them double the supply of money and factor this into their loan and deposit pricing. The idea is that the current amount of deposits is not so much based on the curren[t] supply of base but the supply expected in the future." (That was from commenter HJC, with my emphasis added and one assumed typo fixed).

Now that is what I call a real and important critique of the money multiplier, as exposited in the textbooks. Because the textbooks are implicitly assuming a permanent increase in the money base, but none of them AFAIK make that assumption really explicit and talk about the difference between the current monetary base and the expected future monetary base. And it is a critique that has important real world implications, like for the US right now. And it is us Market Monetarists who should be making that critique.

Compare two different cases:

1. The Bank of Canada announces it will permanently double the monetary base relative to what it would otherwise have done. It recognises that doing so will permanently (approximately) double the price level relative to the price level path that it would otherwise have chosen under 2% inflation targeting, and it wants this permanent doubling of the price level to happen.

2. The Bank of Canada announces that a computer glitch will cause the monetary base to double, but only for one month. Because the techies are absolutely certain that they can fix it, but are currently all on holiday. The Bank of Canada assures everyone that normal programming of 2% inflation targeting will resume shortly, and that it will take steps to ensure that this computer glitch will have zero permanent consequences for the price level, if necessary by tightening monetary policy in future to restore the previously planned path for the price level.

In the first case I would expect an (approximate) doubling of currency in public hands and doubling of deposits. The money supply would double, as would the price level and all nominal variables like NGDP. Nominal interest rates would rise temporarily, then revert to normal.

In the second case I would expect approximately nothing to happen. The banks would roll their eyes and sit on (approximately) all the extra base as reserves for one month. (Since Canadian banks normally hold very small amounts of reserves, because none are legally required, and the base is currency+reserves, the stock of reserves would much more than double.) The overnight rate would fall 0.25% to equal the deposit rate (the rate of interest the Bank of Canada pays on reserves), and one month interest rates on liquid assets would fall a little too, but approximately nothing would happen to loans and deposits and the stock of currency in public hands.

I think that the current US case is much closer to case 2 than to case 1. Yes, the US banks have been sitting on the extra reserves for much longer than one month, but it is the conditionality that matters more than the duration. As someone (sorry I forget who) once said: the Fed has put out a large punchbowl of free booze, but no individual bank wants to drink much unless the other banks drink too, and the Fed says it will take away the punchbowl as soon as they start drinking. The full punchbowl just sits there. The Fed would need to announce that it wanted the price level (or NGDP) path to be permanently higher to give them permission to start drinking. And if it did that, a much smaller punchbowl would work much better than the current uselessly large one.

Other critiques of the money multiplier totally miss the point. Like "loans create deposits!" or "base is endogenous!" or "banks don't lend reserves!" or "other things affect the money supply too!". This critique matters because the textbook exposition is designed to show (among other things) how the central bank's control over its own balance sheet (which is all it really controls) allows it to control the money supply, in the same sense that I control my car. (Yes, the position of the steering wheel is endogenous, given the bends in the road I have chosen to take.) The current state of the balance sheet matters less than the expected future balance sheet, and the expectations about the conditions under which the central bank will change that balance sheet.

Comments

[Tom Brown]

Great Nick! Nice post... too bad you didn't get my horse analogy in there rather than the car one... but other than that I like it. Lol :D

[Tom Brown]

BTW, this gets to a question I had for Mark Sadowski yesterday: I bungled the question a bit by bringing up Canada vs US, but the basic idea was "as an empiricist, how can he factor in intention?"
http://www.themoneyillusion.com/?p=26518&cpage=1#comment-327750

[Ralph Musgrave]

In Canada and elsewhere there are no rules as to what reserves commercial banks have to hold, thus I suggest that in that scenario what reserves banks currently have or think they will have in a year or two is utterly irrelevant.

It’s often claimed that commercial banks need a minimum stock of reserves to enable them to settle up between themselves, but even that is not true. That is, if there was no central bank at all, commercial banks could use other assets to settle up: shares, property, bonds, etc.

[Squeeky Wheel]

I admit I still don't understand how changing the base is supposed to cause anything when required reserves is 0%. So neglecting that ignorance...

1) What if capital rules are more binding than reserve rules?

For example, CB says that government bonds + reserves must be >=X% of liabilities L. Assume that required reserves is 0, and bank B has a minuscule amount of reserves, plus enough government bonds to just barely cover the capital rule. CB engages in QE and buys all the bonds, swapping them for reserves. CB says this move is permanent. Now bank B still meets the capital rule, but only barely. Bank B is unable to make further loans. There is no change in money in circulation and no change in velocity. Why would any seller raise his prices?

2) What if there are no (additional) credible borrowers?

CB doubles the monetary base. All commercial banks look at the two entities in this world: It has already loaned Nick the maximum amount Nick is capable of repaying and SqueekyWheel is a dead-beat who never repays loans. No new loans are created. Why would SqueekyWheel raise the price of wheels?

3) What if velocity drops?

One of our seven consumers in the consumption-sale chain moves his spending from Monday to Sunday. Now all other consumers in the consumer chain need to hold much more money. The CB expands the monetary base to accommodate the desire for more money. However velocity has decreased proportionately. Hence, price level remains unchanged.

[DismalEconomist]

What timeframe would you expect prices to double? Instantaneously, 1 year, 50 years? If the Bank of Canada said we are going to double monetary base on December 31st of this year, do all suppliers/retailers instantly increase their prices today in anticipation? Interest rates would probably fall immediately, but it would take a considerable amount of time for this change to trickle through the economy as a whole and influence demand enough to double consumer prices. Does the time horizon of that price change matter, I think so but can't say for certain how much?

Counterintuitively, depending on the state of household savings, isn't it also possible that a drop in interest rates means that savings needs to now rise in order to meet a future liability, or that interest income falls and the combined effect of both of these reduces demand, offsetting at least part of the extra demand that would come from lower rates, thus causing the impact on prices to be less than a 100% increase.

[Frank Restly]

Squeeky Wheel,

See:

http://marketmonetarist.com/2012/03/18/how-unstable-is-velocity/

It doesn't explain exactly how a central bank hits a nominal GDP target, but at least it recognizes the potential trade offs between the supply of money and the velocity of money -

"If the NGDP target is 100% credible the correlation between growth in M and growth in V to be exactly negative 1."

[Min]

"1. The Bank of Canada announces it will permanently double the monetary base relative to what it would otherwise have done. It recognises that doing so will permanently (approximately) double the price level relative to the price level path that it would otherwise have chosen under 2% inflation targeting, and it wants this permanent doubling of the price level to happen. . . .

"In the first case I would expect an (approximate) doubling of currency in public hands and doubling of deposits. The money supply would double, as would the price level and all nominal variables like NGDP. Nominal interest rates would rise temporarily, then revert to normal."

In the first case I would expect a crisis of confidence in the central bank.

[Nick Rowe]

Squeeky:

0. Desired reserves, not required reserves are what matter. Required reserves matter only insofar as they affect desired reserves. Plus, desired currency matters too. Create either excess reserves over desired, or excess currency over desired, and individual banks and individual people will try to get rid of it, and will create deposits.

1. Capital is endogenous. Banks can and will raise more capital if it is profitable to do so. (And to the extent that banks have real capital, as opposed to nominal, it will rise automatically).

2. First, solve for the equilibrium, where P and NGDP are doubled, so nominal money demand is doubled, and nominal loan demand doubles. Then think about the process of getting there. Starting in equilibrium, one more loan of $1 is all we need to get the ball rolling towards the new equilibrium. Or one bank buying something for $1.

3. It won't. Velocity is a real variable. If it were going to drop it would have dropped anyway. If anything, velocity will temporarily rise due to the temporarily higher expected inflation.

