Mosler on Treasury Rates and Fed Policy

By Dan Kervick

It’s starting to look like QE might be indirectly responsible for a dangerously volatile situation in US financial markets.  And 10-year Treasury notes hit a 2-year high following today’s Fed statement.

But, in my opinion, it’s not the intrinsic nature of the policy itself that has created the danger, but all of the ridiculous and misleading hullabaloo and punditry that has surrounded it. I don’t blame the Fed for using asset purchases to hold down long-term interest rates. But I do blame all of the market pundits and neo-monetarist theorists out there who have grossly misrepresented these asset purchases as something they are not: an all-embracing attempt to manage aggregate demand, gross spending and employment by “pumping money into the economy.”

Unfortunately, there now seem to be a not-insignificant number of people who believe:

1. The Fed has been “keeping the economy afloat” by its asset purchases which “pour money into the economy.”

and

2. The future volume of Fed asset purchases depends inversely on its perception of the health of the private sector economy.

The second claim is certainly true, but the first is ridiculous.  But as a result of believing both of these things together, Mr. Market now responds to good news by taking bearish actions, thus turning good news into bad news. And it responds to bad news by breathing a sigh of relief and taking economically positive steps that turn the bad news into good news!  If indicators tick up, the markets believe the Fed will taper its bond-buying early, and so their misguided believe in claim #1 leads them to act as if they just received bad news.  If indicators are sour, they assume that means the Fed will not taper early and they turn bull.

The result could be that the popularity of the neo-monetarist theories of central bank omnipotence have produced distorted psychological dispositions that amount to a self-regulating mechanism for keeping the economy is a stagnant, low-output equilibrium.  This is nuts.  I think it would be wise for the Fed to communicate more clearly and bluntly about what asset purchase policies do, and what they don’t do.  The economy can’t function normally if economically important groups of fools think a maestro is controlling it it all.  Of course some monetarists, who are deeply authoritarian at heart, have sometimes actually endorsed having an economy run by this kind of slavish awe and psychological manipulation.

Here’s what Warren Mosler has to say about the current situation.  Mosler’s recommendation is that rather than set the Fed Funds rate only, and then letting the market set Treasury rates in response to Fed communication, the Fed should simply target risk-fee interest rates and peg the entire set of Treasury rates along the yield curve at their target ceiling:

… the ‘markets’ have become concerned about ‘supply side’ effects of ‘tapering’, fearing fewer Fed purchases will mean higher term rates, and particularly higher mortgage rates, even as the Fed funds rate remains well anchored at current levels.

So what we are seeing is a ‘market determined’ decision to hike longer term rates based mainly on supply fears and not on fears of ‘excess demand’ driving up rates.

And adding to the volatility is the notion that should the Fed someday decide things have ‘normalized’ and the economy no longer ‘needs’ negative real rates, that would translate into the Fed shifting the Fed funds rate to maybe 3-4% to be at some ‘neutral’ spread to ‘inflation’ etc.

So it’s pretty much binary- we stay near 0 (life support) until it’s time to ‘normalize’ to 3-4%.

All of this makes it rational for market indifference levels to shift a full 1% or more-turning term financing on or off- on nuances of Fed language.

All of which comes back to the fact that the term structure of risk free rates, with floating fx, is necessarily a policy decision, best left to political decision vs market decision, because with floating fx ‘market decision’ is necessarily nothing more than forecasting reaction functions of those setting the rates. That is, the ‘feedback’ from today’s rates determined by market forces is nothing more than what markets think the Fed will vote into policy down the road.

And I see no value in paying the real price of the volatility/uncertainty we are seeing to get that kind of information.

So best to cut the highly disruptive volatility that goes with letting markets determine longer term risk free rates by having the Fed peg the entire term structure of risk free rates with, for example, a bid for the entire tsy curve at their target rate ceiling, along with an open ended securities lending facility.

That is, in my humble opinion best for the FOMC to vote on the entire term structure of risk free rates than today’s policy of voting on just the Fed funds rate and using ‘language’ to influence the curve.


