Feed the dragon (but not too much)
Yesterday I developed a monetary post using a dragon parable. In a bizarre coincidence Frances Woolley has beaten me to it. Here’s my version:
There’s a enormous dragon living just outside a village of peasants. If the dragon is not adequately fed it lashes out and destroys part of the village. If it’s overfed then it produces lots of noxious gases. This is not as damaging as an underfed dragon, but not optimal either. There is a bell that rings whenever the dragon needs to be fed, but the bell often malfunctions.
I’d like to use this parable to better explain my post from a few days back, where I argued that the recent US recession was triggered in late 2007 by the Fed’s failure to provide enough $100 bills for tax evaders, drug dealers and foreigners (TDFs).
It just so happens that TDFs have a growing demand for $100 bills. If the Fed doesn’t meet that demand, we get a recession. If they feed too many $100 bills into the economy, then we get excessive NGDP growth (aka “demand-side inflation.”) Not as bad as a recession, but not optimal either.
This post met with lots of objections. Some pointed out that the Fed can’t control how much of the new base money goes to the TDFs in the form of $100 bills. That share is “endogenous.” Yes that’s true, but has no bearing on my post. The same is true of the M2 money supply. The Fed controls the base, not the ratio of M2/MB. That ratio (called the “M2 multiplier”) is determined by the public (and the Fed via reserve requirements.) And the ratio of $100bills/MB is also determined by the public. However the Fed can target either M2 or the quantity of $100 bills, if it wishes to do so.
My point was that a recession occurred because the Fed did not meet the rising demand for base money in late 2007 and early 2008. And secondarily, that that increased demand came mostly from TDFs wishing to hold more $100 bills. Go back to the dragon analogy, and now assume there are three dragons. The big one represents the hunger for $100 bills by TDFs. Another dragon, less than half as large, represents the demand for currency for transactions. And a third, roughly 1/10th as large, represents the demand for bank reserves in 2007. The villagers throw food into the dragon den, but can’t control how the dragons divy it up. Damage from underfed dragons is proportional to the size of the dragons, and the extent to which they are underfed.
Another criticism was that the entire monetary base is endogenous. Nick Rowe has some excellent posts that clarify the “endogenous money” issue. Whenever the central bank pegs one variable (say exchange rates, gold prices, or M2, or the base, of the fed funds rate, or inflation), then all other variables become endogenous. But once again, that has no bearing on my argument. I wasn’t claiming the Fed targeted the base, much less $100 bills. They don’t, and they shouldn’t. They tend to target the fed funds rate in the very short run, and then adjust the fed funds rate every so often in order to target inflation or NGDP over longer periods of time. But that fact has no bearing on whether a shortage of $100 bills caused the recession.
Let’s see how things change if the Fed has a fed funds target. Some people argued that any extra demand for $100 bills would be smoothly accommodated by the Fed, which would supply enough cash to keep interest rates stable. Like the bell in the dragon example, the fed funds rate often does send out timely signals. But on occasion the bell fails to ring when the dragons are hungry, and on occasion the fed funds rate does not send out a warning that the TDFs need more $100 bills. Late 2007 and early 2008 was one of those failures. You could say the recession was “caused” by a malfunctioning bell, but I prefer to say it was caused by underfed dragons lashing out at the villagers.
The interest rate signalling mechanism only works well if the Wicksellian equilibrium interest rate is constant. In late 2007 and early 2008 the Wicksellian equilibrium rate fell sharply. The Fed did cut the actual fed funds target somewhat, but not enough to keep the TDFs well fed. As a result the supply of currency, which had been trending upward at roughly 5% per year for many years, suddenly stopped growing. And this was associated with a sharp slowdown in the rate of growth of M*V, where M is defined as the monetary base. In an accounting sense, the sharp slowdown in the growth in the base caused the recession.
However this is not my preferred way of thinking about the problem, as it may (wrongly) suggest that if only the Fed had kept the base growing at 5% per year we would have avoided a recession in late 2007. Maybe, but maybe not. It’s quite possible that 5% more base money would have led to 5% less velocity. I prefer to talk in terms of the Fed’s control of M*V, where a monetary policy failure occurs when the Fed allows M*V to grow too fast, or too slowly. But that’s not good enough for most people, they want the process explained using what Nick calls “concrete steppes.” They ask “What caused velocity to suddenly plunge in late 2007 and early 2008?” To those people I say, “Velocity did not plunge, indeed it rose slightly.” If you want “concrete steppes” using the old M*V=P*Y workhorse, then I’d say the recession was caused by sudden stop in base growth, and a failure of V to rise enough to offset this “monetary tightening.”
There are no policy implications from the fact that during normal times most of the growing demand for base money comes from TDFs squirrelling away lots of $100 bills. If the Fed runs a sensible monetary policy, they will smoothly accommodate those shifts in base demand. But that will only occur when and if the Fed replaces the faulty “fed funds target bell” with a much more reliable “NGDPLT futures bell.”
PS. Astute readers like Nick, Bill and Saturos will notice that I’ve avoided the question of whether the MOE role of money is central to the entire process. They might argue that only the smaller two dragons can actually damage the village. At worst, an underfed big dragon takes food away from the smaller two dragons. So the big dragon can be a problem, but only indirectly. I think the big dragon can directly damage the village.
Tags:
31. December 2012 at 10:51
Scott, I take your last paragraph to imply that in your model, a failure to accommodate base demand means the supply of base relative to the demand for base at a steady 5% NGDP level path. Thus, the MoA role is simply implicit in the M*V convention.
BTW, the currency component of M1 is growing at about 10% p.a. after a dip in 2010. If this and 4-5% ngdp is enough to keep unemployment falling 75 bps per annum, 4-5% NGDP may be the Feds new implicit target, consistent with the Fed’s forecast.
31. December 2012 at 11:08
Happy New Year to Prof. Scott Sumner and the readers of this blog! 2012 was a good year for MMT. It is amazing to think that one guy’s blog can nudge the monetary policy for all the major central banks and thus improve the lives of nearly every human. Wow! Here’s to 2013 being the tipping point year for NGDPLT and (un)common sense!!
31. December 2012 at 11:33
Where do gamblers fit into the trichotomy?
