Sorry this is late—there were a lot of comments and I amtraveling. Before we begin, a note and plea: we are getting an increasingnumber of emails with comments and questions (sent to NEP email and to mine).Please understand I cannot respond to those—I get hundreds of emails a day andit would consume all of my time to respond individually. That is why I amcollecting comments on the Primer page and responding to all at once. I knowsome people are having trouble posting comments (I do too!) but what I’ve foundis that “three’s a charm”: if you hit “post” three times it inevitably works.Sorry about that.
Some of the comments were quite long and dealt with severalissues. So this week I am posting most of the text, followed by my response.
Iris: So from my point of view, what you obviously neglectis to insert all the additional references to presupposed material, which us,especially with less “previous knowledge of MMT” could help to betterunderstand the matter you’re dealing with. My impression is for example, thatnot only general accounting basics are necessary but also knowledge about thestructural system of inter-firm-, inter-bank- and governmentbanking-institutions’ accounting to private sector. Would it be asked
to much of a favor for you to perhaps offer, at least via footnotes, thesehints too?
to much of a favor for you to perhaps offer, at least via footnotes, thesehints too?
A: Well there is always a trade-off. We don’t want to getinto defining the meaning of “the” (as a former President tried to do). I dopresume some knowledge and it is tricky just how much should be assumed and howmuch explained in detail. Sometimes the reader with less background is probablygoing to have to accept some statements without fully understanding all thedetails behind them. I have been leaving out detailed accounting for tworeasons: it is more detail than most will want, and it is hard to produce thesein Word (and I have no idea how hard it is for our techie to post them to theblog). However, since there have been a number of requests, I will devote ablog to “T Accounts”.
Hugo: Ok, “excess deposits” results in increased”demand for profitable savings vehicles”. And if demand for savingsvehicles exceeds supply, the market will adjust. Savers will have to acceptlower yields on their savings. Firms would find it more easy to borrow money.Interest rates for corporate bonds would likely decrease (in thefinancial markets?). But I feel the transmission to an increase in governmentbonds is somewhat weak here? What you are suggesting seems to be: “Excessdeposits seeking profitable savings vehicles” -> “excessreserves in the interbank lending system” -> “overnight ratemaintenance” -> “bond sales” -> “equilibrium”..?But how does “excess deposits” necessarily lead to “excessreserves” here?
A: Before I get into a response, Guest responded to Hugo,trying to straighten this out:
-Guest:It is a result of double-entry book keeping. Whenever government creditsdeposits of someone in the banking system, it alse credits banks reserves tocreate asset that offsets banks deposit liability.
A: OkI am not sure what an excess deposit would be. When I get my paycheck mydeposit goes up and surely I do not think it is excessive. I then buyconsumption goods and services. I might also make some portfolio decision overwhat is left, allocating some of my accumulated savings to higher earningfinancial assets. Hugo says I might bid up bond prices, on both private andgovernment debt, lowering interest rates on the outstanding stock. Right. He isnot convinced that a government deficit will put downward pressure ongovernment bonds, however, because he does not see how my “excess deposits”creates “excess reserves”. Remember that reserves are on the asset side of thebank’s balance sheet while deposits are on the liability side. When governmentmakes a payment, both sides go up—the bank’s reserves at the Fed are credited,and my demand deposit is credited. Most of those additional reserves will beexcess reserves (details on this are complicated as reserve requirements arecalculated after a lag—let us ignore those details for now). Banks make aportfolio decision: buy something that earns higher interest rate. First theycan lend in the overnight, fed funds, market, pushing that rate down. Next theycan buy a close substitute, treasuries (government bonds), and then diversifyinto other assets. (Note: unless they buy treasuries from the central bank,this shifts reserves about but does not reduce aggregate reserves.) Sincecentral banks target an interest rate (ie: the fed funds rate in the US) theywill react once the interest rate falls below the target. They will begin tobuy treasuries. That eliminates the excess reserves and the downward pressureon interest rates.
-Luigi: “The impact of the deficiton bank reserves has been emphasized by the neo-chartalist school (Bell2001; Wray 1998), but neochartalist writers do not explicitly draw on theconclusion that it supports the complete exogeneity of the long-term rateof interest”. Parguez writes this in 2004, it’s right? How MMT considerlong-term rates of interest?