Ralph: banks use central bank money for the same reason we use bank money and central bank money ourselves. We don't settle our accounts with the supermarket in land and houses.

Tom: yep. Car don't have expectations, which means the car analogy doesn't work as well as horses. But a better analogy still would be a herd of cows, who all want to stick together. We used to send one guy out in front, holding a sack, so the cows would all follow, knowing where we wanted them to go. Or get the "boss cow" to go where we wanted her to go, and let the others follow. It was always much easier to get them into the milking parlour than somewhere new and unfamiliar.

[Nick Rowe]

Dismal: Good question. Sometimes, like in the case of a currency reform where old dollars are replaced with new dollars, and everyone understands and can solve for the new equilibrium, it happens instantly. Dunno. I don't understand the Phillips Curve very well.

"Counterintuitively, depending on the state of household savings, isn't it also possible that a drop in interest rates means that savings needs to now rise in order to meet a future liability, or that interest income falls and the combined effect of both of these reduces demand, offsetting at least part of the extra demand that would come from lower rates, thus causing the impact on prices to be less than a 100% increase."

No. Fallacy of composition. For every lender there is a borrower. But the increase in the price level will change the distribution of wealth, which might have real consequences, which is why I kept sticking in "approximately".

[The Market Fiscalist]

If everyone woke up on Monday morning and felt really optimistic about the future and started buying up stuff, and then on Tuesday businesses went to their banks and started asking for loans to meet all this new demand then those excess reserves would start getting eaten up really quickly even if the CB did and said nothing. On Wednesday the CB would have to start tightening and telling people that any impression they may have given that the increase in the base was permanent was based on a misunderstanding.

If everyone woke up on Monday morning and felt the same as the previous Monday but the CB announced that the increase in the base was permanent (or at least would be kept at whatever level was needed to hit an NGDP target) then its possible (but far from certain) that on Tuesday people would wake up feeling optimistic and start spending just like in the spontaneous optimism case.

In other words: Its largely expectations about the future that drive both demand for loans and willingness of banks to issue loans. If expectations are low then the banking system may well have excess reserves. While expectations that the CB will reduce the base as soon as banks start to lend against it would definitely dampen people's overall economic expectation (and expectations that the base will be permanently increased increase them), its not clear to me that it this rather than spontaneous pessimism unrelated to CB policy that is the root cause of the current recession.

[Tom Brown]

Squeeky Wheel,

1) This is admittedly a very unrealistic example, but I put it together based on John Carney's write up at CNBC, and it demonstrates how a bank can create it's own capital by charging fees to new lenders:
http://tinyurl.com/kjwfzz2

So, I'm not necessarily arguing against your point, just that it's theoretically possible for a bank to simultaneously expand capital and loans. Also, I'm using a definition of Capital there as in (Tier 1 + Tier 2)/(sum of risk weighted assets). Does that correspond to your definition? I guess so, because both reserves and bonds have a zero risk weighting. BUT, I looked briefly at your Seeking Alpha profile, and I'm probably not telling you anything you didn't already know (me being a complete amateur)... but if you *do* happen to glance at my link, I have another wherein I try to explore the difference between capital and equity. John Carney actually did look at my first link there and blessed it, so I at least feel I've adequately captured what he was trying to convey, but nobody has ever reviewed what I wrote on this one:
http://tinyurl.com/m9geu3u

[Frank Restly]

Nick,

"3. It won't. Velocity is a real variable. If it were going to drop it would have dropped anyway. If anything, velocity will temporarily rise due to the temporarily higher expected inflation."

A rather bold prediction considering the evidence.

[Tom Brown]

Nick, you write in response to Squeeky Wheel:

"3. It won't. Velocity is a real variable. If it were going to drop it would have dropped anyway. If anything, velocity will temporarily rise due to the temporarily higher expected inflation."

I agree that velocity (V) is a real variable and should return to the same steady state value from which it started, all else equal, however your assertion that if anything it will temporarily drop seems to be at odds with your example in this post:

http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/03/the-sense-in-which-the-stock-of-money-is-supply-determined.html

"By cutting the rate of interest, the central bank increases the quantity of loans from the central bank, which creates more money. Eventually P and/or Y will increase and the quantity of money demanded will increase in proportion to the quantity created."

Recall in that example Md = P*Y. You described a scenario in which the quantity of money supplied (call it Ms) immediately rose when the quantity of loans rose, and then Md eventually moved through a combination of P and/or Y **increasing** (I'll take that to mean that neither P nor Y decrease) until Md = Ms.

Well, Ms = M = P*Y/V, right? So if M moves immediately to Ms > Ms0, and only "eventually" P and/or Y catch up by increasing, that implies that P*Y doesn't immediately increase, which implies that immediately V increases by a factor Ms/Ms0... and only eventually does it decrease to it's starting value, which it reaches when Md = Ms.

[Tom Brown]

"however your assertion that if anything it will temporarily drop "

should read

"however your assertion that if anything it will temporarily rise"

[Scott Sumner]

Nick, Excellent post. MMs have always claimed that it's not the current monetary base, but rather the expected future path of the base that matters. One of the things it matters for is the money multiplier.

Frank, There is a lot of empirical evidence that permanent increases in the money supply growth rate raise velocity (Cagan, etc.)

[Ralph Musgrave]

Nick, I don’t agree that commercial banks need base money for same reason as you and me need commercial bank money to do our shopping. If there was no central bank, commercial banks would let inter-bank debts lie for longer than they do in the real world: i.e. they wouldn’t settle up every inter-bank debt immediately using shares, property, etc. Moreover, any seriously in debt bank ought to see what’s going on and lend less, and thus in effect pay off its debts to other banks. Thus while exchanging property, shares, etc, is inherently expensive, it wouldn’t happen all that often.

In contrast, strikes me the real flaw in my above comment has to do with the fact that the state demands that taxes be paid using the state’s money, so the private sector (banks and non-bank entities) has to have a stock of the stuff.

[Tom Brown]

Nick, I can't believe how screwed up my question above is... replace this sentence too:

"which implies that immediately V increases by a factor Ms/Ms0... and only eventually does it decrease to it's starting value, which it reaches when Md = Ms. "

with

"which implies that immediately V decreases by a factor Ms0/Ms... and only eventually does it increase to it's starting value, which it reaches when Md = Ms. "

[Max]

I think you're being overly dismissive of "temporary" effects. In the 2008 financial crisis, people were liquidating assets because everyone else was liquidating assets, not because they thought the Fed's long range policy had changed. A "temporary" decrease in short term interest rates might have helped a lot.

[Tom Brown]

So let me try to be more succinct and clear. In your example Nick, we have starting off in equlibrium:

Md0 = Ms0 = P0*Y0 = P0*Y0/V0, where we then must have V0 = 1

After Ms increases to Ms1 we have immediately:

Ms1 = P0*Y0/V1, where V1 = V0*Ms0/Ms1 < V0 since Ms1 > Ms0, but because P and Y did not increase immediately, we still have:

Md = P0*Y0 (at time = t0 + epsilon)

Eventually P goes to P1 > P0 and/or Y goes to Y1 > Y0 such that we are in equilibrium again at:

Md1 = Ms1 = P1*Y1, and V1 = V0 = 1 again.

[Nick Rowe]

TMF: fair point. But then we get into "sins of ommission vs sins of commission" by the central bank, and "relative to what would have happened otherwise". Certainly, for a given time path of the base, lots of other things can change the money supply too.

Scott: thanks! Yep, this is really just apply to the money multiplier the same point we generally make about base and NGDP.

Max: Fair point. But that is more about the importance of how the bank responds to a temporary liquidity shock than about the effects of an exogenous shock to the central bank's behaviour.

