1) Suppose the Fed increased interest on reserves from 0.25% to 8% tomorrow and simultaneously began a program to sell few trillion of the assets on it’s balance sheet and announced a new inflation target of 0%. What does the Book of Arnold predict will happen to inflation over the next two years?
2) Suppose the Fed cut the interest rate on reserves to -2%, announced a plan to buy an unlimited amount of financial assets until a market based forecast of NGDP 5 years from now reached $22.5T (5% year over year growth). What does the Book of Arnold predict will happen to NGDP over the next two years?
I think of the Fed as a bank. It makes profits in a weird way. It requires banks to hold reserves, and then it imposes a tax on those reserves by paying a below-market interest rate. Funded this way, it buys assets that earn a market rate of interest. (The Fed also profits from assets obtained with zero-interest-rate currency.)
So in the first exercise, the Fed’s cost of funds would rise from 0.25 percent to 8 percent. If this were to happen to any bank, it would soon be insolvent. For the Fed, this would mean having to go to Congress and beg for a large appropriation to cover its losses. That is such a weird and unlikely scenario that I do not think that any prediction can be made about NGDP.
In the second exercise, the higher tax on bank reserves would make their business less competitive, perhaps even unprofitable. We could see a big decline in bank balance sheets and an increase in shadow banking, or maybe just an increase in reserve-minimization tactics, like sweep accounts. Because reserves would plummet, the Fed’s liabilities would shrink, not rise. The only way that the Fed could expand its balance sheet would be by using currency to pay for its increase in assets.
I am not sure what the net effect would be on nominal GDP. Perhaps in the short run, you get bank failures and other forms of financial disruption, causing actual and forecast nominal GDP to decline.
The phrase “buy an unlimited amount of financial assets until…” makes me want to draw a cartoon with Janet Yellen telling investors “we will administer an unlimited amount of beatings until morale improves.” Again, I wonder how much the Fed could actually buy before running afoul of Congress.
Look, if you get rid of any constraints on the size and maneuverability of the Fed, then sure, they can do something to nominal GDP. And if you get rid of any constraints on the size and maneuverability of my body, then I could play in the NBA.
#1 would bankrupt the majority of US banks and any investment firms that were levered long the market. Anyone with short term debt would wobble and even the the Federal government would be struggling with insolvency in 3-4 years- NGDP would clearly fall in the short to medium term.
#2 is more difficult as it isn’t clear as to what the fed does once NGDP futures hit that mark, and it isn’t clear how the NGDP futures market would work as a vehicle for savings. If the Fed committed to continuing to buy assets until those contracts expired then you have a screwy situation. Anyone holding financial assets would demand a massive premium from the Fed to sell which would functionally drive interest rates of anything acceptable by the Fed into the negative range. Actual NGDP should be expected to run at ~1.5-2% below NGDP futures for the duration of the contract if the market is liquid enough.
Lovely.
But what about the simple prospect (somewhat likely?) that the Fed should sell assets (aka reduce the Balance Sheet) suck up some of the paper (currency & credits) it issued?
Nice and easy, of course; and with no “apparent” pattern.
That’s very clarifying as regards your disagreement with, say, the Scott Sumner view of the Fed.
I was confused and thought your disagreement was more theoretical: that the Fed was the mostly impotent tail instead of the dog, and that the market and inertial, self-validating social expectations set the rates.
But that sentence seems to indicate that your view is more that these theoretical constraints have a limited operational envelope, that the Fed could theoretically have major impacts if it greatly expanded its firepower, but that other – maybe insurmountable – political constrains would still keep the Fed confined to its pen.
Am I reading that right?
At some point, it becomes a different Fed, for example, if it quintuples the size of its balance sheet.
huh. So is your entire objection to “the fed can target NGDP if it wants to” contained in “congress wouldn’t let them”? In that case, would you agree that the People’s bank of China could target NGDP if it were told to to so by the Chinese communist party?
It is a bit more than that. The Fed would have to be several times larger than its already somewhat-bloated self. It would be a different Fed than what we have now.
Wouldn’t that depend on what NGDP path they target? Assuming they can successfully target NGDP, and are doing so, demand for base money is going to be a lot more at 0% NGDP growth than 10% right? So we should expect to see a much more bloated central bank at 0% than at 10%.
So do you disagree with that? Or would you say that targeting NGDP at a reasonable rate of 4% or 6% would be in the range that causes bloating? Or something else?
” Assuming they can successfully target NGDP, and are doing so, demand for base money is going to be a lot more at 0% NGDP growth than 10% right? So we should expect to see a much more bloated central bank at 0% than at 10%.”
What do you mean by “base?”
Central bank reserves, cash, and coins.
