Answers: Taking IOR to Zero
I want to thank all the commenters on my last post — at Angry Bear, at Asymptosis, and at Mike Norman’s blog. You’ve provided me with exactly the education I hoped to achieve. Here’s hoping others benefited similarly.
I asked: what would happen if the the Fed cut the interest rate on reserves from its current .25% to zero. I was not suggesting it should be done. I simply wanted to understand what would happen if that one variable changed.
I want to summarize the conclusions I’ve come to based on all the discussion.
This is me speaking, based on sifting and considering all the responses. I won’t link to all the excellent comments that brought me to this. (Though I do want to highlight the Angry Bear comment by Bad Tux beginning “At 0% IOR”. It arguably explains things better than I do here, and at significantly less length.)
First, the market monetarists responses. Scott Sumner said (in comments a while back on his blog, which I linked from my original post):
It could be slightly expansionary, or if accompanied by other moves, wildly expansionary.
I’m presuming he says “slightly expansionary” based on the theory Mark Thoma gives us in a November 17 post that Cameron was nice enough to link to on Angry Bear:
it would slightly lower the incentive for banks to hold cash rather than loaning it out, and more loans would help to spur the economy
So banks could lend a quarter point cheaper, or loosen their lending requirements slightly. Assuming there’s some decent amount of demand at lower rates (elasticity of demand is appreciably > 0), or that good borrowers are asking for loans but being turned down cause they’re too risky, this could have a small effect. Scott’s “other moves” presumably include NGDP level targeting by the Fed, but that’s all beyond the contained question I asked here.
James Oswald — who cites himself as a market monetarist but who seems to understand and adhere to much MMT thinking in his other comments and writings — said at Asymptosis:
There is no reason to think they [reserves] would not decrease back to the pre-IOR levels, at least over time, pushing around around 1.4 trillion dollars of high powered money into the economy and triggering significantly higher inflation.
This doesn’t seem to make any sense at all. (And I rather doubt that the Sumners and Beckworths of this world would agree with it.)
In (simplistic) theory, taking IOR nominally negative (the extreme case) would make banks want to instead hold physical currency, with its higher (zero) nominal return. Continuing the simplistic theory, that more-liquid money would be lent and spent more.
But:
1. There are significant costs and management headaches associated with holding currency — trucks, warehouses, security guards, all that rot.
2. It’s completely unclear why banks holding warehouses full of currency would have any incentive effects on borrowing and lending — hence real-economy purchases/velocity. Lenders and real-economy borrowers do their thing because they see valuable risk/return opportunities in the real economy. Changing the form of banks’ holdings will not affect that real-economy reality. Recent history: the QE trades — giving banks reserves in return for bonds — doesn’t seem to have had such an effect, if the massive runup in excess reserves is any testament.
3. Explaining #2: For banks, currency is (see #1) less liquid than reserves. They’re not carrying it in their pockets so they can buy gum at the corner store. They want to make loans; are they going to make them in cash?
4. Even if they did make the move to currency:
A) They couldn’t all do it; there’s not enough currency around.
B) The effect would be to reduce the amount of currency “in circulation” (it’s stuffed under banks’ mattresses), presumably prompting exactly the opposite of what market monetarists suggest:
a. Less real-economy spending/circulation/velocity and
b. Deflation — dollar bills would be harder to come by, so they’d be more valuable relative to real goods
At least in the discussions I’ve been perusing, this “currency” theory of “pushing” “more-liquid” money into circulation doesn’t make any sense. At all.
Market monetarists do seem to at least loosely and implicitly adhere to the (questionable) theory that people and businesses spend more because they hold money in more-liquid form — and they might even confute bank’s incentives and behavior with people’s incentives and behavior at times — but still this currency thinking is probably not a good or widely held market-monetarist theory. In any case it deserves unequivocal debunking.