Tom: if it takes time for the increase in the base to affect both NGDP and the money supply, V could go either up or down temporarily, depending on which one increases first. The algebra alone won't tell us that.

[Tom Brown]

"The algebra alone won't tell us that."

I agree, I was just trying to relate it to your previous example where P and/or Y seemed to be strictly increasing, not decreasing.

But I guess your broader point is that P and/or Y may be increasing and/or decreasing, but they eventually end up in a place in which their product is higher than their previous product so that Md = Ms, and V is back to where it started, which says nothing about it's trajectory to get there.

[Frank Restly]

Scott,

"Frank, There is a lot of empirical evidence that permanent increases in the money supply growth rate raise velocity."

How do you know that any increase in the money supply growth rate is permanent - because a central banker tells you it is? Central bankers get appointed and eventually replaced by politicians. Politicians get appointed and eventually replaced by the voting public. And so the permanence of any money supply is ultimately one decided by the voting public.

Also, how do you establish singular causality - supply of money caused an increase in the velocity instead of say:

1. Increase in legally obligated payments (taxes, interest payments, etc.)
2. Increase in supply / variety of goods
3. Increase in population

There may be some correlation to the money supply growth rate and the velocity of money, but I don't think there is enough to establish causality.

[Mark A. Sadowski]

Tom Brown,
Over at Money Illusion you said the following:

"Re: “money multiplier” … why not just call it the “broad to base ratio?” That’s save a lot of heartache wouldn’t it? Plus it’d save one syllable if you say “ratio” like many of us do."

This question caused me to go on a somewhat fruitless quest to try and figure out when the phrase "money multiplier" was first used. Along the way I found the following on the history of the simple model of multiple deposit creation:

http://www.richmondfed.org/publications/research/economic_review/1987/pdf/er730201.pdf

James Pennington seems to be the first one to explicitly explain the process of multiple deposit creation in writing in 1826. Alfred Marshall appears to have been the first one to give it mathematical expression sometime after 1877. But the current version found in most elementary and intermediate textbooks was originated by Chester Arthur Phillips in 1921.

The article also mentions that the formula relating the monetary base to broad money supply, involving the currency ratio and reserve ratio, was first derived by James Meade in 1934. The formula was explained again by Friedman and Schwartz in 1963, and Phillip Cagan in 1964, and this appears to be when it first became popular, particularly for historical analysis.

However, scanning these writings I can find absolutely no incidence of the phrase "money multiplier", not by C.A. Phillips, or Friedman, or anybody.

In fact the earliest closest approximation I can find of the phrase is by Karl Brunner in 1961, who used the phrase "monetary multiplier". This was as part of *his model* of the money supply creation process, which I'm sure most endogenous money enthusiasts would be absolutely appalled by.

Incidentally I think it was also in 1961 when Karl Brunner coined the phrase "base money". So it is with some irony that over time "monetary multiplier" seems to have changed to "money multiplier", and that "base money" has changed to "monetary base".

In any case, I'm somewhat stumped and cannot believe the phrase itself is of so recent an origin. Perhaps this is a question for David Glasner.

[Tom Brown]

I asked a simple question "why not call it the broad to base ratio instead of the money multiplier" at Scott's, and Mark really dug in!... so if you're interested in the history of the phrase "money multiplier" here's Mark:
http://www.themoneyillusion.com/?p=26518&cpage=1#comment-327879

[Tom Brown]

Shoot, next time I'll read before posting. :(

[Nick Rowe]

Mark: neat! My guess is that "multiplier" came into use because the Keynesian "multiplier" became commonplace. Just a guess.

[HJC]

Nick: Oh no, it looks like I've become a market monetarist when I wasn't looking! (Yes, thanks for fixing the typo. I feel very honoured to have triggered an entire post!)
I am uncomfortable with the base money change arriving as manna as per Patinkin. He was always careful to maintain the previous distribution of money when it was magically doubled. But this only really is applicable to the abstract unit of account. Central banks do repo to change the base, this is not neutral in its distribution, exiting debts are nominal etc, nor permanent. That's why I mentioned Metzler's model, he incorporated open-market operations and found that there is no unique full-employment real interest rate. It's not a monetary theory model either - price changes come via output changes, not directly from money changes.

[Tom Brown]

HJC, I notice you mentioned "unit of account." Do you happen to have a copy of Jurg Niehans book "The Theory of Money?" It's been a pet project of mine to try and figure out whether or not what I call "the definition" should be attached to the UoA or the MoA or should properly stand on its own as a separate concept. See footnote (1) here to JP Koning's post:
http://jpkoning.blogspot.com/2014/03/credit-cards-as-media-of-account.html

[Tom Brown]

"Oh no, it looks like I've become a market monetarist when I wasn't looking!" ... relax... grab yourself a purple robe and a cup of Kool-Aide and stay awhile.... Haha!... I'm actually fighting the urge to "join up" myself, and so far my doubts are strong enough to keep me out of the compound.

[HJC]

Tom: Yes I do have a copy. It's at work though, I'll dig it out on Monday. My reference here was more to Patinkin though.

[Tom Brown]

Awesome! JP thinks the "Tom Brown Multiple" as he dubbed it, should be attached to the MoA, and Sadowski thinks the UoA. Some of us are still confused (Marcus Nunes maybe?... but me certainly). Supposedly that book established a baseline, but neither JP nor Mark have a copy. Yes, I realize this is a bit of an off-topic request. But if my question makes sense, maybe the answer lies within. :D

[Tom Brown]

Nick you write

"Starting in equilibrium, one more loan of $1 is all we need to get the ball rolling towards the new equilibrium. Or one bank buying something for $1."

Do you think the "people of the concrete steppes" could be partially mollified if you proposed a definite channel through which the CB actually get's that going? Like telling them you definitely have on your list of things to do to throw a lit match on the pile of oily rags (even though you know in your heart it will burst into flames through spontaneously combustion anyway).

[Peter N]

"1. Capital is endogenous. Banks can and will raise more capital if it is profitable to do so. (And to the extent that banks have real capital, as opposed to nominal, it will rise automatically)."

Banks need to improve their capital to assets at risk ratio. They can do this by changing either the numerator or the denominator. If investors lack confidence in banks and price capital too high, it's more profitable to shed assets.

The situation is aggravated by balance sheet incentives to avoid recognizing nonperforming assets. This is tolerating insolvency to avoid illiquidity (zombie loans), something which governments often connive in (Japan and Spain for example), even though it makes the situation worse in the long run.

Also the change to Basel III capital rules has produced a large need for capital at a time when bank stock prices are very low and raising capital is unattractive.

Of course the Euro currency union and the special limitations of the ECB may make Europe a special case, though the example of Japan argues otherwise.

[JKH]

Regarding the use of the term "multiplier", this is from the first text I mentioned yesterday (Stager 1979):

"Although there is an important similarity between the national income multiplier and the expansion of bank deposits, particularly in the formulae for calculating the final results, these two processes should not be confused in any way. They deal with quite different features of the economy. The term "multiplier" is reserved for the particular process of an expansion or contraction in national income resulting from an initial change in spending. This term cannot be used for the expansion and contraction of the money supply. In fact, there is no similar generally accepted term in the latter case; one simply refers to the deposit adjustment process."

[Tom Brown]

JKH... does that constitute the "next page" that Sadowski was referring to?

[Scott Sumner]

Frank, Of course one can never be certain about anything in macro. But when theory predicts X and evidence suggests X then you have about as much certainty as possible.

It's true that the market never knows for sure if a given increase is permanent, but they do the best they can in forecasting.

[dannyb2b]

If the general public had reserve accounts with the fed increases in base would always be permanent right?

[dannyb2b]

continued from previous comment...

