David Altig takes up the question with links back to Marginal Revolution, DeLong, as well as my post from yesterday. (Thanks, David!)
Rather than looking at it as what DeLong calls "Operation Twist", Altig opts for the simpler explanation that it will put a lower bound on the effective fed funds rate. That is, of course, the fundamental effect that this would have in any circumstance--crisis or not. It puts the Fed in as the residual buyer of the funds and thus establishes the floor.
The apparent lack of a (non-zero) lower bound on the funds rate was first noticed over a year ago when there were trades happening at zero percent. Here's what I said in August 2007:
When the Fed began discussing it more seriously in May 2008, I said:
So I am clearly on board with the stated reasoning behind the move. Plus, I think it's just a good policy to eliminate what is effectively a tax on reserves.
But I was struck by DeLong's comment about open market operations on the risk premium rather than on the liquidity premium. The more this drags on and the more we learn, the more I am coming to the conclusion (see here, for example) that this is a problem with the risk premium. Why else would the CP market freeze up despite the massive injections of liquidity, not to mention the CDS market? There seems to be a lot of liquidity out there, but it's not necessarily getting to where it needs to go.
And that got me wondering if paying interest on reserves might, as Tabarrok suggested, accomplish the goal of getting that liquidity where it needs to go in an Operation Twist sort of way. While the Fed is not yet targeting particular assets, we're treading very close to the kind of environment where that might be necessary. (Have you seen a T-bill rate lately?) Having the ability to pay interest on reserves would not be counter to that purpose, even if it wasn't the primary reason. Of course, it should also be noted that the paying of interest on reserves is a permanent change rather than a temporary one meant only for the crisis.
Paying interest on reserves is a good policy for a lot of reasons. The obvious ones and the ones that might still be a stretch--at least for now.
Rather than looking at it as what DeLong calls "Operation Twist", Altig opts for the simpler explanation that it will put a lower bound on the effective fed funds rate. That is, of course, the fundamental effect that this would have in any circumstance--crisis or not. It puts the Fed in as the residual buyer of the funds and thus establishes the floor.
The apparent lack of a (non-zero) lower bound on the funds rate was first noticed over a year ago when there were trades happening at zero percent. Here's what I said in August 2007:
So while I don't have a full and definitive explanation [for the zero percent transactions], it would seem that borrower risk is a factor, and the fact that these are excess reserves (which earn no interest) is also a factor. In that case, the low end of the range could stay low until the reserve picture gets back to normal.
When the Fed began discussing it more seriously in May 2008, I said:
I'll go on the record that this is a good idea. It will help to smooth out the recent fluctuations in the funds rate that garnered so much consternation at this blog among other places. It would prevent interest rate policy from getting in the way of policies for directly injecting liquidity into the financial markets by effectively keeping a floor on the funds rate even during a big injection of liquidity.
So I am clearly on board with the stated reasoning behind the move. Plus, I think it's just a good policy to eliminate what is effectively a tax on reserves.
But I was struck by DeLong's comment about open market operations on the risk premium rather than on the liquidity premium. The more this drags on and the more we learn, the more I am coming to the conclusion (see here, for example) that this is a problem with the risk premium. Why else would the CP market freeze up despite the massive injections of liquidity, not to mention the CDS market? There seems to be a lot of liquidity out there, but it's not necessarily getting to where it needs to go.
And that got me wondering if paying interest on reserves might, as Tabarrok suggested, accomplish the goal of getting that liquidity where it needs to go in an Operation Twist sort of way. While the Fed is not yet targeting particular assets, we're treading very close to the kind of environment where that might be necessary. (Have you seen a T-bill rate lately?) Having the ability to pay interest on reserves would not be counter to that purpose, even if it wasn't the primary reason. Of course, it should also be noted that the paying of interest on reserves is a permanent change rather than a temporary one meant only for the crisis.
Paying interest on reserves is a good policy for a lot of reasons. The obvious ones and the ones that might still be a stretch--at least for now.
I am not very good with the alphabet soup, so I always go back to the yield curve and (from signal processing days) compare it to a Gaussian Fourier transform, devoid of excess cyclic behavior.
The current curve drops like a rock at three months rather than having a smooth tail. This tells me that things happening on the three month or less basis are very noisy and not measured. I think our financial crisis verifies this.
So, an Operation Twist that just smooths out the tail, and drops long term rate down a bit would look like an economy running efficiently.
In 1961, they wants to invert the curve, make it less Gaussian. In this case, we are making it more Gaussian. It might work in that it forces us toward equilibrium, not away from it.
But, the sweet spot is very tiny, and whether the Fed can do this without inducing excess tail wagging remains to be seen.
Treasury should do it, and Treasury's purpose should be to minimize interest rate expenses. To the extent that Treasury can keep federal interest rate expenses minimizes it will force the rest of the economy back toward equilibrium.
"But, the sweet spot is very tiny, and whether the Fed can do this without inducing excess tail wagging remains to be seen."
Yes. I'm not even 100% convinced they're going to consciously try to do it. But could it work...a little bit... on the margin...? Maybe.
