Reply to Tyler Cowen

Tyler has a new post where he makes this claim:

More concretely, I am not persuaded by the view that a kind of sheer internal commitment to good outcomes, however sincere, can sustain a peg or nominal target.  The outside world always impinges on the logic of commitment, and thus capital is required.  This is also why I do not agree with Scott Sumner’s claim that a truly credible Swiss target, eliminating the need to expand the SNB balance sheet to make it stick, is possible circa January 2015 or for that matter anytime soon.

Just to be clear, the post Tyler links to does not claim that a currency peg eliminates the need for a central bank to buy assets, and indeed the SNB had to buy assets to maintain its recent peg (which was “truly credible.”)  My claim is different, that the switch to tighter money in Switzerland will require an even bigger balance sheet; it will require the SNB to buy even more assets than the previous policy.

In this case, Tyler may be showing too much respect for the decision-making of central banks.  He should show the SNB exactly the same respect that the market showed their recent decision.  Yes, there are scenarios where you could imagine a central bank having capital problems.  That scenario does not fit Switzerland circa 2015, and even if it did the recent policy switch is not an effective solution. Indeed it makes the problem worse.

Update:  Over at Econlog I reply to a comment by Bill Woolsey.



25 Responses to “Reply to Tyler Cowen”

  1. Gravatar of Market Fiscalist Market Fiscalist
    18. January 2015 at 07:47

    I had a Mike Sprol moment when reading the above.

    When a CB says its going to maintain a stated nominal target no much how much new money its need to print it may well find that the stronger its commitment is perceived to be the less money it actually has to print.

    However if the CB lacks capital its unlikely that it will be able to make any commitment as to the value of its its liabilities as no-one will want to hold them in the first place.

  2. Gravatar of foosion foosion
    18. January 2015 at 08:01

    Sorry to be off topic, but what’s your reaction to this from Gene Fama in Oct 2013:

    Mr. Fama dismissed market concerns about the beginning of tapering of the five-year quantitative-easing program as a non-issue that is getting way too much attention.

    Citing the strategy of buying short-term debt in order to finance the purchase of long-term debt, he said, “I don’t think the Fed[eral Reserve] has any role in how high rates are right now.”

    “I don’t understand why everyone is paying attention to this tapering,” Mr. Fama said.

    “The Fed is using one kind of bond to buy another kind of bond,” he said. “What’s the big deal, and why is anyone taking the Fed seriously?”

  3. Gravatar of Assorted links Assorted links
    18. January 2015 at 10:06

    […] Scott Sumner responds on the Swiss central bank.  And here is James […]

  4. Gravatar of Assorted links – Freedom's Floodgates Assorted links - Freedom's Floodgates
    18. January 2015 at 10:14

    […] Scott Sumner responds on the Swiss central bank. And here is James […]

  5. Gravatar of Major.Freedom Major.Freedom
    18. January 2015 at 10:56

    “My claim is different, that the switch to tighter money in Switzerland will require an even bigger balance sheet; it will require the SNB to buy even more assets than the previous policy.”

    Then that is looser money, not tighter money.

    The more a central bank prints, the looser the central bank is, regardless of what is happening to interest rates or prices or NGDP, which are functions of both central bank activity and activity in the broader economy.

    In terms of direct control, central banks control the quantity of base money they print. They do not directly control aggregate prices or spending or interest rates, so any reasonable gauge of monetary policy should be based on what the central bank does directly control.

    My own actions have an indirect effect on the supply of oil. Every individual’s actions, other than producers of oil, have an indirect effect on the supply of oil. If we want to know how “productive” or how “stringent” the producers of oil are, we do not look at how many times oil is exchanged, or what the supply of oil being exchanged is in the broader economy. We look at how many barrels of oil are being produced.

    The same principle applies to money. If we want to know how loose or tight a central bank is, we look at how many units of currency it issues. The more units it issues, the looser that central bank is, the fewer, the tighter.

    The easiest way to understand that NGDP is an absurd gauge for the stance of monetary policy, is to imagine the case of central bankers going on vacation for the weekend, and over the weekend, all holders of money “turn over” their money balances (for goods and services) at a higher pace than before the vacation. In other words, imagining velocity increasing.

    Now if NGDP were the gauge of monetary policy, then we would have to believe that central banks around the world engaged in a loosening of money over the weekend, without even being there. If that doesn’t sink in, imagine all central bankers suddenly disappearing, and the same increase in velocity took place over the weekend.