Increases in base would always be permanent if the public could directly hold reserves like commercial banks and reserves were capable of being used for transactions just like bank deposits. Reserves are transfered directly to the public on a non debt basis without purchasing assets like MBS or treasuries. Reserve expansion is classified as equity on CB balance sheet instead of being a liability. Reserve creation rate determined by ngdp or inflation targeting.

[Tom Brown]

dannyb2b,

"If the general public had reserve accounts with the fed increases in base would always be permanent right?"

Why do you say that? Suppose the Fed purchased $1 of assets, an then the next day it sold those assets. The base would go up $1 and then back down again.

[dannyb2b]

Tom Brown

sorry, check out my second comment for clarification

[Undergrad]

The way that I see the transmission of monetary policy is the follow: the central bank increases the quantity of monetary base supplied. This leads to a disequilibrium: banks, households, and firms are initially holding more monetary base than they collectively demand at the initial money market interest rate. Banks will attempt to reduce their excess holdings of monetary base by lending out their excess holdings of monetary base in the interbank market, purchasing money market instruments, etc. Households and firms will attempt to rid themselves of their excess holdings of monetary base by depositing them in interest-bearing short-term bank deposits, purchasing money market instruments, etc. Banks, households, and firms cannot collectively succeed; instead, money market interest rates will fall until the quantity of monetary base demanded has increased sufficiently to eliminate the initial disequilibrium. The lower money market interest rates will mean that the interest rate that banks must pay on their liabilities is lower. Thus, profit maximization implies that banks will supply more banks loans, lowering the interest rate on bank loans. This will cause the quantity of bank loans demanded to rise. A new equilibrium obtains with a higher quantity of bank loans, and thus, a higher quantity of bank liabilities, including bank deposits, which households are willing to collectively hold because of the lower interest rate. Aggregate demand rises because of the lower interest rate, which is financed through bank loans and other forms credit. This eventually causes the price level to rise. As the price level rises, the demand for the monetary base rises, and thus, by the reverse of the process explained above, interest rates rise back to their initial level. I have assumed throughout that the increase in monetary base is permanent, but the growth rate of monetary base has not changes so that expected inflation is pinned down (and so, I can fudge the distinction between real and nominal interest rates).

In all the above, the monetary base is of central importance, but broad money is not. Where am I wrong? Why does broad money matter? Why is the connection between the monetary base and broad money important?

[Tom Brown]

dannyb2b,

"Increases in base would always be permanent if the public could directly hold reserves like commercial banks and reserves were capable of being used for transactions just like bank deposits. Reserves are transfered directly to the public on a non debt basis without purchasing assets like MBS or treasuries. Reserve expansion is classified as equity on CB balance sheet instead of being a liability. Reserve creation rate determined by ngdp or inflation targeting."

Are these hypothetical "direct" reserves like coins are now (in an accounting sense)? Coins at not a liability on any balance sheet and take on their face value as soon as they are minted. Unlike paper reserve notes, the Fed purchases coins from Treasury at their face value, and while at the Fed they are an asset to the Fed. When the Fed sells $1 in coins to a bank, the Fed loses $1 of assets but also loses $1 of liabilities (either a paper note or an electronic Fed deposit was traded to the Fed for the coins -- and electronic Fed deposits and paper reserve notes both serve as Fed IOUs, losing their value when returned to the Fed).

So just for laughs say people can get coins directly from the Fed. What do they have to do to get them if they aren't selling the Fed assets? Is the Fed lending them the coins? Trading a coin asset for a loan asset?

With the real system we have, the CB can prevent the base from shrinking by simply not selling assets and also replacing ones that mature.

I'm not sure what problem you're trying to solve with your new Fed deposits (which I'm likening to coins, only because the accounting you describe sounds similar).

[Tom Brown]

Undergrad, by not caring about broad money you sound a little like Scott Sumner (to me). You might find this post of Scott's interesting if you haven't seen it:
http://www.themoneyillusion.com/?p=26400
In the second plot, he shows how an increase in the stock of money initially leads to a lower interest rate (moving from point A to point B) but then later prices rise (movement from point B to point C).

[dannyb2b]

These direct reserves would be electronic like the reserve accounts of depository institutions. Im saying if reserves as they exist now are modified so they can be held by anybody and also used as a means of payment like commercial bank deposits in order to increase the functionality of base.

Coins and notes are a liability from the viewpoint of the issuer and so are reserves. This is why asset purchases are performed when expanding base otherwise issuers balance sheet would be negative. Thats why I said that if reserves are issued directly to public without asset purchase or lending then they should be recognized on fed balance sheet as equity. This is more accurate anyway as the public are the feds constituent (assuming the fed is a public entity, if it isnt public then it should be). Financial assets issued to owners are recognized as equity on balance sheet like stocks of General Electric.

[Tom Brown]

"Coins and notes are a liability from the viewpoint of the issuer and so are reserves."

Actually coins are never a liability to any entity. They are only an asset to the entity which owns them.

"if reserves are issued directly to public without asset purchase or lending then they should be recognized on fed balance sheet as equity."

But why would they be Fed equity if someone other than the Fed owns them. Perhaps you can draw out the balance sheets and demonstrate what your are talking about.

The Fed is a hybrid entity. The member banks own its "stock" which I believe pays dividends, but it's a very strange type of stock. Obviously the government has a role in the operation of the Fed as well.

I'm still not sure what problem is being solved. Are you trying to solve the problem of the Fed not being able to directly control the stock of money? It can do that: it has absolute control over the stock of base money, but not absolute control over the stock of broad money.

Why would the Fed issue me these reserves? They just send them to me for free? How does one obtain these new reserves?

Here's an alternative idea... the government as a whole could issue either currency or deposits to pay for things creating money in the process: normal things that governments pay for like salaries, hiring contractors, building bridges, etc. Then when people pay taxes back to the government the money is destroyed. That's the MMT concept as I understand it. At least I understand in that scheme why money comes into existence and why it's destroyed again. But I'm not getting that with your scheme. The balance sheets would help! :D

[dannyb2b]

Tom Brown oops coin is an asset. Reserves are a liability though it doesnt change my example.

[dannyb2b]

"But why would they be Fed equity if someone other than the Fed owns them. Perhaps you can draw out the balance sheets and demonstrate what your are talking about."

I am saying the fed should be a public entity as it creates the national currency. Not quasi public.

Financial assets can be recognized as a liability or equity. Base is a financial asset from point of view of holder. It is more accurate to record these as equity than liability if issued to the Feds constituent (assuming fed is a public entity).

"I'm still not sure what problem is being solved. Are you trying to solve the problem of the Fed not being able to directly control the stock of money? It can do that: it has absolute control over the stock of base money, but not absolute control over the stock of broad money."

The problems being solved relate to monetary policy effectiveness and efficiency.

If the money supply can be expanded on a non debt basis by the fed the demand for money can be met without needing an increase in debt or needing less debt. Higher debt to GDP is correlated to greater financial instability and longer lasting recessions. Deposits are generally expanded through lending which pushes debt upwards.

Higher functionality of base will increase the velocity of the money supply.

Money transferred to peoples account without asset purchases are permanent increases in base as opposed to through asset purchases they might not be.

The ZLB isn't a problem either if base is sent straight into peoples accounts.

Money transfers to the general public have a higher MPC than proceeds from asset sales (QE or OMO's) by large corporates.

"Why would the Fed issue me these reserves? They just send them to me for free? How does one obtain these new reserves?"

The amount of new reserves created will be relative to ngdplt. Yes they would be free.