31. December 2012 at 12:10
I have read enough fantasy to know that the cost of feeding dragons is excessive. They usually want to be fed young maidens, which are often hard to come by. It is a particularly raw deal if you are the maiden (or she is your daughter or sister) that has been selected as dragon fodder.
Inevitably, the kingdom will realise that the cost is no longer acceptible and find a knight to slay the dragon. The kingdom will probably lose a vilage or two in the conflict.
31. December 2012 at 12:28
You use the wrong data & the wrong calculations to arrive at wrong conclusions.
31. December 2012 at 12:46
Unfortunately, I was not in a position to comment on your original post on this subject, Scott, but I am afraid that it was based on a false premise (which I am pretty sure I have discussed with you before).
By 2007, awareness of the Fed’s long-term abuse of the dollar, and apparent willingness to be even more slack in response to America’s gathering problems was becoming widely appreciated outside professional finance. Meanwhile, the euro had become familiar, offered a €500 note, and the ECB was at least talking tough. If you got out more, you might recall that mid-2007 was when that view went viral, as evinced by media stories about Giselle Bunchen requiring payment in dollars http://news.bbc.co.uk/1/hi/business/7078612.stm and JayZ’s latest music video portraying criminals with euro notes http://news.bbc.co.uk/1/hi/7097736.stm. The levelling off in $100 circulation was because the euro was finally beginning to supplant the dollar in the TDF community, and had nothing to do with the financial crisis. Indeed the financial crisis probably reversed it as the very existence of the euro began to questioned.
Nevertheless, the financial crisis was already unfolding by then, with banks beginning to worry about the creditworthiness of their financial industry counterparties. If the Fed made a mistake at this time, it was probably a failure to introduce the IoR fast enough. Essentially, the banks wanted to use the Fed as a central counterparty, but receiving zero interest on the deposit side of that arrangement was prohibitive, so the money market was gradually freezing.
The misleading analysis in your original post illustrates the arrogant foolishness of your view that only “elite” monetary economists, paid as much as it takes to recruit them, should make monetary policy. You need some people who follow economic developments very closely (whatever you think of what he did with the information, Greenspan was always a master of this), and some who really know the monetary plumbing (private sector bankers, maybe?).
Nevertheless, let me take this opportunity to wish you (Scott) and all moneyillusion readers a happy 2013.
31. December 2012 at 13:28
Scott, each villager can also only carry $10,000, or else get thrown into a dungeon.
Do legal restrictions on using legal tender affect the ability to supply, and demand for, $100 bills?
I also thought that the connection between gold stocks and the base was severed in 1933.
31. December 2012 at 14:48
Frances Woolley.
31. December 2012 at 15:08
Tommy, Obviously I think the MoA role is key, but others don’t agree.
Rebeleconomist, You said;
“If you got out more,”
Your comment has nothing to do with getting out more, it’s information that you and I both learned from reading the newspapers. (Yes, I’ve also read the same anecdotes, which are obviously not representative.) I’d add that your anecdotes have no bearing on anything I said in this post. And the stuff about the financial crisis already being well advanced also has no bearing on anything I said. Nor do the comments on Greenspan. It’s easy to ridcule things that you don’t understand. I’d suggest taking a little bit of time trying to actually learn monetary economics. You still might disagree with me, but at least you’d be able to address my specific arguments, rather than simply making childish ad hominem attacks.
jknarr, No, the gold stock continued to drive growth in the base after 1933, in some ways even more directly than during the so-called “gold standard” prior to 1933. I’d guess legal restrictions have some effect on demand for cash, and hence quantity supplied.
Thanks Philo. I actually spelled it correctly, then my computer crashed. The second time I spaced out and spelled it incorrectly. I’m getting senile.
31. December 2012 at 15:12
On the web site of the Atlanta Fed …
http://macroblog.typepad.com/macroblog/2012/12/nominal-gdp-targeting-still-a-skeptic.html
… Dave Altig still does not believe.
“To summarize my concerns, the Achilles’ heel of nominal GDP targeting is that it provides a poor nominal anchor in an environment in which there is great uncertainty about the path of potential real GDP. As I noted in my earlier post, there is historical justification for that concern.”
He does not realize that resistance will prove futile in the Happy New Year!
31. December 2012 at 16:00
ssumner:
My point was that a recession occurred because the Fed did not meet the rising demand for base money in late 2007 and early 2008.
and
If you want “concrete steppes” using the old M*V=P*Y workhorse, then I’d say the recession was caused by sudden stop in base growth, and a failure of V to rise enough to offset this “monetary tightening.”
The fall in base growth wasn’t exactly “sudden”:
http://research.stlouisfed.org/fredgraph.png?g=e9V
It was gradually falling since the 2001 recession (while M2/M3 were increasing on the basis of credit expansion, which generated malinvestment).
The argument that the recession was caused by the Fed not increasing what it has direct control over, i.e. base growth, at a fast enough rate, is perfectly in line with Austrian theory. Applied Austrian theory holds that the early to mid 2000s, the base growth was probably greater than what the market process would have generated, and the malinvestment of this inflation was revealed once the Fed slowed down the rate of base money growth. The investments that depended on a continuously loosening base money, could no longer be sustained.
Connecting this back to MM, the tacit assumption in MM is that “elite” intellectuals can know what the optimal rate of money production for the market is, yet divorced from the market process itself. Austrians know that there is actually no way for anyone, including “elite” central bankers, to know the optimal quantity of money and rate of money production. For this knowledge is grounded on the information that only a private, competitive, market-driven money production can reveal to us.
Just like Bernanke could not possibly know what the “correct” rate of car production should be that is more informed that what can be revealed through the market process, so too does he have no idea what the “correct” rate of money production should be that is more informed than what can be revealed through the market process.
MMs, just like every single monetarist before them, believe that their particular inflation “rule” is the best the central bank can offer. But they are deluding themselves. The problem isn’t the rule, it’s the central bank itself.
Constant gradual aggregate spending that is non-market driven does not cause stable monetary systems and stable markets.
31. December 2012 at 16:14
There is a fairly close substitute for $100 bills, namely $50 bills. You might mention that the Fed didn’t supply enough of *them*, either.
“I’m getting senile.” No, blame the computer!