A: Sounds OK to me. A central bank CAN target a long termrate (ie 30 year treasury bond) and hit it if it wants, but central banksnormally do not. Instead, they target the short end and when they want thelonger term rates to fall, they make statements like “we expect to hold theovernight rate at a low level for the foreseeable future”. That makes holdersof longer maturity bonds more confident that the short term rate will not risesoon—which would cause capital losses. There are a number of approaches to thedetermination of longer term interest rates: expectations theory, habitattheory, and interest rate parity. As a great philosopher once said “you canlook it up”. But in conclusion, yes, MMT agrees that longer rates are complexlydetermined and are not normally exogenously controlled by central banks.
WH10: “Finally, the fear that government might “printmoney” if the supply of finance proves insufficient is exposed as unwarranted.All government spending generates credits to private bank accounts—which couldbe counted as an increase of the money supply (initially, deposits and reservesgo up by an amount equal to the government’s spending).”That’s only halfthe picture for those concerned. Peopleperceive govt spending as being counteracted by bond sales, so the money supplyseemingly does not go up. HOWEVER, is itnot the reality that a significant proportion of bond sales come from bankPrimary Dealers, which ‘spend’ from their reserve accounts, such thateffectively there is a net credit to deposit accounts (as opposed to them beingoffset by purchases of bonds out of deposit accounts)? In other words, it seems we’re almost always’printing money,’ if this is the case.
A: Minsky said: “anyone can create money (things), theproblem lies in getting it accepted”. Yes, we are almost always “printingmoney” in the sense of issuing IOUs denominated in the state money of account.Get over it. On some conditions, that can cause prices of output or offinancial assets to rise. It all depends. There is no automatic channel thatcauses an increase of “money supply” however defined to lead to “inflation”,however defined. And there is nothing that magical with respect to inflationeffects of government spending as opposed to private spending. If I get an autoloan to buy a car, on some conditions that could push up car prices and hencethe CPI. And we could find that some measure of the money supply also hadincreased. If government strokes some keys to add a vehicle to its fleet ofcars, on some conditions that could push up car prices, the CPI, and somemeasure of the money supply. Yes, it is a possible outcome and if you reallywant to point your finger at the increase of the money supply, I guess you can.I would say that it was the increased purchase of autos that in tight markets(full employment, full capacity utilization) would induce manufacturers toincrease prices. Note that could also happen without any additional loans or“printing money”.
WH10: Was there a time when did the U.S. Govt could spendbefore requiring the Treasury’s account to be marked up? If we imagine afiat currency starting out, but Fed overdrafts are not allowed and the sameinstitutional restrictions that we have today are in place, then what are theaccounting statements which allow the government to spend without a positiveaccount? Does this necessitate the existence and willingness of primarydealers that to have their reserve accounts go negative to facilitategovernment spending? Why are they willing to do this?
A: Not exactly sure when the US government decided to tieits shoes together by requiring Treasury to have a credit to its account at theFed before making a payment, but it could date to creation of the Fed in 1913. Whatif there were no Fed? Bank clearing could take place on the books of theTreasury, and the Treasury could simply credit them with reserves whenever itmakes a payment. Even simpler, it could just pay with paper notes or coins. Or,in the old days, with tally sticks. These would be the debt of the governmentand the financial assets of the nongovernment, accepted in tax payment.
Dave: I guess I’ve missed something (though I’ve reviewedthe two previous posts): Given that reserve requirements are defined by the Fed(http://www.federalreserve.gov/…how does the non-governmental sector as a whole acquire “excessreserves” i.e. don’t reserves only grow as much as (in proportion to) thesurplus the non-governmental sector accumulates from deficit spending? Or doyou only mean that SOME agents/banks/actors of the non-governmental sectoraccumulate “excess reserves”? Or….?
A: Banks can get reserves from either the central bank orthe treasury. When treasury buys goods and services, bank reserves arecredited. We normally call that government spending. When the central bank buysfinancial assets from banks (ie: buys government bonds, or private debt, or theIOUs of a “borrowing” bank) that also increases bank reserves. But we do notnormally call that “government spending”. Really it is, but it is spending onassets not on goods and services (so does not show up in GDP).