"Here's an alternative idea... the government as a whole could issue either currency or deposits to pay for things creating money in the process: normal things that governments pay for like salaries, hiring contractors, building bridges, etc. Then when people pay taxes back to the government the money is destroyed. That's the MMT concept as I understand it. At least I understand in that scheme why money comes into existence and why it's destroyed again. But I'm not getting that with your scheme. The balance sheets would help! :D"

The government and the fed need to be separate. Conflicts of interest and excessive burden on the government of managing monetary policy. There needs to be an independent focus on monetary policy and its importance and independence shouldn't be underplayed. Just like the judiciary should be separate to the government. The gov can just tax to fund any project it needs anyway.

[Tom Brown]

"If the money supply can be expanded on a non debt basis by the fed the demand for money can be met without needing an increase in debt or needing less debt. Higher debt to GDP is correlated to greater financial instability and longer lasting recessions. Deposits are generally expanded through lending which pushes debt upwards."

What if the CB just purchased assets other than Treasury debt when there was an excess demand for money, and then sold them again when there was an excess supply? The CB could still keep at approximately zero equity, although it doesn't matter too much, because it won't go bankrupt. What if the assets it buys lose value and it sells them all but needs to further reduce the money supply? Well then it could sell some kind of time deposits maybe. Shoot I don't know...

I don't think I have what it takes to give you the feedback you're looking for. Maybe Nick or someone else can help you! Good luck.

[dannyb2b]

"What if the CB just purchased assets other than Treasury debt when there was an excess demand for money, and then sold them again when there was an excess supply? The CB could still keep at approximately zero equity, although it doesn't matter too much, because it won't go bankrupt. What if the assets it buys lose value and it sells them all but needs to further reduce the money supply? Well then it could sell some kind of time deposits maybe. Shoot I don't know..."

Still not getting around the limitation that to increase broad money supply increased lending has to occur to expand deposits. If central bank money is equal to deposits in terms of functionality then less demand for deposits and hence lending to expand money. A greater proportion of broad money will be Base. Also from people's viewpoint reserves will be a superior electronic money because deposits a the fed will be safer than at commercial banks.

[Tom Brown]

"Still not getting around the limitation that to increase broad money supply increased lending has to occur to expand deposits. If central bank money is equal to deposits in terms of functionality then less demand for deposits and hence lending to expand money. A greater proportion of broad money will be Base. Also from people's viewpoint reserves will be a superior electronic money because deposits a the fed will be safer than at commercial banks."

Say for example the Fed bought land, gold and corporate stock directly from citizens. I don't see how borrowing has to happen to expand deposits in that case (Fed deposits for citizens). The Fed could lend directly to citizens too I guess (which of course would create debt).

I don't see how you sell the concept of the Fed crediting everyone's deposits for free. IMO that's going to be a tough sell politically... tougher than MMT probably... or even straight out socialism.

At a deeper level, I'm not aware of the problems with the debt based system you claim, nor do I have any way to check your claims... I'm afraid I'd be very skeptical no matter what. I'd be fascinated though to sit back and see you have a conversation with Nick or somebody else that's more qualified than I am!

[Tom Brown]

Actually, Nick is not a fan of representing reserves and currency as liabilities of the central bank. I don't really get why (it seems like a natural thing to do to me), so maybe you and he have a bit of overlap on that score anyway! He wrote at least one post on the subject... you can try to find it using the search box at the top.

[dannyb2b]

"Say for example the Fed bought land, gold and corporate stock directly from citizens. I don't see how borrowing has to happen to expand deposits in that case (Fed deposits for citizens). The Fed could lend directly to citizens too I guess (which of course would create debt)."

What about all the people that dont have assets? These have the highest MPC. Imagine money transfers to all the students unemployed etc...

"I don't see how you sell the concept of the Fed crediting everyone's deposits for free. IMO that's going to be a tough sell politically... tougher than MMT probably... or even straight out socialism."

The key aspect is that money gets distributed evenly everywhere by interacting with everyone. Its like blood flowing to the whole body as opposed to just the arm. The amount of money creation wont be much per person, it will almost be like a tax return. Its rightfully free because a public entity is already owned by the public so issuing something to yourself free is correct I suppose. I understand your point but I think people will resist this idea because they don't really understand the nature of a public institution. Its not like this organization is giving stuff away to random people, its issuing paper to its owners.

[dannyb2b]

"At a deeper level, I'm not aware of the problems with the debt based system you claim, nor do I have any way to check your claims... I'm afraid I'd be very skeptical no matter what. I'd be fascinated though to sit back and see you have a conversation with Nick or somebody else that's more qualified than I am!"

http://www.oecd.org/eco/public-finance/Debt-and-macroeconomic-stability.pdf
http://www.frbsf.org/economic-research/publications/economic-letter/2014/march/private-credit-public-debt-financial-crisis/

[Tom Brown]

Danny, I read what you had to say above. It does sound like a radical proposal to me, and honestly I would love to see a good critique... but it's well outside of my capability. You might be absolutely right, but I wouldn't know where to began with it.

On a more prosaic topic, I did find that post Nick wrote:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/03/is-modern-central-bank-money-a-liability.html

[dannyb2b]

I know that Mark Sadowski wont agree that higher income people have a lower MPC becuase he only agrees with measures of income that dont include cap gains. Even though cap gains income is called income he doesnt agree that its income becuase it doesnt count towards GDP.

But transfer payments to all people including the unemployed and students should realize higher MPC than money received from asset sales.

[Tom Brown]

Thanks for the links... ...I'm happy they are both short. You seem determined to have a convo on this. Just to let you know where I'm at... I just ran a brief errand, but the whole way back I was trying to work out Marginal Purchasing.... Marginal Propensity... hmmm... well I just googled it, and now I know what MPC is. Lol. OECD is another new one for me. I'm glad to see it's not a Maoist youth organization.

That'll keep me busy for a bit.

[dannyb2b]

This proposal toward the central bank is very capitalist in that people are seen as shareholders. The propensity to spend of money received is important becuase it will determine the effect on ngdp of money received.

[Tom Brown]

Well, I don't know enough to have an opinion on MPC yet. Also, I don't know if you saw this or not, but I thought it was interesting: Sadowski lists his opinion the goodness and badness of various kinds of taxes:

http://www.themoneyillusion.com/?p=26468#comment-326323

"In other words taxes can be ranked according to their effect on long run growth in the following fashion:

1) Capital taxes (very bad)
2) Labor Taxes (slightly bad)
3) Consumption (moderately good)
4) Property (very good)"

We had a little bit more of a conversation on it here:
http://diaryofarepublicanhater.blogspot.com/2014/03/capital-gains-taxes-paul-krugman-and.html?showComment=1396279611063#c4642485410534569476

So by "consumption tax" ... well that could mean "luxury tax" basically, according to Mark. I was most intrigued by the "property tax" ranked at "very good." I have not followed that debate at all though... I went and posted Mark's comments on Mike's site because Mike makes no bones about being a liberal and is SUPER skeptical of MMists, especially Sumner. Nick says Mike has "Sumner derangement syndrome." So basically I was amusing myself. Sadowski calls Mike a "troll blogger" ... ha!

Well, Mark made some comments there he didn't make on Scott's site, so I thought you might like to know about that. I only dimly was aware of your previous discussion w/ Mark on this subject... I was interested in other things. If you've completed a degree in economics you're way beyond me ... there's just a few things I zero in on purely out of my own curiosity. MPC isn't on my radar.

[dannyb2b]

I dont like the idea of taxing any type of income whether it be labour, capital (profits) or capital gains. I think all taxes should be levelled against spending like consumption, investment and asset purchases. This is becuase taxation affects incentives. You want to incentivise people earning incomes. I think I am more in favor of higher taxes against consumption and lower against investment spending. Spending on asset purchases like (property, shares) I think should be taxed. Also efficiencies would be found in taxing from fewer sources (just spending and opposed to spending and income) IMO becuase it is easier to perform.