The more this drags on and the more we learn, the more I am coming to the conclusion (see here, for example) that this is a problem with the risk premium. Why else would the CP market freeze up despite the massive injections of liquidity?
Unfortunately crowding out needs to be considered a possibility too. It's possible that Treasury has been issuing too many T bills since mid-September. As long as people fear 1 - 2% losses on commercial paper backed accounts, the money market funds can not be in equilibrium until the returns on T bills are negative (which obviously would make the jobs of fund managers rather difficult).
Of course, money market fund insurance changes the balance a bit, but most likely the fact that T bill rates have risen above 0.5% even as commercial paper markets are frozen is a sign that policy intervention is pushing the market away from equilibrium, not towards it. I think Treasury's issuance goal should be to keep T bill rates at zero (and not worry if they go negative every now and then) until such time as the commercial paper market unfreezes.
If I were confident that T-bills and CP were good substitutes, I would be tend to agree. In a non-crisis environment T-bills and high quality CP are better substitutes. But right now, I'm not confident that a zero T-bill rate would unclog the CP market. As you said yourself, if people fear 1-2% losses on CP backed accounts an equilibrium would require the T-bill rate to be negative. How confident are you that the fear of loss will be limited to 1-2%?
I understand your point, but I'm inclined to go with DeLong in that this is a problem with the risk premium. The very idea of getting the Fed to address the risk premium rather than the liquidity premium is one that must be considered very carefully. And I'm not endorsing it at this point. But a year ago (a month ago?) I wouldn't have expected that we'd even be talking about it.
The Commercial Paper Funding Facility is a step in the right direction, but it is unclear to me yet if this will be enough.
I'm trying to work through the title of your post, which seems to have been neglected (or at least elided) throughout the post: "Why pay interest on excess reserves?"
And there's no good reason I can think of.
Suppose I have a bank with $100 billion in assets. I have a reserve requirement of, roughly (depending on the asset mix), $8 billion.
Recall: Banks are in business to make money, and they are uniquely positioned in the economy in that they can take advantage of the multiplier effect to earn returns.
Now, if I put up $8 billion to cover my Reserve Requirement, you might want to make the argument that I should be paid interest on that. (Personally, I consider it a cost of doing business, and charge accordingly when I'm lending, but that's a learned behavior, and one that can change.) And, especially if you're arguing that you will change the behavior in the parenthetic on a marginal basis and generate more lending, I'm sympathetic. "Desperate times call for desperate measures," and this permanent change may be as "good" an idea as raising the Discount Rate above the FedFunds Rate and last for appreciably less time. Worth a try, and a better idea than waiting four weeks to have Goldman Sachs people buying their way into cushy jobs when the leave the current one in three months.
But suppose I put up $12B? Or $16B? Then I'm leaving capital on the table, not making it work. I'm specifically not trying to generate economic activity; I'm just subsidizing my bottom line.
Should I be rewarded for that?
You want to see what happens when you pay interest on required reserves? Worth a try; I know which way I'm betting, but I wouldn't, er, bet the house on it, and would like to be proved wrong. (Think EITC, where, what, about 70% the flows have actually gone to the employer subsidization, but the program is considered a success.)
But rewarding banks for not managing their Treasury to its earning potential? Let's just say that if they really cannot find a customer worth lending the money to at a fair market rate, then outright nationalisation of the entity is inevitable anyway. Why prolong the inevitable, and lose money (in the form of interest payments) in the process?
Ken,
I did not neglect the title of the post. I am, however, implicitly assuming that it is understood that you have to pay interest on all reserves (including excess) in order to have the effect of putting a floor on the funds rate--which is the main objective. David made the same implicit assumption in his post of a similar title. Neither one of us started from scratch to explain all the details. But since you ask...
Here is the FAQ from the NY Fed: http://www.newyorkfed.org/markets/ior_faq.html
Here is a paper from the KC Fed that I cited on this blog in 2006:
http://www.kc.frb.org/PUBLICAT/ECONREV/PDF/2q97WEIN.pdf
Reserve requirements operate on the quantity of reserves, not their price. Paying interest only on required reserves would only transfer money from the Fed to the banks because that money has to be held on reserve anyway. It would not change incentives, nor would it put a floor on the funds rate. Paying interest on all reserves puts the Fed in the position of the residual buyer (for as much as the market is willing to supply) and establishes the floor for the funds rate.
Paying interest on all reserves makes the reserve requirement unnecessary, as the KC Fed paper explains. If you want them to hold more reserves, raise the interest rate. Now, that may not be on the horizon for a while--it's just not the right time to talk about that yet--but it may be the next step in the evolution of U.S. monetary policy.
As for your concern that this encourages banks to forgo profitable investments, keep in mind that these are not long term funds. These are overnight reserve balances trading at very low rates. According to the NY Fed, the initial rate paid on reserves will be the lowest target fed funds rate for the maintenance period less 75 basis points. Currently that would be 0.75%.
I'm not worried that this will siphon off funds that would go to productive investment. But I'm pretty sure that it will put a floor on the rate and make the trading desk's job easier.