31. December 2012 at 17:16
Scott,
If you want to convince more people then you have to more clearly show how Lehman’s collapse was the result of falling NGDP expectations and and not mispriced assets. Your arguments on that point are more theoretical than empirical, and until you can find some smoking gun in the data non Market Monetarists will not be convinced.
31. December 2012 at 18:08
I’m confused.
You say “My point was that a recession occurred because the Fed did not meet the rising demand for base money in late 2007 and early 2008”
and then later “I’d say the recession was caused by sudden stop in base growth, and a failure of V to rise enough to offset this “monetary tightening.””
The first statement makes sense to me – people wanted to increase their cash balances for precautionary reasons so they reduced spending and without CB actions to increase the money supply to match the demand we had a recession.
The second statement appears to contradict the first, and implies that people were expecting rising income but when it failed to materialize they cut back on spending and caused a recession. Is that actually what you are saying happened ?
31. December 2012 at 18:45
Prof. Sumner,
Could you at least explain why emerging markets such as China tend to have much higher rates of inflation than relatively wealthy countries?
31. December 2012 at 19:40
The question of cash in circulation is fascinating. Would a “tax and crime amnesty” to anyone repatriating $100 bills in lumps of $1 million or more help the USA economy?
I suspect these TDF’ers like to spend their loot.
BTW, do food stamps increase the money supply?
What if NYC went back to subway tokens, and they became widely traded as cash? What if NYC tokens were minted up by several hundred billion worth, and people accepted them as cash?
And is it time for a $1000 bill? The $100 has been the king bill for a long time, but $100 ain’t what it used to be….
31. December 2012 at 20:28
@sumner
I find these anecdotes where you replace reasoned analysis with a frankly dubious metaphor slightly trying. More equations, less dubious similes!!!! Then again, I doubt I am representative of your typical reader :). You remain my favourite blogger. 🙂
I do no really understand how people can doubt your account of this crisis. The logic is compelling, the data doubly so. Those who doubt scott’s account should reflect that he is basically arguing that the fall in NGDP caused a (massive) deflationary shock. If he is right it should show up in the inflation statistics…and it does!!! Given that sticky wages make it impossible for industries not currently blessed with strong productivity growth to lower wages, Scott’s story predicts that inflation by sub-sector should have a median inflation close to zero, regardless of the CPI. Any remotely competent analysis confirms this.
Scott is right. If you don’t believe this, I point you to Daniel Kahmenn’s book, “thinking fast and slow” to explain all the ways in which you are deeply irrational.
@serious thinkers:
There is no such thing as a monetary policy that is optimal in any and all circumstances. NGDP, for example, is basically the policy of stabilising aggregate demand. This fails in the following two cases:
1) RGDP growth exceeds the NGDP growth target. In this case, the central bank, by targeting NGDP, induces an unnecessary recession due to deflationary pressure.
2) There is a sound reason why aggregate demand should be falling. E.g. If an asteroid wiped out half of the the population of the US, it would be foolish to stabilise demand, it should fall by half, obviously. :). In this case NGDP targeting would unfairly favour workers over pensioners, for example.
Probably there are more that I didn’t think of. Nevertheless, NGDP targeting is the right policy now, as we are in a demand side recession, which a demand stabilisation policy would (obviously) have made impossible.
=========================
Other points: Rebel, wtf? I literally have no idea how to interpret your comments.
MF: The austrian sector have contributed not a single useful idea to economic thought. They are useful in the way that a good devil’s advocate can help you think through your ideas more fully, but the stories told by Hayek have no bearing on reality. They are elgant, and somewhat compelling, but data confirms their wrongness. A good thinker discards two favourite hypothesis before breakfast: discard austrianism and join the ranks of the open minded. And I say this as a devout catholic……
Scott, you do need to build an empirical case on this blog for your argument. There are plenty of data minded individuals out of them, and no matter how nice your stories, without data they (we) will not be convinced. There must be data that confirms beyond doubt that the US is (was?!?) in a demand side recession that was crushing industry. Find it, and publish it. I had to personally check out the UK inflation data to be convinced by your story. I doubt I am alone. Get Bill Mcbride onside and you will have won over 50% of young finance professionals – get it done!!!!
Happy new year.
31. December 2012 at 20:47
Suppose that there are two instruments – Benjamins and Freddies. Freddies are interest bearing, and thus are subject to rate risk and have a lower opportunity cost, but are liquid and have no counterparty or default risk. Investors want to hold a mix of these two. In particular, Freddies are a necessary ingredient in the manufacture of synthetic securities to suit required risk and return profiles, a very profitable business if conducted on a large scale.
Now suppose that some of the Freddies are suspected of having counterparty and default risk. The alternatives are accepting more risk, doing less business or substituting Benjamins for Freddies. None of these is very satisfactory. Transaction volumes fall; institutions experience liquidity problems; demand for those Freddies that remain unblemished drives their rates to 0. Growth in the broader aggregates like M4 goes into reverse, even as the Fed pumps in more base money.
The trend is accelerated by mark to market rules, which turn unrealized nominal losses into actual losses, destroying capital and forcing deleveraging and dumping of discredited assets at fire sale prices. The capital requirements and mark to market rules which were supposed to be improvements prove to have a major downside – aggravation of systemic risk.
1. January 2013 at 05:54
My view (and I think that of Rowe and Saturus) is that the problem was a shortage of deposit-type media of exchange, not 100 dollar bills, currency used to make ordinary purchases, or even the reserve balances banks use to clear checks.
When the Fed makes open market purchases, it directly creates deposit balances for those selling the securities as well as additional reserve balances for their banks.
The Fed chose to restrict the creation of deposit-type media of exchange to keep spending in the economy from growing too fast. (Though they framed this as raising their policy interest rate to keep inflation from rising.)
They could have done the same if there was no such thing as hand-to-hand currency. They could even do the same thing if the demand for reserve balances was zero in equilibrium.
Suppose the Fed is giving melons to the villagers. The villagers eat the melons and throw the rinds in the trash pile. The dragon tenders prick up the rinds and give them to the dragons. There are only two dragons.
Peasants from the countryside come to the village and eat some melons. They take home their rinds and fertilize their fields.