I guess I agree with Mark. Since consumption accounts for about 65% of gdp you have to careful in taxing this as it could have a big effect gdp in short term. I would like to see an economy where a greater portion of gdp is investment and smaller is consumption so I would carefully try to increase taxes on consumption and reduce taxes on investment marginally in order to incentivise a more investment oriented system. Taxation isn't going to be the only tool that can address the underinvestment issue though.

[Tom Brown]

What about a "property tax?" That's what Mark ranks as the best.

[Squeeky Wheel]

Nick - thanks for indulging me. However, I'm still not convinced of the causality. So I'm pushing on corner cases.

"1. Capital is endogenous. Banks can and will raise more capital if it is profitable to do so. (And to the extent that banks have real capital, as opposed to nominal, it will rise automatically)."

I originally posited that the banks were capital constrained. So the original setup assumes that either the banks are unable to raise capital or it was not profitable to do so. Why would an increase in monetary base make an otherwise unprofitable capital raise become profitable?

"2. First, solve for the equilibrium, where P and NGDP are doubled, so nominal money demand is doubled, and nominal loan demand doubles."

The scenario was that there were no credit worthy borrowers (an extreme case of all subprime is underwater). Dead-beat squeeky wheel can demand all the loan he wants, but no bank will lend out the money. Nick in this case has maxed out his credit worthiness. Nick will need to get a raise *before* any bank will lend to him. Since no bank will lend to SqueekyWheel to pay more for worthwhile blog posts, Nick can't raise prices, can't borrow money, and then no change in price level.

You argued many times before that inflation requires that sellers be induced to raise their prices, and I don't see the inducement. I only raise my prices if I think everyone else has more money. Until the monetary base changes results in more deposits or currency, everyone else doesn't have more money, so I don't raise prices.

One last thought I've had for a while is on the CB's ability to force an expansion. Assuming a bond/reserve swap style of QE and a non-coercive legal structure, the CB cannot force anyone to sell bonds. This was my argument against Williamson - selling bonds doesn't force inflation down, the inflation must be down first before I sell the bonds. So two scenarios could be constructed if the CB wants to do QE: 1) CB tries to buy bonds at or below pre-QE prices. Given the increased overall demand, no one agrees to sell to the CB at the old prices and QE fails to occur. Or 2) the CB buys at the new prices - ie higher prices than before the announcement. In this case the nominal value of all nominal assets increases and that itself forms the start of inflation (through wealth effect, etc). it strikes me that the ratio is empirically true, but the mechanism is needed for causation. Something has to induce prices to rise - the CB can do that through buying something (bonds/securities?) at higher prices. However if the CB simply prints a lot of currency and locks it in a vault, the base increases, velocity decreases, and nothing happens (until the vault is unlocked).

[Tom Brown]

property tax = wealth tax I think. It's not on income or consumption I guess.

[dannyb2b]

Im not a big fan of property taxes other than to pay for whatever services the state provides to owners of property.

[Tom Brown]

Squeeky Wheel,

"However if the CB simply prints a lot of currency and locks it in a vault, the base increases,..."

Actually, I don't think that's true: the face value of reserve notes doesn't actually get created until those notes leave the Fed for the private sector. They lose their face value again when they are returned to the Fed. The Fed buys them for production cost (not face value) from the Bueaur of Engraving and Printing. Coins are different.

[Tom Brown]

Of course I don't know the details in Canada!... could be different I guess.

[dannyb2b]

Under current system taxing wealth is good because wealth is so concentrated. But otherwise no I think. It depends on the system.

[Squeeky Wheel]

Tom,

I'm not an accountant, but your stuff looks good. I was trying to capture the idea that an increase in bank reserves does not necessarily cause bank loans. As JKH has often pointed out, as a practical matter bank loan officers don't consider the bank's reserve position when deciding on whether to make loans. So I'm guessing that there can be other (real?) constraints on bank lending which would result in a monetary base increase having less effect on lending than the 'multiplier' would normally imply.

[Tom Brown]

... coins are different, but still if they aren't in a commercial bank vault or out in circulation, they aren't base money. Thus they aren't base money if they're stored at the Fed:
http://en.wikipedia.org/wiki/Money_supply

[Squeeky Wheel]

Ralph,

"thus while exchanging property, shares, etc, is inherently expensive, it wouldn’t happen all that often"

I disagree. When Singapore introduced the FAST payment system (< 30 sec round-trip inter-bank retail transfers - similar to the UK system), it went from settlement then clearing to clearing before settlement. In effect the banks became exposed to eachother's settlement risk. Despite this being for low valued retail transactions, the banks required a fully collateralized system - all banks must post collateral for their maximum settlement upfront.

I suspect that a real property based settlement system would be similar.

[Ramanan]

Nick:"...allows it to control the money supply"

Controlling something means predictively making that something move in the desired direction. The central bank cannot do that. A car driver has a good control overall on the speed and direction of the car. This is not the case with "controlling the money supply". So the analogy is far from perfect.

Of course the central bank has an influence on credit conditions and hence the money stock but that is a slightly different thing than saying "controlling".

And before anyone points out, these are not semantic things. Whenever central banks have tried to control the money stock by announcing targets, they have failed miserably.

[Tom Brown]

Squeeky, thanks. I'm not an accountant either. I got interested in how all that from Cullen Roche, JKH, and Fullwiler, but Carney's argticle filled in the capital requirements part for me. I heard from multiple sources that "reserves don't limit banks' lending, capital does." I have since learned to qualify that a bit... reserves don't constrain bank lending *as long as* the CB has a completely accommodative policy (i.e. it's targeting a fixed overnight rate). However, say the reserve level in the US is $2.5T and checkable deposits are at $1.5T (Those figures are in the ball park), then lending could be constrained in a hurry if the reserve requirement were raised to 2.5/1.5 = 167%, with no further accommodation from the Fed (i.e. no more OMOs).

Inflation targeting (I.T.) is less accommodative than overnight rate targeting. Fixed MB level targeting is not accommodative at all. I see it as a matter of degrees, like I see endogeneity and its complement, exogeneity. So I do think that MB constrains bank lending under I.T., but not at a hard limit: more like it sets a mean and variance around which it can bounce. Certainly both MB and the overnight rate will have a degree of endogeneity under I.T., but MB won't be as endogenous (and the overnight rate won't be as exogenous) as under interest rate targeting.

My current favorite analogy is that it's like a set of jointly Gaussian variables. Then the magnitude of the correlation coefficient between any two of them describes the degree of exogeneity of one wrt the other. The complement of this magnitude describes the degree of endogeneity. That's how I visualize it!... nothing official.

[Nick Rowe]

Ramanan: One important difference is that my car has a good speedometer, that responds very quickly to the car's actual speed. Central banks get data on the money stock with a lag. Cruise control cannot even see whether the car is going uphill or downhill, which has a big effect on the speed, but cruise control can very quickly adjust for the hills, even though it can't see the hills, because it can "see" the speedometer.

A second important difference is that my car has finite power and brakes. There is no limit to how much money a central bank can print, unless it runs out of paper and ink. Or buy back, unless it runs out of assets to sell.

The Bank of Canada abandoned money targeting, not because it couldn't target money, but because it saw that the relationship between money growth and inflation wasn't constant enough, and it wanted to bring inflation down, so it switched towards targeting inflation directly.

[Tom Brown]

... and the other big difference is he's actually riding a horse instead of driving a car. :D

[Nick Rowe]

HJC: Yep. I can't remember the last time I hoisted something from comments, and wrote a post about it. You are very honoured!

But it was a very good comment. And it was a good criticism. The only good substantive criticism of the money multiplier I have come across, despite all the haters out there.

You are now a Market Monetarist, whether you want to be or not!