Cross-posted from Rugged Egalitarianism

Follow @DanMKervick

22 responses to “Mosler on Treasury Rates and Fed Policy

  1. I wonder how Warren squares that with his claim that “I would make the current zero interest rate policy permanent.” See:

    http://www.huffingtonpost.com/warren-mosler/proposals-for-the-banking_b_432105.html

    I agree with the latter statement of his rather than what he says in the passage cited above by Dan about the Fed adjusting rates.

    I.e. I think that ideally, aggregate demand should be influenced by adjusting fiscal rather than monetary policy (as I think do most MMTers?). That’s because adjusting interest rates is distortionary: it works only via investment. In contrast, fiscal policy can be almost distortion free. I.e. one can easily increase ALL TYPES OF government spending, and increase household spending via e.g. tax cuts.

  2. Auburn Parks

    Hey ladies and gentlemen.

    I’ve been trying to get to the bottom of this and I would like to get some friendly insight. OK, so let me set the stage. This is the beginning of a conversation I had over at PragCap with one of the regular contributors over there.
    I started by saying that deficit spending is not redistributing bank dollars, but that you could accurately portray (without getting too far out there) that taxation and spending up to that amount tax collections could be described as redistributing inside money (bank deposits).
    He responded with their usual line, that the Govt is not currently allowed to create money (its some weird ideological thing they’ve got going on over there).
    Anyways, here’s the important parts. I asked how deficit spending could simply be redistributing inside money when Congress couldn’t even pay back the principle on maturing securities if it wanted to, thats the Fed’s job.
    I.E. Congress runs budget surpluses of exactly $1 dollar per year for 35 years. So they accumulate (destroy) a net $35 of private sector wealth. So if Congress wanted to “pay back” the debt in a surplus environment, how would that be possible in this example given their framework? My claim would be that over the 35 year period of surpluses, nearly the entire stock of outstanding securities would have matured. And without issuing any new bonds, the debt would be nearly eliminated. And this, without congress having to tax $16 trillion of surplus wealth from the private sector to then redistribute the inside money to bond holders. No its the Fed job to simply transfer your money from one Fed account to another using its magic money computer spreadsheet. Congress plays no role here.

    And nobody could come up with a coherent response to this very simple example. I mean we just had four years of surplus 13 years ago. Did Congress appropriate all that surplus money to come out of the TGA to pay back maturing securities? First, you cant find that line on a budget from the period and second, if Congress spent all the surplus money that it acquired through taxation, then by definition….THERE WOULD BE NO SURPLUS!

    Please friends, somebody comfort me and tell me I’m not crazy.

    • Sorry, I’m not following. The Fed doesn’t pay the Treasury’s debt.

      • Auburn Parks

        So what you are saying is that there is congressional appropriated money from the budgets during the Clinton surplus years that came out of the Treasury General Account, and was spent on maturing securities principle repayment?

        And if that is the case, then the PragCap claim that Congress can only redistribute inside bank deposit money would be correct? Even though the Govt, is the monopoly supplier of securities and reserves (NFA’s) Congress cannot currently create its own new bank deposits?

        • Well, first of all, Congress makes all the rules and under constitutional law is ultimately responsible for monetary policy. So if there are any rules currently in effect as to what Congress itself can and cannot do in conducting its fiscal policy, those are rules that Congress has imposed on itself, and that it can decide to un-impose any time it wants. It has legally delegated some of its monetary authority to the Fed, but can un-delegate it wants to by changing the laws.

          Second, in MMT decriptions of the role of government, “government” refers to the combined government and the government’s balance sheet is the consolidated balance sheet of the entire government, which includes the central bank. The central bank can simply emit dollars. It doesn’t need to get them from some other source. Whether or not it has positive income in any given period, and whether or not it even operates with a balance sheet in a condition of positive “equity”, is solely a matter of monetary policy choice, not an operational constraint.

          Third, there are good reasons for thinking of treasury securities as simply another form of money. Warren Mosler has a post this week in which he makes that case. Coincidentally, I was just authoring my own post on Warren’s proposal, and should have it up within a few hours.