Most food is actually produced by the peasants in the countryside and most don’t use melon rinds for fertilizer.
The Fed decides that the peasants are overweight and send less melons.
The peasants in the village eat their fill like usual, but their are fewer melons left for the peasants from the countryside. They are starving.
While the peasants in the village remain plump, when averaged with those from the countryside, average weights are coming down.
At the same time, rumours spread that that a blight is speading in the countryside. While the blight does reduce the output of food, worse, it poisons the crop in a way that isn’t obvious. You eat the food, and you may get very sick and perhaps die. Not every melon from a field is bad, just some.
Food production is dropping in the countryside, but demand is dropping more than supply.
Starving peasants are coming to the village looking for melons, the Fed usually provides when the blight strikes.
The Fed meets. They decide to check on weights. Well, it so happens that a group of very heavy American tourists are visiting the village that day, so average weights shoot up through the roof.
The Fed refuses to send more melons.
Now, there is panic in the countryside. People rush to the village to weight in line for the limited quantity of melons. As they sit in their refugee camps on the outskirts, they aren’t growing any food!
Finally, the Fed gets the message and sends in huge amounts of melons. There is a mountain of rinds outside the village. The dragon keepers have no problem.
Looking back on the situation, Sumner notes that records showed that the trash pile of the rinds was low for a time. (This was when the Fed decided people needed to lose weight.)
Sumner says that there must have been dragon attacks. Obviously, because there were too few rinds.
Everyone else says that they saw no evidence of dragon attacks, but rather the problem was starvation in the countryside.
Sumner says that this is impossible, because the peasants in the countryside can always grow there own food.
Most economists instead blame the rumours of the blight. Or maybe they blame the blight. There is a big debate as to whether the rumour was even true. They don’t even believe that the dragons exist.
I would blame the deaths on the starvation and the starvation on the failure of the Fed to deliver enough melons. I would dismiss worries about the dragons. If the dragons were a problem, they never in fact “lash out” at anyone. It is rather that eat food that otherwise would go to peasants in the village and the countryside.
In fact, villagers and peasants usually have stocks of food and don’t depend on regular melon deliveries from the Fed.
But rumors of a blight? That leads to disaster. And when the Fed gets confused by a pack of American tourists passing through, all confidence is lost.
Fortunately, the refugee camps only include a minority of the peasants from the countryside. Today, the dragons are fed, the peasants in the village are doing great and have huge piles of melons in front of their house (they even keep them in Fed provided refrigerators!) Most peasants in the country side are going food on land that seems blight free and have ample food stocks.
It is only those hungry peasants remaining the refugee camps that are the problem.
Oh, the Fed is still worried about the weight problem. And they would act if peasants begain streaming towards the refugee camps or starvation caused weights to fall.
But they are just a little uneasy now about the refugee camp problem. Maybe the “fiscal authority” should do something about that.
1. January 2013 at 06:08
Philo:
Why is deflation are problem?
If productivity grows rapidly, and potential output grows faster than the growth rate of the target path, then there is deflation.
So?
If we consdier the microeconomics, millions of firms find that they can produce more output with the same resources. Their average and marginal costs are lower. Their profit maximizing level of output is higher, which they sell by lowering their prices. They can do this because their marginal and average costs are lower.
This is nothing like a situation where demand falls and so they can’t sell as much at current prices. While it may be more profitable (or less loss indusing) to lower prices a bit and dampen the decrease in the amount they sell, the entire situation is painful.
1. January 2013 at 06:10
constantly diminishing k factor in the real economy (rising speculation far from direct spending in production/consumption + Exogenous demand for US/European money and quasi-money -> Triffin dillema), which leads to a (delta) money and quasi-money GDP multiplier smaller than one…
1. January 2013 at 08:01
Jim, It almost seems like Altig doesn’t view NGDP as a nominal variable. How is it not a nominal anchor?
JoeMac, You said;
“If you want to convince more people then you have to more clearly show how Lehman’s collapse was the result of falling NGDP expectations and and not mispriced assets. Your arguments on that point are more theoretical than empirical, and until you can find some smoking gun in the data non Market Monetarists will not be convinced.”
What we have here is a failure to communicate. Let’s review:
1. Every thinking person should know that the NGDP expectations crash lowered asset values and made Lehman’s situation worse. Surely there is no dispute over that point.
2. Nobody would claim that the NGDP crash was Lehman’s only problem
So tell me again what I need to convince people of, and please be more precise.
Ron, You misunderstood my argument. Demand for cash generally rises at about 5% per year. There were no unusual movements in cash demand in late 2007 and early 2008. No sudden “precautionary demand.” What was unusual was the sudden sharp drop in the growth rate of the supply of base money. That caused the slowdown in NGDP growth, at least in an accounting sense. There’s no mysterious cash demand changes that require explanation.
Travis, I answered your question in a previous commen thread.
Phil. I agree about the asteroid strike, which is why I actually favor targetng NGDP per capita, but often leave off the “per capita.”
I disagree with your point 1. The example whare RGDP changes dramatically for supply side reasons is precisely the case where NGDP targeting is far superior to inflation targeting. You can’t get a recession if RGDP is rising faster than the NGDP target.
PeterN, The substitute for Freddies would normally be T-bills. Only when rates fall to zero do $100 bills become a good substitute.
Bill, You said;
“When the Fed makes open market purchases, it directly creates deposit balances for those selling the securities as well as additional reserve balances for their banks.”
That’s only the short term effect. Within a few days the new base money filters out into currency held by the public.
I don’t understand the melon example–how does that address my claim that the biggest part of base demand is $100 bills held by TDFs?
1. January 2013 at 08:04
Great post. It’s been obvious to me that this is the way you think about the economy for about a year now, but many commenters and visitors seem not to have picked it up yet. Hope it doesn’t pain you too much to repeat yourself so many times. (I still think you need to update your “key blog posts” link.)
“The interest rate signalling mechanism only works well if the Wicksellian equilibrium interest rate is constant.”
Funny idea – didn’t Wicksell define that as “that rate which if not maintained will result in continual inflation or deflation”? So this is just an admission that interest rates in and of themselves tell you nothing.
“I think the big dragon can directly damage the village.”