    The point is that because NGDP is the outcome of both central bankers and individuals in the broader economy, it is wrong to attribute that outcome solely to any one party, just like it would be wrong to attribute the supply of oil exchanged between individuals in the broader market, as the outcome solely to the producers of oil.

    No, this fact does not get nullified by the existence of a wizard oil producer who can conjure any number of barrels of oil into existence, and he says to himself “I am going to conjure whatever number of barrels of oil into existence such that I later observe a constant rate of X barrels of oil exchanged per month.” What would be happening in this case is still that individuals in the broader market would be partly responsible for how many barrels of oil are being exchanged each month, with the wizard oil producer being the other party partly responsible.

    But since the wizard oil producer cannot predict the future, he will not be able to “target” the number of barrels of oil that get exchanged. The individuals in the broader market remain partly responsible for how many barrels get exchanged.

    The above analogy is why Sumner wants an NGDP futures “market”. The wizard central bankers cannot target NGDP if they can’t predict how many times any given supply of dollars will get “turned over”.

    Going back to the meaning of “looser” and “tighter”, if a higher NGDP means looser money and a lower NGDP means tighter money. If two economies had identical central banker activity, but in one of them, the individuals in the broader market exchanged dollars at a higher rate, then we would have to believe that two identical sets of central bankers, each engaging in identical activities, are nevertheless different in terms of “looser” or “tighter.” But clearly the people who are really looser and tighter in the two economies are both the central bankers and the individuals in the broader market.

    It is absurd to change one’s mind regarding the actions of a central banker, if individuals in the broader market decide to exchange dollars fewer or more often. The absurdity is due to attributing omnipotence to a single party despite the fact that the outcome considered is not the sole outcome of that party.

    It is much more reasonable to define the central bankers as neither looser nor tighter with respect to each other, but who each live in a different economy with different cash holding and spending desires in the broader economy.

    MMs claim that NGDP is a reasonable metric for looseness and tightness because doing so will bring attention to employment. NGDP is a better metric, allegedly, because if unemployment rises and NGDP falls, then MMs think it is then reasonable to tell the world: “Unemployment increased. Since we cannot explain the entirety of the present increase to present real side factors only, it must be the case that present monetary policy is a detrimental factor. Since unemployment rose, we therefore should say “monetary policy is too tight”, so that central bankers will print more money.”

    It is a flaw to attribute an outcome that is the product of both central bankers and individuals in the broader economy, to solely the central bankers at any given time. There is of course the ready-made counter-claim defense that “We are not saying CBs are in direct control of employment. What we are saying is that we want to eliminate unemployment caused by bad, i.e. “tight” monetary policy.”

    Notice the circular logic. They start with the theory that present tight money (defined as decreased NDGP) is the cause for a positive percentage of present unemployment. Since employment is caused by a myriad of factors, they then move to the claim that not all of the present unemployment is caused by present tight money, but rather, some positive percentage of unemployment is caused by tight money. What percentage? We’re supposed to prove the initial theory, by reassuming it, and say, “We should look at how much NGDP decreased of course!” The theory is “proved” by being assumed true as a premise.

    Then we’re shown charts of NGDP changes and employment changes, as if this is supposed to prove their theory of causation.

    MMs do not understand that present decreased NGDP and present increased unemployment are not causally related to each other, but are both the effects of a different prior set of causes. Sumner indicated that he would rather not think about or study or research these causes. He wants to rest content with spurious present correlations between NGDP and unemployment.

    Yet because NGDP is the outcome of both central banks and individuals in the broader market, it is not correct to blame present monetary policy on present unemployment. We have to look at prior actions that lead to this outcome if we want to know why unemployment rose and NGDP fell. No, we cannot say that NGDP is lower today because the central bank tightened up, because again, NGDP is a function of more than just the central bank. Central bank activity, and NGDP, can move in opposite directions when it comes to additions or subtractions. Central banks can remove (add) money in the banks while money is turned over more (fewer) times in the broader economy such that NGDP rises (falls). It is not reasonable to define a central banker selling (buying) treasuries as loosening (tightening) money simply because the combined effect of this and activity in the broader market leads to the outcome of higher (lower) NGDP.