          Here’s the Treasury’s monthly statement of receipts and outlays. In Table 3, you can see the gross outlays for interest on Treasury debt.

          http://www.fms.treas.gov/mts/mts0713.pdf

          I don’t believe Congress needs to appropriate money for debt service. It authorizes certain forms of spending and appropriates money for that spending spending, and raises revenues via taxation. It then allows the Treasury to issue debt up to a statutory limit to carry out any authorized spending for which the appropriations are inadequate. Once the debt has been issued, it is an obligation of the US government and the Treasurer doesn’t need further authorization to pay it. But the amount of the debt service obligation is going to determine how much new debt must be issued to carry out all authorized spending, especially in fulfilling the contracts and other obligations that have already been made.

          • As a committed MMTer, of course I agree with your first paragraph 100%. Which is why this discussion is on the minutiae of operational details.

            “Second, in MMT decriptions of the role of government, “government” refers to the combined government and the government’s balance sheet is the consolidated balance sheet of the entire government, which includes the central bank.”
            Yes, I too view the CB as part of the Govt. And accept the reality of consolidated balance sheet analysis.

            “The central bank can simply emit dollars. It doesn’t need to get them from some other source. Whether or not it has positive income in any given period, and whether or not it even operates with a balance sheet in a condition of positive “equity”, is solely a matter of monetary policy choice, not an operational constraint.”
            However, the Fed has no current mechanism for currently creating and injecting net new bank deposits into the economy. It swaps currently existing bonds for reserves or deposits, depending on who the bond holder is (reserves for banks, primarily or deposits for non-bank bond holders, but its still an asset swap).

            “Third, there are good reasons for thinking of treasury securities as simply another form of money. Warren Mosler has a post this week in which he makes that case. Coincidentally, I was just authoring my own post on Warren’s proposal, and should have it up within a few hours.”
            I too consider them money, however the guys over at pragcap do not. And this discussion is about their views and how to reconcile them with MMT perspectives. This whole thing is about the operational process of redeeming securities into deposits and reserves, not about whether USTS’s are “money”

            “Here’s the Treasury’s monthly statement of receipts and outlays. In Table 3, you can see the gross outlays for interest on Treasury debt.”
            Its obvious that Congress appropriates money to pay the interest. I would have never made the original comment about something that simple. Again, I am talking about the operational details when the budget is in surplus and no new securities are being issued to rollover the maturing principle on already existing securities.
            If the budget was perfectly balanced for 30 years. Would Congress have to appropriate the funds to come out of the TGA to pay back the maturing securities principle?
            For example, if congressional budget outlays and spending for FY 2014 are exactly equal at $1 trillion each. $200 billion for defense, $200 B for SS, $200B for infrastructure, $200 B for Govt institutions and $200 B for Medicare. As securities mature, would Congress have to divert money from other spending programs to pay back the maturing principle?
            To continue the example, say in FY 2015, we have $200 billion in projected maturing securities. Would we have to cut spending or raise taxes an equal amount and reappropriate that money for bond repayment?
            This is the question I am trying to answer. If the answer is yes, Congress must pay for the maturing principle out of either new taxes or cutting spending in order to maintain a balanced budget, well then I see that as a problem in discussion with the MR guys. Their model putting the Govt as simply a redistributor of bank created inside deposit money would be more accurate than saying that Congress creates money every time it deficit spends. Which is how I’ve always described and the nexus of my cognitive dissonance, and I absolutely hate cognitive dissonance.

            • However, the Fed has no current mechanism for currently creating and injecting net new bank deposits into the economy. It swaps currently existing bonds for reserves or deposits, depending on who the bond holder is (reserves for banks, primarily or deposits for non-bank bond holders, but its still an asset swap).

              Yes, it does. The Fed now pays interest on reserves, which isn’t a swap. It’s just a net addition of money.

              • right, I overlooked that one. But that seems to be the only way that happens currently, and IOR was just implemented recently and probably wont be around for long. So I think its probably best to think of that as a special circumstance, not a normal part of CB operations.

                • Not sure about that. IOR isn’t really considered part of “unconventional” monetary operations. It was idea that was around for a long time about how to better target the policy rate.

        • But this idea that Congress can only re-distribute “bank money” is quite wrong in my opinion. When a taxpayer pays a tax by issuing a payment order on a bank deposit account, the bank ultimately pays the Treasury via a transfer of some of its reserve balances to the Treasury’s general fund. So in taxation the government is collecting back some of its own liabilities. The Treasury’s general fund is an account at the Fed, just like a bank’s account at the Fed.