Action at a distance, eh? Of course the big dragon hardly ever even gets near the village. If the other two dropped dead, the poor nominally-rigid villagers would either prevent the value of the currency dragon from rising (tautologically implied by nominal fixedness); or the currency would appreciate partway, then get stuck partway below equilibrium (again, due to logical implication of stickiness) instead of rising the rest of the way to equilibrium as RGDP falls (MoA on its own has no impact on RGDP); or the currency appreciates all the way to equilibrium with no unemployment, prices fully adjusting.
Bill, I’m afraid you’ve totally lost me. But I suspect I’d agree with you if I knew what you were talking about.
In other news, this is scary: http://www.bloomberg.com/news/2012-12-24/plan-to-limit-fed-s-mandate-is-folly.html
1. January 2013 at 09:29
Banks can hold no reserves precisely because the Fed doesn’t allow shifts in currency demand to disturb the money market.
In other words, the Fed doesn’t let drug dealers set the Fed Funds rate.
None of this has anything to do with the “equilibrium” Fed Funds rate. There’s no reason to think that currency users have any special power over that, right?
1. January 2013 at 10:52
“If you got out more,” I don’t literally mean that, Scott. It is just a common British expression, teasing you for apparently not following celebrities like Gisele Bundchen and JayZ. Sometimes these things cross the Atlantic, and sometimes not. Sorry.
I note, however, that you have not engaged either substantive point I made, either that about the ready alternative explanation for the slowdown in US currency growth around 2007 or that about the need for a range of types to formulate monetary policy.
1. January 2013 at 11:53
“The substitute for Freddies would normally be T-bills. Only when rates fall to zero do $100 bills become a good substitute.”
Consider T-bills a form of Freddie, but not available in all the durations and interest rates needed for something as complicated as the fixed income securities market. Longer duration sovereign bonds, corporate AAAs, GSE guaranteed securities and derivatives specially constructed/insured to have AAA ratings are all needed for a multi-trillion dollar market. Think, for instance, what money market funds need to do to pay dividends and still make a profit.
The shortage of Freddies will force the rates of risk free instruments to 0, where they can, indeed substitute for Benjamins (at a small but significant sacrifice). In other cases where Freddies would be used as collateral or a store of value, Benjamins will be used for lack of good alternatives or for protection from currency risk (attack on sovereign bonds will affect exchange rates eg. the Euro).
It’s perfectly possible for tight money to start with the broadest forms of money. In this case you would expect to see changes in the larger aggregates lead those in smaller, which is what you did indeed see. Look at Divisia M4.
1. January 2013 at 15:40
Monetary policy was too tight – but Benjamin’s weren’t in short supply. The cyclicality in money flows (MVt) is unvarying. If you actually review the data, the roc in currency had nothing to do with the recession / depression.
And currency doesn’t change hands as fast as transaction accounts are debited. What the economic fraternity calls a base is meaningless. Currency has no expansion coefficient. Your thery is incorrect. From the standpoint of the banking system, (fractional reserve banking) is function of the velocity of bank deposits, not a function of their volume.
And one cannot use absolutes when comparing metrics. Only rates-of-change (roc’s) are telling
1. January 2013 at 17:13
NO CORRELATION BETWEEN SHORTAGE OF BENJAMIN’S & 2008.
DATE……………CURRENCY…………ROC’s
1/1/2000 ,,,,,,, 1126.9 ,,,,,,, 0.01
2/1/2000 ,,,,,,, 1097.4 ,,,,,,, 0.00
3/1/2000 ,,,,,,, 1108.9 ,,,,,,, 0.01
4/1/2000 ,,,,,,, 1125.7 ,,,,,,, 0.03
5/1/2000 ,,,,,,, 1100.4 ,,,,,,, 0.01
6/1/2000 ,,,,,,, 1101.8 ,,,,,,, 0.01
7/1/2000 ,,,,,,, 1102.8 ,,,,,,, 0.01