Or more generally: “the kind of money that every other kind of money promises to be redeemable in. The kind of money that serves as a unit of account”
So M2? There does not seem to be a discernible relationship between nGDP growth and M2 growth over the past 35 years.
Not M2, literally just cash, reserves, and coins. M2 includes bank deposits and other things that are not base money.
NGDPLT can only exist as a tautology. For it supposes two things that cannot both be true. On the one it it claims monetary intervention is required because prices in the real world do not adjust fast enough (ie sticky wages). On the other hand it supposes the central bank can trade in a market that is efficient and where prices are not sticky.
There is another problem with activist monetary policy. At some point the bank’s book will become so large it will influence the market. We had this once. It was called LTCM. It failed. So too will activist monetary policy.
Huh? The central bank is trading in securities markets that are a lot more efficient than labour markets. Nobody is assuming any market is perfectly efficient.
When the Fed pays interest on reserves isn’t this new money? It would seem to me that they can pay any interest on reserves they want because they can create any balance of reserves they want. I’m not sure I follow #1 in your argument.
“I think of the Fed as a bank. It makes profits in a weird way. It requires banks to hold reserves, and then it imposes a tax on those reserves by paying a below-market interest rate.”
This is historically true.
However, this is not an accurate description of Fed activity in the post-2008 ZIRP model, in which excess reserves have played a major role and a much more important role than required reserves.
Since 2008 the Fed’s monetary expansion, designed to help liquify credit markets and avoid deflation, has been “pushing on a string” i.e. confronting stubborn declines in velocity as a combination of slack loan demand, post-crisis bank conservatism, regulatory pressures on solvency/capital ratios, and interest paid on reserves have caused banks to effectively sterilize the feds quantitative easing by maintaining burgeoning balances of reserves.
The Fed, in order to maintain a positive floor on nominal interest rates (and later, I think, with the motive of sterilizing quantitative easing which was really intended to skew credit allocation and generate wealth effects), began paying interest on excess reserves in October of 2008 and has done so ever since. How they will treat excess reserves as they raise short rates (an operation fraught at this time with the risks of intensifying flows out of EM-funding carry trades and into the Dolllar with exchange rate effects problematic on both sides) is an interesting question.
If they don’t keep interest on reserves competitive with short rates and they don’t raise the level of required reserves then excess reserves will drain out of the Fed system with monetary and/or short-rate issues (via re-establishing fractional reserve multipliers on hitherto reserves and/or bank’s substitution purchases of short paper e.g. t-bills driving short rates below Fed target) as well as credit allocation and fiscal implications (the latter two via the Fed’s sales of treasuries/agencies and concomitant back-up in yields). If the Fed does keep paying interest on large amounts of (excess) reserves in-line with rising short rates or raise required reserves to high levels (while continuing to pay interest on reserves) they will be making massive and politically unpalatable amounts of money to primary dealers depositing at the Fed.
Rossle:
You cite Arnold’s, “I think of the Fed as a bank … profits …[and later] … losses …” depiction of the Fed.
You then state, “That is historically true.”
I have to say that Arnold’s depiction/characterization of the Fed in this way is not only not true, it hasn’t ever been true. Arnold’s depiction of the Fed is a rather gross non sequitur – a complete logical fallacy.
How could the accountancy concepts of “profit” and/or “loss” possibly apply or even be relevant to the Fed – the one sole entity in the universe that has the pre-existing statutory authority to “print/create” money?
I apologize to you in advance if this comment seems to be picking on, or criticizing, you. I don’t have that intent at all.
Thanks for the thoughtful reply. It’s a good point that the Fed would make huge losses if it jacked up the interest rate on reserves to 8%. And it is of course a ridiculous example, but I think it’s useful to imagine two extreme cases, one in which the Fed torpedoes NGDP and one in which it causes NGDP explode and then reason that somewhere in between lies an intermediate case where the Fed can push some nominal variable toward a reasonable target.
If I am interpreting your reply correctly, it is basically that the Fed is constrained legally or politically from taking the kind of drastic actions that would clearly push NGDP up or down. I think implicitly you also have some model in mind in which the Fed’s impact on NGDP is highly nonlinear: Fed can either accept the status quo (the Fed looks impotent), cause a depression, or cause hyperinflation, but nothing in between. I am curious as to what makes a model of the economy work like that.
I doubt that the Fed could cause a Depression. By itself, it can only cause inflation by growing tremendously. The non-linearity is that inflation expectations are anchored, so that people have to perceive a regime change in order for inflation to get much higher. To get hyperinflation, people need to be convinced that they cannot safely lend to the government to finance its deficits, leaving money-printing on an ever-increasing scale as the only option.