          • Sure, when a taxpayer G pays taxes, the process works how you describe. Taxpayer G’s bank eliminates $X from his deposit account and then eliminates the same amount $X from the bank’s reserve account at the Fed to complete the tax obligation. Then the TGA at the Fed gets credited with that amount of reserves from the bank.

            Then Treasury spends the money on programs in the opposite fashion, its reserve account gets debited $X and the SS recipient H’s bank’s reserve account gets credited with $X of reserves, which in turn they pass on to the SS recipient as $X of deposits.
            So, its not illogical to consider this process as redistributing Taxpayer G’s deposit money to SS recipient H’s deposit account by way of reserve exchanges amongst the banks and TGA at the Fed.

            Now, lets look at the accounting of deficit spending. Bondbuyer A buys an $X bond by exchanging his deposit for a bond. Person A’s bank reserves are debited by $X at the Fed. And a securities account at the Fed gets credited by $X. The TGA gets credited $X in reserves.
            Next, Congress deficit spends $X on SS recipient B . The TGA gets debited $X in reserves, SS recipient B’s bank has its reserve account credited by $X and the bank in turn credits SS recipient B’s deposit account $X.

            I think that covers these interactions, of course please correct me if I’m wrong. Moving on to bond repayment analysis. And this is the part I’m having trouble with.
            When bondbuyer A’s security matures (in a deficit environment), her bank account gets credited with $X + interest. Her bank’s reserve account gets credited $X + interest of reserves. A securities account at the fed gets debited by $X + interest. The TGA reserve account gets debited by $X + interest to pay back the maturing bond. And it replaces those reserves with a new bond sale repeating the process. This would be how the TGA account never gets net lower on redeeming maturing securities in a deficit environment, by issuing new bonds to cover the reserve depletion of old maturing bonds.
            If the Govt’s budget was balanced and no new bonds were sold to replace the TGA reserves lost due to bondbuyer A’s bond maturation, then Congress would have to move money around to accommodate the principle repayment?
            If your comment about the Fed not repaying principle with its own money, only the Treasury repays maturing principle out of its TGA is true, then my initial comment and argument have all been wrong. I supposed that in a balanced budget environment, when bondbuyer A’s bond matures, the Fed credits her bank’s reserve account $X + interest an the Fed debits the Securities account $X and the TGA the amount of the interest, and the bank then credits bondbuyer A’s deposit account $X + interest (the sum of the debit from the securities account and the interest payment from the TGA). In this way, the TGA would not be paying back the principle on bondbuyer A’s maturing security and Congressional deficit spending would truly be creating new bank deposits. I fear I might be wrong now.

            Now, I view this operation as largely irrelevant since marking up or down accounts via a computer is an infinitely open-ended process, and any constraints are all self-imposed. However, its not enough to only try to develop and understand MMT for ourselves, we must spread the word to the uninitiated. And when the guys over a Pragcap make a logical, coherent argument or claim about the operational reality I am trying to reconcile that with what I perceive to be conflicting claims about reality.
            To me, it comes down to how many of MMT claims are currently in place right now, and how many or what laws would need to be changed, to make MMT descriptions more accurate moving forward?

            • bubbleRefuge

              Not following the use case here. Can you simplify it?

              PragCap appears to be spinning around the operational details in order to obfuscate the fact that government deficit spending creates deposits into only the private banking system creates deposits.
              This cannot be logically argued. They reject the government as monopoly issuer of the currency and claim this power has been ceded to the banks. All of it is an attempt to justify the vampire squid that is the financial sector ( after he realized the anti-finance wave that exists in the MMT community) for which the author of PragCap derives his probably large income from.

              Having said that, I like the PragCap site because of its MMT view + discussion of the financial markets. Something which doesn’t exists since Mosler shut down his comments section.

              • Auburn Parks

                That’s how I’ve always perceived the situation bubble, but now I’m not so sure. Lets take for example these accounting entries of what happens when both a non-bank person buys a T-bond (it seems we get a net zero in new bank deposits) and when a bank buys a T-bond (here it seems we do in fact get net new bank deposits and its the reserves that net to zero).