8/1/2000 ,,,,,,, 1094.5 ,,,,,,, -0.02
9/1/2000 ,,,,,,, 1088.8 ,,,,,,, -0.05
10/1/2000 ,,,,,,, 1092.2 ,,,,,,, -0.03
11/1/2000 ,,,,,,, 1092.1 ,,,,,,, 0.00
12/1/2000 ,,,,,,, 1111.7 ,,,,,,, 0.00
1/1/2001 ,,,,,,, 1100.4 ,,,,,,, -0.02
2/1/2001 ,,,,,,, 1089.8 ,,,,,,, -0.01
3/1/2001 ,,,,,,, 1111.2 ,,,,,,, 0.01
4/1/2001 ,,,,,,, 1126.7 ,,,,,,, 0.02
5/1/2001 ,,,,,,, 1114.7 ,,,,,,, 0.02
6/1/2001 ,,,,,,, 1126.3 ,,,,,,, 0.03
7/1/2001 ,,,,,,, 1139.7 ,,,,,,, 0.04
8/1/2001 ,,,,,,, 1144.5 ,,,,,,, 0.05
9/1/2001 ,,,,,,, 1193.6 ,,,,,,, 0.07
10/1/2001 ,,,,,,, 1159.2 ,,,,,,, 0.05
11/1/2001 ,,,,,,, 1169 ,,,,,,, 0.07
12/1/2001 ,,,,,,, 1208.5 ,,,,,,, 0.09
1/1/2002 ,,,,,,, 1191.8 ,,,,,,, 0.06
2/1/2002 ,,,,,,, 1178.3 ,,,,,,, 0.06
3/1/2002 ,,,,,,, 1196.8 ,,,,,,, 0.06
4/1/2002 ,,,,,,, 1196.9 ,,,,,,, 0.05
5/1/2002 ,,,,,,, 1186 ,,,,,,, 0.04
6/1/2002 ,,,,,,, 1194.7 ,,,,,,, 0.00
7/1/2002 ,,,,,,, 1200.4 ,,,,,,, 0.04
8/1/2002 ,,,,,,, 1182.3 ,,,,,,, 0.01
9/1/2002 ,,,,,,, 1185.8 ,,,,,,, -0.02
10/1/2002 ,,,,,,, 1196.6 ,,,,,,, 0.00
11/1/2002 ,,,,,,, 1205.3 ,,,,,,, 0.02
12/1/2002 ,,,,,,, 1245.5 ,,,,,,, 0.04
1/1/2003 ,,,,,,, 1225.3 ,,,,,,, 0.02
2/1/2003 ,,,,,,, 1225.2 ,,,,,,, 0.03
3/1/2003 ,,,,,,, 1244.8 ,,,,,,, 0.04
4/1/2003 ,,,,,,, 1259.3 ,,,,,,, 0.05
5/1/2003 ,,,,,,, 1266.3 ,,,,,,, 0.07
6/1/2003 ,,,,,,, 1284.4 ,,,,,,, 0.08
7/1/2003 ,,,,,,, 1287.6 ,,,,,,, 0.08
8/1/2003 ,,,,,,, 1291.9 ,,,,,,, 0.07
9/1/2003 ,,,,,,, 1285.9 ,,,,,,, 0.03
10/1/2003 ,,,,,,, 1288.5 ,,,,,,, 0.05
11/1/2003 ,,,,,,, 1293.6 ,,,,,,, 0.06
12/1/2003 ,,,,,,, 1332.2 ,,,,,,, 0.07
1/1/2004 ,,,,,,, 1301.5 ,,,,,,, 0.03
2/1/2004 ,,,,,,, 1306.3 ,,,,,,, 0.03
3/1/2004 ,,,,,,, 1337.5 ,,,,,,, 0.04
4/1/2004 ,,,,,,, 1342.8 ,,,,,,, 0.04
5/1/2004 ,,,,,,, 1333.1 ,,,,,,, 0.03
6/1/2004 ,,,,,,, 1347.4 ,,,,,,, 0.05
7/1/2004 ,,,,,,, 1338.5 ,,,,,,, 0.04
8/1/2004 ,,,,,,, 1352.2 ,,,,,,, 0.05
9/1/2004 ,,,,,,, 1348.8 ,,,,,,, 0.01
10/1/2004 ,,,,,,, 1351 ,,,,,,, 0.04
11/1/2004 ,,,,,,, 1370.7 ,,,,,,, 0.05
12/1/2004 ,,,,,,, 1401.2 ,,,,,,, 0.05
1/1/2005 ,,,,,,, 1361.2 ,,,,,,, 0.01
2/1/2005 ,,,,,,, 1354.7 ,,,,,,, 0.02
3/1/2005 ,,,,,,, 1381.4 ,,,,,,, 0.03
4/1/2005 ,,,,,,, 1369 ,,,,,,, 0.02
5/1/2005 ,,,,,,, 1369.1 ,,,,,,, 0.01
6/1/2005 ,,,,,,, 1384 ,,,,,,, 0.03
7/1/2005 ,,,,,,, 1365 ,,,,,,, 0.01
8/1/2005 ,,,,,,, 1376.5 ,,,,,,, 0.00
9/1/2005 ,,,,,,, 1363.1 ,,,,,,, -0.03
10/1/2005 ,,,,,,, 1364.9 ,,,,,,, 0.00
11/1/2005 ,,,,,,, 1373 ,,,,,,, 0.01
12/1/2005 ,,,,,,, 1396.6 ,,,,,,, 0.01
1/1/2006 ,,,,,,, 1374.9 ,,,,,,, 0.00
2/1/2006 ,,,,,,, 1361.5 ,,,,,,, -0.01
3/1/2006 ,,,,,,, 1394.3 ,,,,,,, 0.01
4/1/2006 ,,,,,,, 1393.2 ,,,,,,, 0.02
5/1/2006 ,,,,,,, 1391.6 ,,,,,,, 0.01
6/1/2006 ,,,,,,, 1378.6 ,,,,,,, 0.01
7/1/2006 ,,,,,,, 1368.1 ,,,,,,, 0.00
8/1/2006 ,,,,,,, 1370.1 ,,,,,,, 0.00
9/1/2006 ,,,,,,, 1346.9 ,,,,,,, -0.04
10/1/2006 ,,,,,,, 1359.8 ,,,,,,, -0.01
11/1/2006 ,,,,,,, 1368 ,,,,,,, 0.00
12/1/2006 ,,,,,,, 1387.1 ,,,,,,, -0.01
1/1/2007 ,,,,,,, 1368.4 ,,,,,,, -0.02
2/1/2007 ,,,,,,, 1346.9 ,,,,,,, -0.03
3/1/2007 ,,,,,,, 1377.8 ,,,,,,, 0.00
4/1/2007 ,,,,,,, 1391.8 ,,,,,,, 0.02
5/1/2007 ,,,,,,, 1385 ,,,,,,, 0.01
6/1/2007 ,,,,,,, 1370.2 ,,,,,,, 0.02
7/1/2007 ,,,,,,, 1368.1 ,,,,,,, 0.01
8/1/2007 ,,,,,,, 1372.2 ,,,,,,, 0.00
9/1/2007 ,,,,,,, 1355.7 ,,,,,,, -0.02
10/1/2007 ,,,,,,, 1368.8 ,,,,,,, 0.00
11/1/2007 ,,,,,,, 1369.9 ,,,,,,, 0.02
12/1/2007 ,,,,,,, 1394.2 ,,,,,,, 0.01
1/1/2008 ,,,,,,, 1375.3 ,,,,,,, -0.01
2/1/2008 ,,,,,,, 1365.3 ,,,,,,, -0.01
3/1/2008 ,,,,,,, 1400.5 ,,,,,,, 0.02
4/1/2008 ,,,,,,, 1406 ,,,,,,, 0.03
5/1/2008 ,,,,,,, 1396 ,,,,,,, 0.02
6/1/2008 ,,,,,,, 1406.8 ,,,,,,, 0.04
7/1/2008 ,,,,,,, 1417.3 ,,,,,,, 0.04
8/1/2008 ,,,,,,, 1400.9 ,,,,,,, 0.02
9/1/2008 ,,,,,,, 1441.1 ,,,,,,, 0.03
10/1/2008 ,,,,,,, 1462.2 ,,,,,,, 0.06
11/1/2008 ,,,,,,, 1513.2 ,,,,,,, 0.11
12/1/2008 ,,,,,,, 1631.8 ,,,,,,, 0.17
1/1/2009 ,,,,,,, 1580.2 ,,,,,,, 0.12
2/1/2009 ,,,,,,, 1551.4 ,,,,,,, 0.11
3/1/2009 ,,,,,,, 1594.2 ,,,,,,, 0.13
4/1/2009 ,,,,,,, 1626.9 ,,,,,,, 0.15
5/1/2009 ,,,,,,, 1617.1 ,,,,,,, 0.15
6/1/2009 ,,,,,,, 1661 ,,,,,,, 0.15
7/1/2009 ,,,,,,, 1656.9 ,,,,,,, 0.13