So: Numerous cogent and convincing commenters agree that taking IOR to zero would have negligible first-order effects on lending and spending. And (invoking authority here) Mark Thoma agrees, in the post cited above:
It probably wouldn’t do much
But – considering the practical, workaday effects on the financial system such as those depicted in the currency fantasy above — Thoma links to an article by Todd Keister from the NY Fed. In short, IOR of zero would break a whole lot of financial entities’ business models. The gang at Mike Norman’s blog point to the problems already facing primary dealers, which could be (greatly?) exacerbated by a drop to zero. StreetEye on Angry Bear says that it would trash the main-street banking model. And etc. Various institutions would die or just withdraw their services/trades from the financial system.
The second- and third-order effects of such eventualities could have profound negative impacts on the real economy.
Which perhaps explains another thing I’ve been wondering about: why did the Fed institute IOR in the first place, and why did it do so when it did?
We can at least give the Fed credit for understanding the business models of various financial entities. When they saw interest rates heading toward zero, they instituted IOR to prevent the systemic lockup/breakdown described above.
IOW, nothing (much) to see here folks. Move along.
Sorry if I’m so dull that I had to go through all this to figure it out.
Make sense?
Cross-posted at Asymptosis.
BTW, I wish to thank you too, because this was a useful exercise for me too. The person who mentioned Thoma’s post on this question in particular deserves kudos because while I don’t agree with all of Thoma’s assertions that the inter-bank rate would hit 0% if not for IOR (there is sufficient frictional overhead in the system to keep it from doing such because the cost of lending money to another bank is non-zero), clearly there is much to think about there and if you accept some of Thoma’s assumptions as true, IOR makes more sense than it otherwise would make.
Note that my own thinking on this matter has moved, over the past six months, from “IOR is evil” to “IOR at current rates doesn’t make any real difference in the current economic environment other than as a way to plump bank balance sheets” as data comes in and is added to my model of how things work in today’s effective zero-bounds environment (in which many traditional mechanisms appear to not be functioning the same as they work in a normal environment). In a normal economy IOR *would* be evil because it gives incentive to hold money rather than lend it. But in an economy with a lack of viable lending opportunities and surplus capital stacked up in every crevice of the financial system due to prior attempts at quantitative easing… eh.
Bad Tux: “In a normal economy IOR *would* be evil because it gives incentive to hold money rather than lend it. But in an economy with a lack of viable lending opportunities and surplus capital stacked up in every crevice of the financial system”
Which is why I asked about the Swedish experience, where the central bank took the IOR negative. Aren’t they doing rather well?
The question of viable lending opportunities is in part systemic. Secretary Paulson claimed that the the bank bailout was so that the banks could start lending again. Stlll waiting. Meanwhile, the banks are doing rather well and handing out generous bonuses. Still waiting. . . .
If the Fed forced more lending by eliminating the IOR, where would the money lent go? Would it not return to the banks? Yes, some banks might be hurt at the expense of other banks, but what would be the systemic effects?
Min: “If the Fed forced more lending by eliminating the IOR, where would the money lent go? Would it not return to the banks?”
Yes, spending (or not spending, a.k.a. saving) does not affect the aggregate quantity of money in private hands. Spending equals income. But more spending on real goods — more velocity — *is* higher GDP.
IOW it’s the flows that matter, not (or more than) the stocks.
Bad Tux: “other than as a way to plump bank balance sheets”
If I were imputing Fed motives, based on my current understanding I’d suggest that this is an ancillary “benefit” for them to a policy that they really had to put in place.
I think it’s useful to frame the question in terms of the target rather than the tool. IOR is just the FFR maintenance tool of choice in an excess reserve environment post-QE world. It’s not more good or evil than the announcement/open marker operation tools used for rate maintenance pre-crisis.
So the relevant policy question remains “What should the Fed’s FFR target be?” just as it was pre-crisis when a different set of rate maintenance tools was used.
I’m with you on not much happening for the reason I stated earlier; that they would then adjust the balance sheets, because ending IOR signals the end of “extend and pretend.”
But not getting rid of it leaves the BIG problem, alluded to above by Min: it’s risk-free arbitrage and therefore distortionary, especially for firms that are supposed to be in the business of intermediation.
IOR is an idea only an economist could be dumb enough to love. The rest of the world calsl it what it is: stealing tax monies.