Yes, to get strict Quantity-Theory results, a la Patinkin, new money would have to be helicoptered down proportional to existing stocks (and new bonds too). But the only difference this makes is distribution effects. A back of the envelope calculation suggests it doesn't make a big difference if the central bank uses OMOs or helicopters to increase the base. Base is around 5% of NGDP, so if we double the price level by permanently doubling the base, that is a one-time inflation tax of 2.5% of GDP used to reduce the national debt by 2.5% of GDP. A much bigger non-neutrality would be the effect of the higher price level on the real value of the existing bonds. That would halve the debt/GDP ratio. Plus the non-neutralities that result from sticky prices.

So I don't think the differences between OMO and helicopter money are big enough to worry about. Especially since that calculation is for a doubling of the base and price level, which is a very big change in monetary policy.

[Nick Rowe]

dannyb2b: I do have a bank account with the Bank of Canada. There are several Bank of Canada $20 notes in my wallet. Sure, it's a paper account, and not an electronic account, but it's the same thing. I have an account with the Fed too, and the Bank of England.

[dannyb2b]

"I do have a bank account with the Bank of Canada. There are several Bank of Canada $20 notes in my wallet. Sure, it's a paper account, and not an electronic account, but it's the same thing. I have an account with the Fed too, and the Bank of England."

Yeah but electronic money is generally more efficient. Otherwise probably nobody would have bank accounts. The fed could expand money by posting it to every citizen or just credit electronic accounts.

" Central banks get data on the money stock with a lag. "

Not the base

"The Bank of Canada abandoned money targeting, not because it couldn't target money, but because it saw that the relationship between money growth and inflation wasn't constant enough, and it wanted to bring inflation down, so it switched towards targeting inflation directly."

So instead it decided to target interest rates. Hmmm

Central banks have perfect data on Base becuase they created it all.

[HJC]

Nick: But (sorry to harp on) Metzler's paper showed that there is a difference between helicopter money and OMOs. And does your doubling of money double the price level if we are not at full employment? I don't think so. (I need to keep my post-Keynesian credentials intact!)

[dannyb2b]

HJC

Do you have a link to Metzler's paper?

[HJC]

Its title is 'Wealth, Saving, and the Rate of Interest' by Lloyd A. Metzler, 1951. I have it in a book, so sorry I don't have a link. Maybe it's not as influential as it perhaps was back then.

[Nick Rowe]

HJC: there is indeed a difference between helicopter money and OMO. I'm saying it's not very big. For a doubling of the base and price level, the difference is equivalent to a change in the debt/GDP ratio of 2.5% of GDP. Base is small, so the number of bonds bought in an OMO to double the base is small too.

If we start away from full employment, we still get a doubling of the price level, **relative to what would have happened otherwise**. (Of course, that might mean the price level would have halved otherwise, and so doubling the base had no effect on the price level, ** relative to what it was previously**.)

[Nick Rowe]

danny: "So instead it decided to target interest rates. Hmmm"

No it did not. Instead it decided to target inflation. I don't know of any central bank that targets interest rates. The only people who think otherwise have very short time-horizons. Like 6.5 weeks.

[HJC]

Nick: Got it, point taken. Your second paragraph, have you got any extra blog posts on that?

[dannyb2b]

I thought the BOC abandoned monetary targets to target the interest rate.

http://www.bankofcanada.ca/core-functions/monetary-policy/key-interest-rate/

Its now targeting the interest rate to target inflation right? Before it was targeting the supply of money to target inflation, growth and employment?

[Nick Rowe]

HJC: I think I did a blog post on it years ago, but I don't think I could find it. But it wouldn't add much to what I have just said.

A lot of debates about whether X causes Y under what conditions could be avoided if we were all just clearer about this distinction. "Other things equal" is another way of talking about the counterfactual conditional, but it isn't as clear.

danny: at a very high frequency (hours and days) the BoC adjusts the base using repos to hit its target for the overnight rate of interest. At a lower frequency, it adjusts the target for the overnight rate 8 times per year (if needed) to hit its inflation target.

In the late 1970's, it announced a declining target path for M1 growth rate. And it adjusted the interest rate to try to hit that monetary target. And it adjusted the base to adjust the interest rate. Then it abandoned the target path for M1, and replaced it with an informal (implicit) target for inflation, then a formal explicit target for inflation.

You get a very different picture, depending on whether you zoom in or zoom out. You see different patterns depending on the zoom.

[dannyb2b]

Nick Rowe

Do you agree that if the fed issued money directly to public in expansionary monetary policy the money supply would expand without debt expanding as much as the current system because a greater proportion of broad money would be base as opposed to bank deposits. Bank deposits are generally created through lending.

As a result of lower debt greater monetary policy efficiency would prevail because higher debt to GDP is correlated to greater financial instability and longer lasting recessions.

[Nick Rowe]

danny; reframe; fixed exchange rates are riskier than flexible exchange rates. Beta banks are riskier than alpha banks, because beta banks have fixed exchange rates.

[dannyb2b]

Nick Rowe

So if I understand correctly banking with the alpha bank is preferable than with the beta banks. If the fed is alpha then it would be superior if the general public could hold electronic accounts with it instead of only the beta banks. Does this mean we agree?

[Tom Brown]

Frances Coppola has a post on the money multiplier now too:
http://pragcap.com/on-the-persistence-of-inadequate-ideas/comment-page-1#comment-172291

[Tom Brown]

And yet another view of the money multiplier, this one from Jason Smith... (I get a mention in this one):
http://informationtransfereconomics.blogspot.com/2014/04/broad-money-narrow-money-and-interest.html

[Tom Brown]

... well now that I look more carefully, perhaps "money multiplier" isn't the right way to describe it, but it does have to do with the ongoing discussion between Nick Rowe and David Glasner.

[Mark A. Sadowski]

Among endogenous money enthusiasts Ramanan often writes things which I find interesting and non-objectionable. However, in this particular instance I have to pick on him.

http://www.concertedaction.com/2014/03/31/the-phrase-money-multiplier-itself-is-inaccurate-and-misleading/

March 31, 2014

The Phrase Money Multiplier Itself Is Inaccurate And Misleading
By Ramanan

"Some people

http://www.themoneyillusion.com/?p=26479

point out that the critique “there is no money multiplier” is wrong because it is a ratio whatever said. No! The phrase “money multiplier” itself is wrong because the phrase itself captures a wrong causal story. A phrase is a small group of words standing together as a conceptual unit and hence the phrase “money multiplier” is inaccurate and misleading.

So take the textbook Keynesian multiplier first. It suggests that a rise in government expenditure leads to a rise in output more than the increase in the expenditure. The ratio of rise in output to the rise in expenditure is the multiplier.

But this is not the case with the “money multiplier”. There is no direction of causality from a rise in bank settlement balances to the rise in the money stock. This is true even if the central bank is doing QE/LSAP, i.e., purchasing assets on a large scale. So if the central bank purchases government bonds in the open market, it leads to a rise in banks’ settlement balances at the central bank and also a rise in the money stock. But the rise in the settlement balances could not have been said to have caused the rise in broad monetary aggregates such as M1, M2 etc. It is the act of the central bank purchase which leads to a rise in the stock of both narrow and broad monetary aggregates..."

The first problem I have with this is the implicit assumption that the statement "a rise in government expenditure leads to a rise in output more than the increase in the expenditure" is obviously correct. In fact one of the most notable facts about the Keynesian multiplier is that empirical research has had great difficulty in detecting nonzero Keynesian multipliers, never mind multipliers greater than one, perhaps due to problems of endogeneity, but more likely due to the fact that monetary policy should routinely offset fiscal policy, unless the central bank is incompetent or lazy.