8/1/2009 ,,,,,,, 1650.5 ,,,,,,, 0.09
9/1/2009 ,,,,,,, 1640.2 ,,,,,,, 0.01
10/1/2009 ,,,,,,, 1663.5 ,,,,,,, 0.05
11/1/2009 ,,,,,,, 1682.3 ,,,,,,, 0.08
12/1/2009 ,,,,,,, 1723.6 ,,,,,,, 0.08
1/1/2010 ,,,,,,, 1673.4 ,,,,,,, 0.03
2/1/2010 ,,,,,,, 1684.4 ,,,,,,, 0.04
3/1/2010 ,,,,,,, 1728.8 ,,,,,,, 0.04
4/1/2010 ,,,,,,, 1715.2 ,,,,,,, 0.04
5/1/2010 ,,,,,,, 1707.4 ,,,,,,, 0.03
6/1/2010 ,,,,,,, 1731.6 ,,,,,,, 0.06
7/1/2010 ,,,,,,, 1717.6 ,,,,,,, 0.03
8/1/2010 ,,,,,,, 1738.8 ,,,,,,, 0.03
9/1/2010 ,,,,,,, 1740.6 ,,,,,,, 0.01
10/1/2010 ,,,,,,, 1766.6 ,,,,,,, 0.06
11/1/2010 ,,,,,,, 1827.9 ,,,,,,, 0.09
12/1/2010 ,,,,,,, 1870.5 ,,,,,,, 0.08
1/1/2011 ,,,,,,, 1854.7 ,,,,,,, 0.08
2/1/2011 ,,,,,,, 1857.8 ,,,,,,, 0.09
3/1/2011 ,,,,,,, 1908.5 ,,,,,,, 0.10
4/1/2011 ,,,,,,, 1917 ,,,,,,, 0.12
5/1/2011 ,,,,,,, 1933.2 ,,,,,,, 0.11
6/1/2011 ,,,,,,, 1953.2 ,,,,,,, 0.12
7/1/2011 ,,,,,,, 1993.5 ,,,,,,, 0.13
8/1/2011 ,,,,,,, 2100.8 ,,,,,,, 0.15
9/1/2011 ,,,,,,, 2098.1 ,,,,,,, 0.12
10/1/2011 ,,,,,,, 2127.1 ,,,,,,, 0.15
11/1/2011 ,,,,,,, 2163.3 ,,,,,,, 0.16
12/1/2011 ,,,,,,, 2206.9 ,,,,,,, 0.16
1/1/2012 ,,,,,,, 2206.5 ,,,,,,, 0.15
2/1/2012 ,,,,,,, 2191.9 ,,,,,,, 0.13
3/1/2012 ,,,,,,, 2243.5 ,,,,,,, 0.15
4/1/2012 ,,,,,,, 2265.9 ,,,,,,, 0.14
5/1/2012 ,,,,,,, 2245.6 ,,,,,,, 0.07
6/1/2012 ,,,,,,, 2263.4 ,,,,,,, 0.08
7/1/2012 ,,,,,,, 2304.4 ,,,,,,, 0.08
8/1/2012 ,,,,,,, 2327.4 ,,,,,,, 0.08
9/1/2012 ,,,,,,, 2352.6 ,,,,,,, 0.07
10/1/2012 ,,,,,,, 2409.9 ,,,,,,, 0.09
11/1/2012 ,,,,,,, 2408.7 ,,,,,,, 0.10
1. January 2013 at 18:47
People say that cash “burns a hole in their pocket”. But just as there are saved DDs (esp. prior to the elimination of Reg. Q ceilings), cash too is saved “for a rainy day” (& cash could represent a duplicative operation).
The Fed was too tight – but this is not reflected in the BOG’s statistical releases. And the “trading desk’s” liquidity injections have historically displayed a delayed effect on velocity (Vt) See G.6 statistical release (debits & deposit turnover).
Roc’s for MVt are ex-ante (because of fixed lags) – they are not coincident indicators. If you search for the Fed’s mistake using absolutes – you won’t find it.
Vt fell (along with income velocity Vi). Vt didn’t rise (because of dis-intermediation within the non-banks).
Vt’s never been correctly calculated. An outflow of savings from non-bank customer accounts (e.g., at shadow banks), or savings flowing into the CBs (where S doesn’t = I), should be subtracted, while savings flowing through the non-banks (where savings are matched with investment) should be included (like interbank demand deposits are excluded in tabulations of the money stock).
It’s no happenstance: “no one at the Fed tracks it” (v.p. Fed).
Pritchard, Ph.D., economics, Chicago 1933: “the correlation of the time series is remarkable”
2. January 2013 at 03:49
“Banks can hold no reserves precisely because the Fed doesn’t allow shifts in currency demand to disturb the money market.”
P.S.: I took some time to understand Mile’s exchange rate proposal. Only got the whole idea after reading Buiter in the FT blog. Basically, there would be different units of accounts, one for bank/electronic money and other for paper currency.
The exchange rate would not be fixed, it would correspond to the accrued negative interest on bank/electronic money at the time. For instance, if the negative interest rate of 5% year was set in january 2012, in December 2012 a deposit in paper currency of 105 should correspond to 100 in bank money and a withdrawal of 100 of bank money should correspond to 105 of paper money. On the other hand, in june 2012, the exchange rate of paper currency against bank money would be 102.5/1. E
Interesting, but complicated to deal with in everyday life.
2. January 2013 at 04:43
I think the problem was shortages of deposit-type liabilities and not shortages of currency.
These shortages were directly created by the Fed.
If the demand for currency had stopped growing or had began to shrink rapidly, the exact same shortage of deposits would have occured.
