« Very good sentences | Main | Will he usher in a reign of futarchy? »
Interest on reserves, continued
I was intrigued by this passage, from Interfluidity:
Interest rates are, for the moment, excruciatingly low. But a subsidy to the banking system, once put into place, will be quite hard to dislodge. So, let's imagine that the Fed will pay interest on bank reserves in perpetuity, that it will pay such interest at or near the risk-free short-term interest rate, and that the expansion of the Fed's balance sheet is more or less permanent. How large a subsidy to the banking system do the interest payments on reserves represent? Some problems are arithmetically challenging, but not this one. The present value of a perpetual stream of market-rate interest payments is precisely the amount of the principal. Therefore, the present value of the Fed's de facto commitment to pay interest to banks on $800B of freshly created reserves is $800B. We fought and wailed and gnashed our teeth over potentially overpaying for TARP assets. Meanwhile, we are quietly allowing the Fed give away, as a direct, literal subsidy, more than the entire $700B that Paulson was allowed to play with. Note there is no question about this being an "investment": The interest payments that the Fed is now making to banks on its suddenly expanded balance sheet are not loans. The banks owe taxpayers absolutely nothing in return for this windfall.
I take that calculation to be a very rough one, and possibly an overstatement, but the point remains of interest. It also can be argued that interest on reserves is a bad signal for at least two reasons:
1. It signals the Fed fears being left holding the intra-day Fedwire bag if a major bank goes under, and
2. It signals the Fed thinks major banks need such a subsidy.
The cited post is interesting throughout.
Posted by Tyler Cowen on December 22, 2008 at 02:53 PM in Economics | Permalink
Comments
It is a great post at interfluidity.
On topic, the NPV of that amount of money might be = to the total amount, but it is doled out over infinity, so in more practical terms, it is far less than the huge number SW says. But still, his point about it was never their money so why give 'em interest is well taken.
Slightly off topic:
He gets closer to the issue about fiat/electronic money than most - that we can do whatever the hell we want with these numbers in our computers to make people do what we want them to do. Negative interest rates are possible with electronic money. The market is demanding them, why not give them what they want?
Totally off topic, but related to our Keynes discussions:
The post before is even better. How can there be overcapacity? A funny question.
Posted by: mickslam at Dec 22, 2008 4:27:31 PM
What interest pays Fed on the reserves now in the 0 interest rate environment?
Posted by: andy at Dec 22, 2008 4:50:56 PM
I would argue that is a tax reduction, rather.
Posted by: IWantCookieNow at Dec 22, 2008 7:11:20 PM
The Bank of Canada has been paying interest on reserves for over a decade (at 25 basis points below the overnight rate target). Yes, it is a tax reduction. It's a reduction in the tax rate on demand deposits (and all other deposits against which banks need or choose to keep reserves). Like all reductions in a tax rate, the loss in revenue depends on the elasticities. More reserves will be held than if there were no interest on reserves.
Posted by: Nick Rowe at Dec 22, 2008 7:56:23 PM
I was going to ask the same question as Andy. According to the Fed website, "The interest rates on required reserve balances and excess balances for the reserve maintenance periods ending December 17, 2008 were calculated using the upper bound of the target range of 0 to 0.25 percent established by the FOMC on December 16, 2008."
http://www.federalreserve.gov/monetarypolicy/reqresbalances.htm
I assume in plain English this means the Fed will pay 0.25% on required and excess reserve balances until there is another change in the Fed funds target, but I could be wrong. Needless to say, if this is accurate, I can't see how this won't screw with the markets: if a bank can earn 0.25% keeping reserves with the Fed and less than that if it lends to another bank through the Fed funds market, who would do the latter aside from Fannie and Freddie which apparently still don't earn interest on reserves.
Posted by: Ricardo at Dec 22, 2008 9:01:20 PM
Cutting interest rates is not going to stop deflation.
There really is only one way to do it and that is to accept inflation in everything else. And that's what its trying to do with super low rates.
Of course, with confidence shot, the effects aren't likely to be more than some refinanced mortgages, which will create a few paper shuffling jobs but will hardly move millions into productive jobs.
Even the billions handed out to banks are just getting invested in US government debt. So that's not helping either, except if you are collecting campaign contributions.
So the Fed should consider another way to put demand back into the economy.
If the Fed were to buy $500 billion in US government debt at the next auction, that would keep $500 billion in circulation and enlarge demand.
Would that be enough? If not, we have $10 trillion in debt to buy back.
How do we avoid inflating the total economy instead of just counteracting the effects of deflating assets? Peg the dollar to a stable commodity like gold and sell the debt on the open market when the dollar's value starts to drop.
Its not rocket surgery. Or am I missing something?
Posted by: Alan Brown at Dec 22, 2008 9:56:45 PM
Great post!
Would you like a Link Exchange with our new blog COMMON CENTS where we blog about the issues of the day??
http://www.commoncts.blogspot.com
Posted by: Steve at Dec 22, 2008 10:53:51 PM
i hope the author is buying bank stocks...
Posted by: kp at Dec 22, 2008 11:44:20 PM
"...but the point remains of interest". I had to read this line six times to fully appreciate its awesomeness.
Posted by: mpkomara at Dec 23, 2008 12:59:20 AM
The 800 bn in reserves is not a gift to the banks. As I understand it, the fed wanted to pay interest so that when it bought assets with cash, the cash would immediately be put in the reserves by the banks and wouldn't need to be sterilized by the Treasury selling bonds. The fed effectively gets sterilization without needing to go through an extra step.
Posted by: travis at Dec 23, 2008 1:31:30 AM
If the interest rate is zero, you can't calculate present value at all. That's a clue that the calculation of present value has turned into noise when the interest rate is small.
A penny a year doesn't imply a present value in the hundreds of billions, for example.
The underlying real world fact is that the time preference of money isn't reflected by interest rates any longer; without which present value mathematically doesn't have any meaning.
Posted by: Ron Hardin at Dec 23, 2008 5:29:16 AM
It is a tax benefit from a long term perspective only for the required reserves which is much less than 800bn usd. The last FED data shows 53bn usd required reserves and used to be around 40 bn usd before the crisis.
For the extra 750bn, the FED bought assets which receives interest from the market. So the asset side of FED balance sheet probably receives more interest than the reserves (liability side).
Posted by: Mariano at Dec 23, 2008 6:37:59 AM
But a subsidy to the banking system, once put into place, will be quite hard to dislodge.
Oh, don't worry, it will be very easy to dislodge. With bankruptcy comes all manner of dislodgements. (Is that a word?)
Strap on your helmets and put on your seatbelts. It's time for a wild combination of quantitative easing followed by Keynesian money explosions.
Posted by: K T Cat at Dec 23, 2008 11:10:39 AM
Alan-
Is there more deflation on the horizon? I had thought (with little to no expertise) that we've already seen deflation in most assets and they've stablized at or near their new found deflated prices.
What else will deflate or how much more will assets deflate?
Posted by: guy in the veal calf office at Dec 23, 2008 12:56:08 PM