    On topic now, if the SNB grows its balance sheet because of a change to its money printing rule, then this is reasonably defined as loosening, even if NGDP falls. For here NGDP would be “tightened” by both the SNB and individuals in the broader economy. The SNB is more active, but individuals in the broader economy are less active by more, when it comes to money.

    The main reason viewing the world in this way is better than the “omnipotent central bank” view, is that it compels us to analyse why NGDP fell and to find a more complete answer. I don’t like “Since we define falling NGDP as tight money, then the answer for why NGDP fell is that the central bank tightened monetary policy”. To readers who were never satisfied with that argument but didn’t know exactly why, well that is the reason. There are better explanations than the circular “Falling NGDP was caused by tightening monetary policy which we define as falling NGDP.”

    The reason why NGDP fell 2008 has more to do than with just central bank activity in 2008. It has to do with central bank activity (and activity from individuals in the broader economy) prior to 2008, but that is for another day…

  6. Gravatar of ssumner ssumner
    18. January 2015 at 11:19

    foosion, I suppose he’s missing the signaling aspect of policy.

  7. Gravatar of BC BC
    18. January 2015 at 12:39

    “Yes, there are scenarios where you could imagine a central bank having capital problems.”

    I can understand that a bank may need capital to *increase* its currency’s value, but would a bank ever have capital problems in trying to *devalue* its currency as the SNB was trying to do? Even an individual with zero capital can issue promissory notes and drive those notes’ value to zero, right?

  8. Gravatar of Simon Simon
    18. January 2015 at 13:21

    Scott, could you please reply to Kevin Drum’s request for explanation? There appears to be some honest misunderstanding on his part, especially towards the end of his post.

  9. Gravatar of Ray Lopez Ray Lopez
    18. January 2015 at 20:25

    @Simon- thanks for that great article by the liberal Kevin Drum of MotherJones. Well done and a very good read; as Drum predicts MM will crumble in the first panic. And, much as I’ve gotten to like our own Scott Sumner (sad by true), you have to ask, what exactly does Scott bring to the party that Krugman and DeLong and others have not already explained? I’d like to see a chronology of Sumner papers, by year, to see how long Sumner has been advocating Targeting NGDP. G. Selgin has been talking about his productivity norm framework (which Sumner says is related to NGDP targeting) since at least 1997. When did Sumner start advocating his framework? I do understand it’s not novel (neither is Selgin’s), but in the modern era (post 1990), when did Sumner start? He could be simply cribbing another’s work.

    Off-topic: I am interested in learning about this, any pointers other than what I can find after 20 minutes with Google?:

    In 1935, economist Irving Fisher proposed a system of 100% reserve banking as a means of reversing the deflation of the Great depression. He wrote: “100 per cent banking […] would give the Federal Reserve absolute control over the money supply. Recall that under the present fractional-reserve system of depository institutions, the money supply is determined in the short run by such non-policy variables as the currency/deposit ratio of the public and the excess reserve ratio of depository institutions.”

  10. Gravatar of Daniel Daniel
    19. January 2015 at 01:42

    So, basically, Kevin Drum believes that running the printing presses at will won’t generate the desired inflation ?

    He’s an idiot.

  11. Gravatar of Michael Byrnes Michael Byrnes
    19. January 2015 at 03:40

    Market Fiscalist wrote:

    “However if the CB lacks capital its unlikely that it will be able to make any commitment as to the value of its its liabilities as no-one will want to hold them in the first place.”

    ‘No one will want to hold them in the first place” is the direction they were trying to go. They were holding the value of the Swiss Franc down relative to the Euro by promising to create francs and buy Euros. If people lost some faith in the franc that would just do their job for them – there would be fewer buyers of francs, which is what they want.

  12. Gravatar of ssumner ssumner
    19. January 2015 at 06:42

    BC, The fear is that the SNB would need capital in the future, when it needed to run a tighter money policy. It’s not a real problem, but I can see how people might worry about it.

    Simon, Thanks, I’ll take a look.

    Ray, My first published paper advocating NGDP futures targeting was 1989. I presented the paper at the 1987 AEA meetings. But what difference does it make? NGDP targeting was advocated in the 1920s and 1930s.

  13. Gravatar of Chuck Chuck
    19. January 2015 at 16:11

    The market “respected” the SNB’s decision by revaluing their banknotes upward. The SNB is considered more creditworthy than before.