Shortages of deposits cause deceases in spending on output.
That is because deposits are used as media of exchange.
The Fed directly changes the quantity of deposits through open market operations.
Banks do not shovel currency out on the sidewalk. There is no process by which money created by the Fed filters out as currency within a few days.
There is a process by which the Fed creates base money to meet currency demands.
If the Fed is using open market operations to impact the policy rate, the output gap or inflation, then this is not what occurs. It is the impact of the open market operations on deposits that is relevant.
If there were no currency, it would work the same way.
2. January 2013 at 05:50
Scott, in a sense I agree with Bill, except that I think you were trying to make a different point. I read your post as a criticism of the concrete steps approach, and nothing more. When you say “in an accounting sense” I read you to mean, “without any relationship to proximate cause and effect.” It is possible without getting into the nuances of deposits versus currency that the large TDF segment was actually decreasing its currency demand even faster than trend was reversing, while the non-TDF’s (though initially small) were rapidly increase their demand. The failure to provide $100s for TDF’s then is only a cause in sense that even though the TDF’s were sated, more $100s would have overfilled them and they would have pushed more out to the hungry settlement holders (which have a very strong relationship to deposit demand, unlike TDFs).
It’s also true in an accounting sense that the Fed didn’t provide enough currency to me, personally. It doesn’t matter how big I am as a percent of currency holders, just like it doesn’t matter for the TDFs. Enough currency to me would have fixed the problem in an accounting sense, and we learn nothing about where the delta in marginal demand versus supply came from during the period.
2. January 2013 at 07:01
Bill and dlr are both right, I think.
2. January 2013 at 07:22
@scot:
Re point (1). If you have potential RGDP growth which is greater than NGDP growth, then in your framework, by definition, the central bank would be restricting aggregate demand to less than was sufficient to buy all of the economies potential output – you would be creating an output gap on purpose. I suppose this might be beneficial, but this is basically what Japan has been doing for twenty years. The NGDP target should have space to have trend RGDP growth plus a little room for inflation to make wage adjustments in uncompetitive industries easier.
It just seems that point one is a consequence of equating NGDP with aggregate demand (which seems right to me).
2. January 2013 at 08:49
Saturos, You said;
“Funny idea – didn’t Wicksell define that as “that rate which if not maintained will result in continual inflation or deflation”? So this is just an admission that interest rates in and of themselves tell you nothing.”
Yup. And yup, in a world of ratex you have “action at a distance.”
flow5, At least you should use seasonally adjusted data.
Rebeleconomist, If your “alternative explanation” was valid, there would have been no slowdown in NGDP growth. You are attributing it to a drop in demand for currency. Peter Temin made the same mistake in his Depression book back in 1976.
Peter N, I certainly agree that changes in other financial asset markets can impact base demand. But just as with more base demand from drug dealers, I expect the Fed to accommodate those shifts so that we don’t get falling NGDP.
Bill, You said;
“Banks do not shovel currency out on the sidewalk. There is no process by which money created by the Fed filters out as currency within a few days.”
Sure there is—it’s call falling interest rates, which is the opportunity cost of holding cash. If the Fed injects more base money than the public wants to hold, interest rates fall until the public wants to hold that much currency. Only when rates fall to zero do the banks hold lots of ERs.
If the Fed was truly accommodating the public’s demand for currency in a passive way, there would have been no slowdown in NGDP growth in late 2007 and early 2008. But there was, hence the currency stock was too small.
dlr, If I understand you correctly, I agree that we don’t know where the demand was coming from at that time–although you can get a pretty good estimate by looking at changes in $100s and $50s versus changes in $5s $10s and $20s.
Phil, I strongly disagree, as I’m sure all the other MMs do. You’d simply have “good deflation”. Austrians would also disagree. Japan had bad deflation because even NGDP is falling.
2. January 2013 at 09:32
TDF is a slick acronymn for the black market, but it reflects more about hoarding (& a distrust in banks) – not spending. The proof is in the math. There’s no correlation between currency & gDp. With the exception of the Great Depression, the growth of bank deposits is sync’d up with gDp growth.
Phil: Policy makers have the flexibility to upwardly adjust any n-gDp target. The idea is not to suppress real-output, but to prevent the greater damage that occurs from a decline in incomes during recessions & or the loss & redistribution of income during excessive inflation. Even though there is limited upward & downward price flexibility, roc’s in inflation can be forced to “revert to mean”. My guess is that real-output on the other hand is even less flexible than prices.
2. January 2013 at 11:18
Bill,
I liked your village/countryside scenario and have pondered versions of it many times. A couple of thoughts:
In years past, peasants from the countryside would come to the villages to find work in their prime, and then return to points of origin to retire. However, opportunity costs for doing so now are higher and less certain. Two means for those in the countryside to compensate: disability payments mean not having to leave the countryside and its lower living expenses. For those with good health but inadequate local work, selling off assets to the cityfolk helps, as they pay more than locals would for the privilege of living with lower tax rates and relative housing costs.
The danger of refugee camps that could eventually include country folk is nonetheless real, as this group of “solutions” exists in delicate equilibrium, especially in terms of what government can provide. For now the “refugee camp” is mostly a limited group of homeless and unemployed of village or near-village origins.
2. January 2013 at 13:32
“Peter N, I certainly agree that changes in other financial asset markets can impact base demand. But just as with more base demand from drug dealers, I expect the Fed to accommodate those shifts so that we don’t get falling NGDP.”
The shortfall in broad money is in the trillions. Something like $6 trillion of formerly 0 risk assets usable for low information financial transactions has been revulsed from that status. That’s a lot to compensate for with the injection of a much smaller amount of base money.
It may well be that the QE approach of exchanging good collateral for revulsed collateral is more effective. Perhaps it should have been started sooner, before there was such a large loss of confidence.
4. January 2013 at 11:42
“At least you should use seasonally adjusted data”
I don’t “fudge” the numbers.
9. January 2013 at 17:31
[…] other worthy entries in the same instant genre of Middle-Earth macro metaphors are Scott Sumner’s Feed the dragon (but not too much), Morton’s Central Banking in Middle-Earth, or: The Much-Maligned King Thror and Eric Crampton’s […]