Joe: OK, so we’re starting to get to the answer of”What if people don’t want to buy the bonds?” Perhaps some examplenumbers, accounts etc. would make thing a bit more concrete as ‘portfoliopreference’ is rather vague. Also, the idea the deficit spending comes first,to provide the reserves to purchase the bonds, seems logical(money must existbefore you can buy bonds), but doesn’t the treasury need a positive balance inorder to spend? Bond sales increase the balance, so there’s a very strongillusion that the proceeds from bond sales are recycled into the treasury’saccount (which I believe is the traditional, pre-1971 view). Did theinterpretation just change in 1971; pre-1971 money from bond sales went intotsy account, post-1971 cash assets are converted to bond assets while new moneyis put into tsy accout? And how can the deficit be mandated to be covered bybonds, if you have to wait for preferences to adjust, there must be some timelag between spending and bond ales? (sorry, lots of questions, I’mpatient, hopefully it’ll all clear up in the coming weeks)
A: Not sure how numbers would help. In the US, where we tiedthe government’s shoes together, the Treasury first sells bonds to specialbanks that buy them by crediting the Treasury’s deposit account. Treasury movesthe account to the Fed before spending. These bonds will be bought by thespecial banks, so at this point the portfolio preferences of the nongovernmentsector do not matter. Deficit spending will increase bank reservesdollar-for-dollar (cash withdrawals will reduce that a bit). As discussed thebanks will try to buy earning assets such as government bonds. The Fed andTreasury coordinate how many bonds will need to be sold by the two of them tooffer earning assets as alternatives to reserves, to allow the Fed to hit its interestrate target. A complication is that in the Treasury’s new issue market, itpursues “debt management”, offering a range of maturities. Occasionally theTreasury might offer a maturity that does not match “portfolio preferences” ofpotential purchasers.
Hugo: According to Vickrey, private capital in the U.S willhave trouble seeking profitable productive investment. Is government bond salesneeded as savings vehicles for the private sector to prevent assets bubbles?
A: Not sure I follow. To prevent asset bubbles, I’d userules, regulation, and supervision of financial institutions. The problemreally is not one of “excess saving”, so trying to “soak up” saving throughgovernment bond sales will not resolve it. If I want to speculate in Martianocean-front condo futures, I do not need any savings. All I need is a bank.
Kostas: “In reality, the Chinese receive Dollars(reserve credits at the Fed) from their export sales to the US (mostly), thenthey adjust their portfolios as they buy higher earning Dollar assets (mostly,Treasuries)”. It would be nice if you could elaborate on how foreigncentral banks get a hold of dollars in their Fed accounts. My understanding isthat this happens when central banks (of surplus countries) intervene inforeign exchange markets in order to maintain their currency foreign exchangevalue (by offering their currency in exchange for foreign assets). Is there anyother way for Bank of China to acquire US$ reserves?
–Dirk: Of course. The People’s Bank of China can borrow/buydollars from abroad. Not only from the US, but from anybody who holds dollars.In case of buying dollars, the counter-party has to accept yuan (not a problem)and the exchange rate might be changing (indeed a problem).
A: Thanks, Dirk, I think you answered.
Neil: “Recipients of government spending then can holdreceipts in the form of a bank deposit, can withdraw cash, or can use thedeposit to spend on goods, services, or assets.” Can’t they also swap itfor another currency with a willing party at an agreed exchange rate? So the ‘shifting of pockets’ surely has anexchange effect as well, not just an interest effect. Or do you see currencyexchange as just another asset purchase and that it will effect themacroeconomy in the same way as any other asset price shift?
A: Yes, I can use a dollar deposit to buy foreign stuff,take vacations abroad, or to buy foreign assets. The dollar deposit will beheld by someone else. My spending abroad can affect the exchange rate.
Andy: What effect,if any, does a reduction in bank depositshave on central banks’ day to day operations? For example if repayment ofprivate debt is greater than bank lending and fiscal tightening by governmentsat the same time.
A: Let us say bank deposits decline due to loan repayment. Whenit comes time to calculate reserve requirements (in the US, more than a monthlater), banks will find they have excess reserves relative to what is requiredon their deposits. They will attempt to individually reduce reserves held bypurchasing bonds (etc). That just shifts the reserves about. But it also pushesthe overnight interest rate down. The central bank responds with an open marketsale of treasuries. So it “forces” the hand of the central bank that reacts tothe interest rate decline.
Suspicious: When will we get the MMP explaining how to credibly regulate a bankingsector ? Banks have always managed to circumvent doctrines, ideologies,regulations, etc. and to wreck havoc the financial system. What’s the purposeof the central bank reserves not being inflationary if banks can loot it viacontrol fraud, and raise prices like in the commodities, and even causehyperinflation if only they were not as greedy as preventing anyone butthemselves to make money on it ?