                Bond purchase accounting (step 1):
                Bond buyer A +100 security -100 deposit
                A’s Bank B -100 reserves
                TGA +100 reserves

                Deficit spending accounting (step 2):
                TGA -100 reserves
                SS recipient M deposit +100
                SS Recipient M’s Bank B +100 reserves

                At this point there would be a net zero change in bank deposits. But Bond buyer A would have a NFA bond

                Paying back the bond accounting (step 3):
                Bond buyer A’s deposit +100 (+interest)
                A’s bank B +100 reserves
                TGA -100 reserves
                Bank B’s Securities account -100

                Assuming that the TGA sold a bond to another non-bank depositor, then bond buyer B’s deposit account would have gone down 100 (basically a repeat of the step 1 accounting) so there is a continual net zero in new bank deposits, and our continuing NFA is the Securities

                Now if the bond purchaser is a bank…..(step 1)
                Bondbuying bank B securities account +100
                Bank B’s reserve account -100
                TGA +100 reserves

                Now the deficit spending (step 2):
                TGA -100 reserves
                SS recipient M deposit +100
                SS Recipient M’s Bank B +100 reserves

                And the bond redemption accounting (step 3):
                Bank B’s securities account -100
                Bank B’s reserve account +100
                TGA reserves account -100

                Now in this example, assuming that the new bond purchaser is also a bank (a repeat of the bank purchaser example step #1) whose initial accounting transaction would be a -100 reserves to make up for that last line of TGA -100. To this extant, we seem to actually get a net increase in bank deposits +100

            • If the Govt’s budget was balanced and no new bonds were sold to replace the TGA reserves lost due to bondbuyer A’s bond maturation, then Congress would have to move money around to accommodate the principle repayment?

              Well this isn’t that much different than what has been happening with the debt ceiling standoffs right? Congress has imposed a ceiling beyond which the Treasury cannot sell additional debt. As a result, if the debt ceiling is reached, the Treasury has to forego some spending that Congress has actually authorized.

              I’m not sure what “moving money around” means in this context. It could raise taxes, or it could cut spending, or it could authorize the sale of more bonds. Or it could do something more unconventional and radical I suppose, like authorizing the Treasury to issue currency directly.

              I supposed that in a balanced budget environment, when bondbuyer A’s bond matures, the Fed credits her bank’s reserve account $X + interest an the Fed debits the Securities account $X and the TGA the amount of the interest, and the bank then credits bondbuyer A’s deposit account $X + interest (the sum of the debit from the securities account and the interest payment from the TGA). In this way, the TGA would not be paying back the principle on bondbuyer A’s maturing security and Congressional deficit spending would truly be creating new bank deposits. I fear I might be wrong now.

              No, I don’t think this can happen. The Treasurer of the US is responsible for paying the principal and interest on the public debt. The Fed is just the fiscal agent. In other words, the Fed is the bank at which the Treasury’s account is held, and thus processes the payments. When either kind of payment is made, a Treasury account is debited.

              We could have a situation in which the Treasury is given an open line of credit at the Fed, so that if the Treasury can’t issue more bonds, the Fed just pays the obligation and charges an overdraft to the Fed. But I don’t think this can happen anymore. Treasury used to have an overdraft facility with the Fed, but it expired long ago.

              The question as to how many of the MMT claims are currently in place now depends on properly interpreting the word “government” in the MMT writings. Government does not mean “Treasury Department.” The government is the whole United States government, of which the Treasury Department is only one branch. The power of the purse of the US government is vested in Congress; so is the monetary authority of the US government. The Treasurer is juts the clerk, authorized to make and receive payments on the government’s behalf, subject to the approval of Congress. The Fed is an agency Congress has set up to exercise that monetary authority.

              Warren Mosler usually uses the term net financial assets in talking about what the government can create. Even under current arrangements, a treasury bill is a financial asset, and the Treasury can issue and sell them. So if the Treasury issues a $10,000 t-bill, sells it for $9,900, and then spends that $9,900 back into the private sector, the private sector now has financial assets of $19,900 where it formerly just had financial assets of $9,900.

              • Auburn Parks

                By moving money around, I mean they would have to either cut other spending or raise taxes in order to maintain the balanced budget. So we should find some proof of all this during the budgets of the Clinton era surpluses. Somewhere there should be some budget lines where Congress had to appropriate X amount of dollars to come out of the TGA to redeem maturing securities since they couldn’t be redeemed simply by issuing new securities.