  14. Gravatar of Bob Murphy Bob Murphy
    19. January 2015 at 19:13

    Scott, you know how you operationally define “tight” money as that which makes NGDP grow below target?

    Well why can’t Tyler Cowen say, “The Swiss policy was not credible, because they dropped it. Period.” ?

  15. Gravatar of ssumner ssumner
    20. January 2015 at 06:16

    Bob, Because it clearly was credible. The fact that they dropped it has no bearing on whether it was credible.

  16. Gravatar of Market Fiscalist Market Fiscalist
    20. January 2015 at 07:22

    “Bob, Because it clearly was credible. The fact that they dropped it has no bearing on whether it was credible.”

    Before the SNB abandoned the policy. anyone who thought “I do not believe the SNB will maintain this policy beyond 3 years – I do not find that a credible commitment” would have been right though, wouldn’t they ? Seen in that way the policy was not credible, which I take to be what Bob is claiming.

  17. Gravatar of ssumner ssumner
    21. January 2015 at 14:24

    Market, But that’s not what “credible” means. It has nothing to do with being right, it’s about being believed.

  18. Gravatar of Bob Murphy Bob Murphy
    22. January 2015 at 09:49

    Scott you are making this difficult. 🙂 Someone who said, “I do not believe the Swiss can keep this up” 6 months ago would have a prima facie case on his side.

    You are saying such a person–including Tyler Cowen–is self-evidently wrong. But I think it’s similarly obvious that the Fed has had loose money since late 2008.

  19. Gravatar of ssumner ssumner
    22. January 2015 at 17:20

    Bob, By your logic someone who said (6 months ago) “I do not believe Scott can keep up teaching for more than 6 more months” would have been proved right, as I just stopped teaching. So does that prove I could not teach more if I wanted to? Clearly the markets thought the SNB could have kept the peg going. The SNB said it could have. No one has explained why it could not have done so. So I’m going to keep assuming it could have kept the peg going, unless someone explains to me why it could not.

    As far as loose money, you need a definition that makes sense. I don’t recall you providing one. Did the German central bank have tight money is 1923? I’d guess you’d say no, so you aren’t using interest rates. Did the Fed have tight money in 1930-32? I’d guess you’d say yes. But they did QE in 1932. So what is your criterion?

  20. Gravatar of Major.Freedom Major.Freedom
    22. January 2015 at 19:51

    “As far as loose money, you need a definition that makes sense.”

    WARNING…thou definition shall tacitly promote NGDPLT.

  21. Gravatar of Bob Murphy Bob Murphy
    23. January 2015 at 11:13

    My debate with Scott is unsustainable. This acknowledges my lack of resources, not resolve.

  22. Gravatar of Market Fiscalist Market Fiscalist
    23. January 2015 at 14:15

    Sounds like “credible” is the wrong word to describe the kind of monetary policy MMs want the CB to commit to. Credible only implies “it is believable that the CB will stick to its target but won’t be that much of a surprise if they drop it”. Which doesn’t sound much use.

    Surely what is required is something stronger like “it is not only believable that they will stick to its target but it is strongly felt they are unconditionally committed to it”. That goes way beyond credible.

  23. Gravatar of Major.Freedom Major.Freedom
    23. January 2015 at 18:52

    “Because it clearly was credible. The fact that they dropped it has no bearing on whether it was credible.”

    So “credible” really means “the public believes in it and continues to act in accordance with that belief”.

    Smoke and mirrors.

  24. Gravatar of Bob Murphy Bob Murphy
    23. January 2015 at 21:18

    One last thing before I drop it:

    Scott, I understand what you are saying, and the distinction between something being credible and something continuing.

    But Tyler referred to the policy as non-credible, and you say he’s wrong. I’m pointing out that he has a pretty good piece of evidence on his side. In contrast, your first response was to say, “It was credible because it was clearly credible.” No, it wasn’t clear to Tyler.

    When you say, “Explain to me why it wasn’t credible,” why can’t we say what the financial analysts are saying? That the ECB was going to devalue the euro and the SNB wasn’t willing to go down that path to the bottom?

  25. Gravatar of ssumner ssumner
    24. January 2015 at 07:14

    Bob, The markets said it was credible, right up until the last day. That’s the only sense in which “credibility” matters. It doesn’t matter if you cherry pick one person who didn’t buy it, we are talking about the market. It was credible to the markets, check out the futures prices.

Leave a Reply