Originally published October 28, 2011 on the New Economic Perspectives blog.
Sorry this is late—there were a lot of comments and I am traveling. Before we begin, a note and plea: we are getting an increasing number 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 and it would consume all of my time to respond individually. That is why I am collecting comments on the Primer page and responding to all at once. I know some people are having trouble posting comments (I do too!) but what I’ve found is 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 several issues. So this week I am posting most of the text, followed by my response.
Iris: So from my point of view, what you obviously neglect is to insert all the additional references to presupposed material, which us, especially with less “previous knowledge of MMT” could help to better understand the matter you’re dealing with. My impression is for example, that not only general accounting basics are necessary but also knowledge about the structural system of inter-firm-, inter-bank- and government banking-institutions’ accounting to private sector. Would it be asked to much of a favor for you to perhaps offer, at least via footnotes, these hints too
A: Well there is always a trade-off. We don’t want to get into defining the meaning of “the” (as a former President tried to do). I do presume some knowledge and it is tricky just how much should be assumed and how much explained in detail. Sometimes the reader with less background is probably going to have to accept some statements without fully understanding all the details behind them. I have been leaving out detailed accounting for two reasons: it is more detail than most will want, and it is hard to produce these in Word (and I have no idea how hard it is for our techie to post them to the blog). However, since there have been a number of requests, I will devote a blog to “T Accounts”.
Hugo: Ok, “excess deposits” results in increased ”demand for profitable savings vehicles”. And if demand for savings vehicles exceeds supply, the market will adjust. Savers will have to accept lower yields on their savings. Firms would find it more easy to borrow money. Interest rates for corporate bonds would likely decrease (in the financial markets?). But I feel the transmission to an increase in government bonds is somewhat weak here? What you are suggesting seems to be: “Excess deposits seeking profitable savings vehicles” -> “excess reserves in the interbank lending system” -> “overnight rate maintenance” -> “bond sales” -> “equilibrium”..? But how does “excess deposits” necessarily lead to “excess reserves” 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 credits deposits of someone in the banking system, it also credits banks reserves to create asset that offsets banks deposit liability.
A: Ok I am not sure what an excess deposit would be. When I get my paycheck my deposit goes up and surely I do not think it is excessive. I then buy consumption goods and services. I might also make some portfolio decision over what is left, allocating some of my accumulated savings to higher earning financial assets. Hugo says I might bid up bond prices, on both private and government debt, lowering interest rates on the outstanding stock. Right.
He is not convinced that a government deficit will put downward pressure on government bonds, however, because he does not see how my “excess deposits” creates “excess reserves”. Remember that reserves are on the asset side of the bank’s balance sheet while deposits are on the liability side. When government makes 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 be excess reserves (details on this are complicated as reserve requirements are calculated after a lag—let us ignore those details for now).
Banks make a portfolio decision: buy something that earns higher interest rate. First they can lend in the overnight, fed funds, market, pushing that rate down. Next they can buy a close substitute, treasuries (government bonds), and then diversify into other assets. (Note: unless they buy treasuries from the central bank, this shifts reserves about but does not reduce aggregate reserves.) Since central banks target an interest rate (ie: the fed funds rate in the US) they will react once the interest rate falls below the target. They will begin to buy treasuries. That eliminates the excess reserves and the downward pressure on 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 the conclusion that it supports the complete exogeneity of the long-term rate of interest”. Parguez writes this in 2004, it’s right? How MMT consider long-term rates of interest?
A: Sounds OK to me. A central bank CAN target a long term rate (ie 30 year treasury bond) and hit it if it wants, but central banks normally do not. Instead, they target the short end and when they want the longer term rates to fall, they make statements like “we expect to hold the overnight rate at a low level for the foreseeable future”. That makes holders of longer maturity bonds more confident that the short term rate will not rise soon—which would cause capital losses. There are a number of approaches to the determination of longer term interest rates: expectations theory, habitat theory, and interest rate parity. As a great philosopher once said “you can look it up”. But in conclusion, yes, MMT agrees that longer rates are complexly determined and are not normally exogenously controlled by central banks.
WH10: “Finally, the fear that government might “print money” if the supply of finance proves insufficient is exposed as unwarranted. All government spending generates credits to private bank accounts—which could be counted as an increase of the money supply (initially, deposits and reserves go up by an amount equal to the government’s spending). ”That’s only half the picture for those concerned. People perceive govt spending as being counteracted by bond sales, so the money supply seemingly does not go up. HOWEVER, is it not the reality that a significant proportion of bond sales come from bank Primary Dealers, which ‘spend’ from their reserve accounts, such that effectively there is a net credit to deposit accounts (as opposed to them being offset 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), the problem lies in getting it accepted”. Yes, we are almost always “printing money” 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 inflation effects 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 hence the CPI. And we could find that some measure of the money supply also had increased. If government strokes some keys to add a vehicle to its fleet of cars, on some conditions that could push up car prices, the CPI, and some measure of the money supply. Yes, it is a possible outcome and if you really want 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 to increase 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 spend before requiring the Treasury’s account to be marked up? If we imagine a fiat currency starting out, but Fed overdrafts are not allowed and the same institutional restrictions that we have today are in place, then what are the accounting statements which allow the government to spend without a positive account? Does this necessitate the existence and willingness of primary dealers that to have their reserve accounts go negative to facilitate government spending? Why are they willing to do this?