                Thanks Dan for engaging, I know that you’re a busy man. I’m pretty sure I have all the accounting right in my response to Bubble above. And if so, it turns out that Congress can currently create net new bank deposits as long as the bonds are being purchased with bank reserves and not non-bank deposits

                • No problem Auburn. This is fun to think about.

                  I don’t think you will find an appropriation for debt repayment. The US Code just requires the Treasurer to pay off Treasury debt. By authorizing the issuance of debt in the first place to pay for authorized spending if necessary, the authorization to pay it off is automatic. If there is a surplus, redeeming the debt is easy, right? The Treasurer has more funds to play with than is needed for all of the other spending, and so can pay debt without rolling all of it over. The new debt incurred is less than the old debt paid off, and so the total national debt declines. If there is a deficit or balanced budget, on the other hand, then the Treasurer still pays interest and the principal on maturing debt, but to do so while meeting all of the other obligations requires issuing more debt than the amount paid off, so the national debt increases.

      • I think Auburn’s trying to reconcile the traditional notion of “paying back the debt” via congress taxing the money and then giving it to bond holders (retiring the bond in the process), vs what Mosler says “the fed just transfers money from a savings account to a checking account”. So it’s a question of the mechanics of different processes of exactly which accounts get which. In Mosler’s scenario, the bond doesn’t actually get retired, right? For the holder of the bond, it is in effect a switch from a savings account (the bond) to a checking account, but the Fed now holds the bond (just like QEII, the fed purchased 600B of treasuries which are now on the Fed’s books, while the original holders of the bonds now have reserves). OR congress could run a surplus, give a bond holder the principal and interest and retire the bond…

        It’s largely academic anyways, since running a surplus long enough to “pay off the debt” is basically impossible, b/c the private sector would be in deficit that long. but it is an important point of process/mechanics to understand. (maybe if you had 100 years of declining population and needed surpluses to manage inflation, *maybe* some of the debt could be paid back, hardly worth thinking about.)

  3. Good post, Dan. Warren’s suggestion would really cut through a lot of the fog surrounding Fed operations.

    Auburn– My understanding is that the Treasury issues bonds and is also responsible for paying the interest and repaying the principal. They may have to issue new bonds to pay off the old bonds. I’m not sure how the accounting and budgeting works for this, however.

  4. Pingback: Mosler on Treasury Rates and Fed Policy | The M...

  5. The economy took off like gangbusters in 1983 when the effective federal funds rate was over 8%. In September of 1983, 1 million jobs were created.

  6. “The Fed has been “keeping the economy afloat” by its asset purchases which “pour money into the economy.””
    I think what the market means is the Fed has been keeping the economy afloat by keeping interest rates low which, in turn, allows more investment. So when QE ends rates will go up of necessity. So, as Warren says, the market rate will go up to a more normal range with no QE support. Warren’s idea would work but they are not going to do it – – we know that.
    My guess would be the tapering of QE will drive the market down – for at least a time- and could bring on a recession. Call it what you want but the market is fearful of higher rates, not that the Fed is pumping money into the economy, except indirectly.

  7. Auburn – maybe the USC §3111 and §3112 can help?

    31 USC 3111 – Sec. 3111. New issue used to buy, redeem, or refund outstanding obligations: http://us-code.vlex.com/vid/used-buy-redeem-refund-outstanding-19220720

    §3111 mentions the TGA: “An obligation may be issued under this chapter to buy, redeem, or refund, at or before maturity, outstanding bonds, notes, certificates of indebtedness, Treasury bills, or savings certificates of the United States Government. Under regulations of the Secretary of the Treasury, money received from the sale of an obligation and other money in the general fund of the Treasury may be used in making the purchases, redemptions, or refunds.”

    31 USC 3112 – Sec. 3112. Sinking fund for retiring and cancelling bonds and notes:
    http://us-code.vlex.com/vid/sinking-retiring-cancelling-bonds-notes-19220715

    ***

    Thinking about the Clinton-era surplus … the fiscal stance may not have to change in order to pay down more debt; it only requires a growing economy which increases the amount of receipts while the government restrains outlays from increasing as much.