The second problem I have with this statement is the absurdity of claiming that the rise in reserve balances cannot be said to have caused the rise in deposits since it was the "act of the central bank purchase" that caused both. QE literally means a targeted increase in the monetary base (or its reserve balance component in the case of Japan) so how can its effect on deposits be disentangled from its effect on reserve balances?

In fact, provided QE consists of purchases of assets held by a non-banks (as has been effectively true, as of today, in the case of 100% of all QE done in the US and the UK), it creates a simultaneous entry on the liability and asset side of commercial bank balance sheets in the form of a commercial bank deposit-reserve balance deposit pair. How can one say that the the central bank's intent of creating a reserve balance deposit did not cause the commercial bank deposit, since, as QE has been practiced, one could not come into existence without the other?

Moreover, using similar techniques to Post Keynesian economists such as Basil Moore, Thomas Palley, Robert Pollin etc. in empirical endogenous money research, I find that the US monetary base has Granger caused commercial bank loans and leases during the period of QE. So if "loans create deposits", as the endogenous money enthusiasts keep chanting, this implies that QE, by encouraging loan creation, has catylized the creation of even more deposits.

Ramanan continues:

"...Now to the case of no QE.

Same story: the rise in banks’ settlement balances could not have been said to have caused a rise in broad monetary aggregates. The more appropriate phrase is “credit divisor”. Here’s Marc Lavoie from his 1984 paper The Endogenous Flow Of Credit And The Post Keynesian Theory Of Money..."

This reminds me of the several things I have read on how the "Keynesian multiplier" should really be called the "Keynesian divisor" by Kevin Hassett, Charles Gave, and so on:

http://muhlenkamp.com/upload/pdf/keynsianDivisor.pdf

none of which I give much credence to either.

Something tells me that economics textbooks are not going to replace the word "multiplier" with "divisor", in the case of either the monetary or Keynesian multiplier, in the near future.

[HJC]

Mark: Hi, if the reserve deposit and the bank deposit "could not come into existence without the other" (in non-bank QE), then any logical inference about direction of causality could work equally well either way. You could just as well say that the creation of the bank deposit caused the creation of the reserve deposit. Ramanan's insistence that the purchase caused both seems pretty sound, for the first round effects anyway.
On your point about Granger causation, it was my understanding that bank credit had largely collapsed in the US and the UK and that instead it was QE keeping broad money stable. I wasn't aware of QE encouraging loan creation (ignoring student loans), have you got any links for this?

[Philippe]

Tom,

"I don't see how you sell the concept of the Fed crediting everyone's deposits for free. IMO that's going to be a tough sell politically... tougher than MMT probably... or even straight out socialism."

Notorious socialist Milton Friedman suggested the idea, he referred to it as dropping money out of a helicopter.

[Mark A. Sadowski]

HJC,
The goal of the purchase is to create a reserve deposit. If the purchase must create a commercial bank deposit in order to create a reserve deposit it seems to me that the immediate cause of the commercial bank deposit is the central bank's decision to create the reserve deposit.

"On your point about Granger causation, it was my understanding that bank credit had largely collapsed in the US and the UK and that instead it was QE keeping broad money stable. I wasn't aware of QE encouraging loan creation (ignoring student loans), have you got any links for this?"

US commercial bank loans and leases are up by about $1.03 trillion in four years, or by 15.8%, and consequently are at record levels in nominal terms:

http://research.stlouisfed.org/fred2/series/TOTLL?cid=100

But even if that weren't true, the question isn't whether commercial bank loans and leases are up or down, but whether they are higher than they would have been in the absence of QE.

And, between 2008Q2 and 2013Q4, US deposits (demand, savings and time) increased by $3.12 trillion, or by 36.2% according to the flow of funds. During the same period of time reserve balances increased by $2.22 trillion.

I have no links for this but I will describe what I have done and am willing to email the estimation output (I've done as much for Dan Kervick for example).

I did my analysis using the Toda and Yamamato technique in Eviews. The data is the monetary base and commercial bank loans and leases at a monthly frequency over the period December 2008 through September 2013. The minimum order of integration such that both the monetary base and commercial bank loans and leases are stationary is four.

I set up a two equation VAR in the levels of the data including an intercept for each equation. The various information criteria suggest a maximum lag length of 2 for each variable. An LM test shows no serial correlation in the residuals.

Then I restimate the levels VAR with four extra lags of each variable in each equation. But rather than declare the lag interval for the two endogenous variables to be from 1 to 6 I left the interval at 1 to 2 and declared the four extra lags of each variable to be exogenous variables.

When I do the Granger causality test I fail to reject the null that commercial bank loans and leases does not cause the monetary base but I reject the null that the monetary base does not cause commercial bank loans and leases at the 5% significance level.

I've also conducted Granger causality tests on UK monetary base and the M4 lending counterpart and the M4ex lending counterpart over the period from May 2009 through September 2013, but the results are not statistically significant.

[dannyb2b]

Phillipe

"I don't see how you sell the concept of the Fed crediting everyone's deposits for free. IMO that's going to be a tough sell politically... tougher than MMT probably... or even straight out socialism."

"Notorious socialist Milton Friedman suggested the idea, he referred to it as dropping money out of a helicopter."

Base expansion in normal times works out to about 500 dollars a year per person. About 10 dollars a week if expansion is performed in weekly increments. Thats hardly socialism. Its capitalim. Its just issuing financial assets to owners of central bank which are the public (or it should be public anyway).

[HJC]

Mark: Thank you for your detailed reply. On the first point, assuming your interpretation of the goal of QE, then I don't suppose you can be wrong, but that wasn't my impression of its goal. Otherwise the talk of loan size counterfactuals and econometric testing takes me well beyond what I can offer in the time I have available, sorry. But it does look like valuable and I may get a chance to get back to it.

[Tom Brown]

Phillipe and dannyb2b, if you can sell those ideas politically, great! If anything I'd guess I'm an easier sell than your average American. My impression of Milton Friedman (from what little I know of him) is he was probably too right wing for my tastes. And I notice we currently don't have money dropping from helicopters or negative income taxes either for that matter. I was just giving you my impression of your chances. I'm not much of a revolutionary or activist... much more of an incrementalist, and thus I'd think if anybody has a chance of selling a new idea, it might be MMists... and the world isn't exactly beating down their door for advice either I notice. So good luck!... and if you can sell one of the MMists on your ideas I'd truly be very impressed... which reminds me dannyb2b: Rowe never confirmed or denied that he agreed with you, right? But did you hear back from him anywhere else?

[Tom Brown]

Philippe (sorry for misspelling your name above) and dannyb2b... I'd posted a question to Miles Kimball recently, and he answered, and it occurred to me that his proposal here (which he pointed me to in his answer):
http://blog.supplysideliberal.com/post/67342414250/pieria-2-the-costs-and-benefits-of-repealing-the
is something you're both probably aware of, right? How does that stack up against yours in terms of being incremental vs sweeping? Also, what's your view of it? Here's my question to Miles BTW:
http://blog.supplysideliberal.com/post/23959666073/what-is-a-supply-side-liberal#comment-1320626945
My impression up till now has been that Miles' proposal was less likely to fly politically than MM, but this will be the 1st time I really give it a good look, so perhaps my view will change.

[dannyb2b]

Tom Brown

I think my proposal is incremental too. The fed is doing the same thing in my proposal just with a broader set of counterparties thats all. The BOE is actually begining to offer reserve accounts to non bank corporates too. I think this is key so that debt to gdp is lowered and spending from base increases is higher. Its pretty simple really.

Im not suggesting we abolish central banking or subsume it into the congress entirely or anything.

[dannyb2b]

Negative interest rates are unnecesary. The only reason we are at the ZLB is becuase demand for reserves is suppressed. If the broader public could hold them then transactions demand for money would shift from deposits to reserves and rates would go up.