A: Not exactly sure when the US government decided to tie its shoes together by requiring Treasury to have a credit to its account at the Fed before making a payment, but it could date to creation of the Fed in 1913. What if there were no Fed? Bank clearing could take place on the books of the Treasury, and the Treasury could simply credit them with reserves whenever it makes 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 government and the financial assets of the nongovernment, accepted in tax payment.
Dave: I guess I’ve missed something (though I’ve reviewed the 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 “excess reserves” i.e. don’t reserves only grow as much as (in proportion to) the surplus the non-governmental sector accumulates from deficit spending? Or do you only mean that SOME agents/banks/actors of the non-governmental sector accumulate “excess reserves”? Or….?
A: Banks can get reserves from either the central bank or the treasury. When treasury buys goods and services, bank reserves are credited. We normally call that government spending. When the central bank buys financial assets from banks (ie: buys government bonds, or private debt, or the IOUs of a “borrowing” bank) that also increases bank reserves. But we do not normally call that “government spending”. Really it is, but it is spending on assets 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 example numbers, accounts etc. would make thing a bit more concrete as ‘portfolio preference’ is rather vague. Also, the idea the deficit spending comes first, to provide the reserves to purchase the bonds, seems logical (money must exist before you can buy bonds), but doesn’t the treasury need a positive balance in order to spend? Bond sales increase the balance, so there’s a very strong illusion that the proceeds from bond sales are recycled into the treasury’s account (which I believe is the traditional, pre-1971 view). Did the interpretation just change in 1971; pre-1971 money from bond sales went into tsy account, post-1971 cash assets are converted to bond assets while new money is put into tsy account? And how can the deficit be mandated to be covered by bonds, if you have to wait for preferences to adjust, there must be some time lag between spending and bond ales? (sorry, lots of questions, I’m patient, hopefully it’ll all clear up in the coming weeks)
A: Not sure how numbers would help. In the US, where we tied the government’s shoes together, the Treasury first sells bonds to special banks that buy them by crediting the Treasury’s deposit account. Treasury moves the account to the Fed before spending. These bonds will be bought by the special banks, so at this point the portfolio preferences of the nongovernment sector do not matter. Deficit spending will increase bank reserves dollar-for-dollar (cash withdrawals will reduce that a bit). As discussed the banks will try to buy earning assets such as government bonds. The Fed and Treasury coordinate how many bonds will need to be sold by the two of them to offer earning assets as alternatives to reserves, to allow the Fed to hit its interest rate target. A complication is that in the Treasury’s new issue market, it pursues “debt management”, offering a range of maturities. Occasionally the Treasury might offer a maturity that does not match “portfolio preferences” of potential purchasers.
Hugo: According to Vickrey, private capital in the U.S will have trouble seeking profitable productive investment. Is government bond sales needed as savings vehicles for the private sector to prevent assets bubbles?
A: Not sure I follow. To prevent asset bubbles, I’d use rules, regulation, and supervision of financial institutions. The problem really is not one of “excess saving”, so trying to “soak up” saving through government bond sales will not resolve it. If I want to speculate in Martian ocean-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), then they adjust their portfolios as they buy higher earning Dollar assets (mostly, Treasuries)”. It would be nice if you could elaborate on how foreign central banks get a hold of dollars in their Fed accounts. My understanding is that this happens when central banks (of surplus countries) intervene in foreign exchange markets in order to maintain their currency foreign exchange value (by offering their currency in exchange for foreign assets). Is there any other way for Bank of China to acquire US$ reserves?
Dirk: Of course. The People’s Bank of China can borrow/buy dollars 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 hold receipts in the form of a bank deposit, can withdraw cash, or can use the deposit to spend on goods, services, or assets.” Can’t they also swap it for another currency with a willing party at an agreed exchange rate? So the ‘shifting of pockets’ surely has an exchange effect as well, not just an interest effect. Or do you see currency exchange as just another asset purchase and that it will effect the macroeconomy 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 be held by someone else. My spending abroad can affect the exchange rate.
Andy: What effect, if any, does a reduction in bank deposits have on central banks’ day to day operations? For example if repayment of private debt is greater than bank lending and fiscal tightening by governments at the same time.
A: Let us say bank deposits decline due to loan repayment. When it comes time to calculate reserve requirements (in the US, more than a month later), banks will find they have excess reserves relative to what is required on their deposits. They will attempt to individually reduce reserves held by purchasing bonds (etc). That just shifts the reserves about. But it also pushes the overnight interest rate down. The central bank responds with an open market sale of treasuries. So it “forces” the hand of the central bank that reacts to the interest rate decline.
Suspicious: When will we get the MMP explaining how to credibly regulate a banking sector ? Banks have always managed to circumvent doctrines, ideologies, regulations, etc. and to wreak havoc on the financial system. What’s the purpose of the central bank reserves not being inflationary if banks can loot it via control fraud, and raise prices like in the commodities, and even cause hyperinflation if only they were not as greedy as preventing anyone but themselves to make money on it?