The factor your reply omits is that Treasury has been increasing the bills held by the public at an average rate of 10% (or $120 billion) per week since mid-September. (You can check Treasury's data yourself at http://www.fms.treas.gov/dts/index.html.) Think about this for a second. Is there any reasonable scenario in which numbers this big would not cause crowding out of private issues?
The risk premium argument is only convincing, when there aren't massive supply changes taking place at the same time.
Yes, supply has increased massively, but so has demand. In fact, they can barely keep up with the demand. I follow your argument, but (a) it does not imply that the risk premium isn't a problem and (b) if the Treasury slowed the flow of T-bills to the market and the risk premium really is the problem it wouldn't fix anything and would probably make things worse.
Did you read Tyler Cowen's post of a reader's comment?
http://www.marginalrevolution.com/marginalrevolution/2008/10/david-murphy-ma.html
But this is precisely my point. By increasing supply to meet demand, Treasury is disrupting the market forces that would cause money to flow back into commercial paper.
This issue should be understood as a choice: Is government going to protect the repo and money markets or is government going to protect the banking system? It should be obvious at this point that you cannot protect both. By preventing the closure of Treasury money market funds and the disruption of the repo market, the government has chosen to close interbank markets -- that is the direct implication of the quantity movements we see in T bills. (Since the resulting reduction in sales of CP end up as loans on bank balance sheets, loans that the banks definitely cannot afford to make right now, this policy will cause every money center bank to be undercapitalized and unwilling to loan on interbank markets.)
While apparently the government imagines that it will be capable of taking over the functions of the banking system, I don't see any signs that anyone is considering how to rebuild a private banking system after the government takes over its interal functions.
The alternative is of course also bad: closure of Treasury money market funds and a big mess in repo markets. On the other hand, we know from centuries of experience that a banking system can function without money market funds and without repo. Personally, I think the choice should be to go with the option we have reason to believe is viable historically and not experiment with the government take over of interbank markets. (I don't know all the connections, but I doubt that money market funds and repo can function in the absence of a banking system -- and I also doubt that the government has the ability to take over interbank markets effectively.)
Addendum: "Since the resulting reduction in sales of CP end up as loans on bank balance sheets ..."
This is due to the backstops CP issuers have with banks. Of course, the fact that many banks rely on CP funding themselves is equally important, since protecting Treasury money market funds by issuing T bills to meet demand is a pretty direct way of bankrupting the banks.
Ok, let’s stop being so naive to what is really going on here. Where do you think the Fed earns profits to pay this interest on reserves?! …mostly from interest on treasuries which is paid by our government, and the government gets the money from YOU through taxing. The Fed has to return all profits after paying operating expenses and dividends to member banks back to the Treasury. Now less of this profit will go back to the Treasury and instead go directly into private banks. Only an idiot would believe the press release from the Fed that this is being done to help set a lower bound on the Fed Funds Rate. In a matter of weeks the spread was quickly increased from 10/75 basis points below the Fed Funds Rate for required and excess reserves respectively to being directly set by the Fed to the same as the Fed Funds Rate of 25 basis points even though the official target is 0-25 basis points. If the current target is 0-25 basis points why would you be worried about a floor!!!! Why would you set it at the upper bound of your target range!!!! HELLO ANYBODY HOME?? HE IS PAYING THIS INTEREST AS RANSOM FOR THE BANKS TO HOLD THE 800 BILLION IN NEW MONEY THE FED CREATED WHEN IT BOUGHT ALL THE TREASURIES AND MORTGAGE SECURITIES AND IS USING YOUR MONEY TO DO IT. This is only going to get more expensive from the over 2 billion a year currently being paid, as interest rates on reserves will only go up with the Fed funds rate to persuade banks not to loan out money and dump it into the market. It’s so obvious Bernanke is a liar. The Fed said they need this new program right now and not in 2011 and the original wording of the bill never intended for interest to be paid on excess reserves, only required reserves, which is why the original estimated cost was only a fraction of what we are paying now. Within days of the Oct. 3 2008 passage of the act allowing interest payments on all reserves the Fed began the massive increase in the money supply to double it to over 1.6 trillion in a few months. This coordinated perfectly with a massive rise in excess reserves to the tune of exactly the 800 billion of newly created money during the same time period. Excess reserves have historically been virtually 0. Now all of a sudden they are 800 BILLION and this is just a cute anomaly!!! The only other time in modern history there has been a notable rise was after 9-11 and this was only about 20 billion, which subsided the next month. Americans now officially pay a banker tax to private banks, and if they don’t they will get hyperinflation. If you have seen all this obvious information and still hold on to dollars you are a moron. This system is coming to an end and this is just a cheap circus trick by the Fed to keep it going a little longer. The spending of government is only increasing to record levels and nobody wants our treasuries anymore. We are about 1 or 2 moves from checkmate. The only way to sop up this money is for the Fed to sell treasuries and securities, I think we all know that ship has sailed. Good luck selling treasuries when the government is conducting record deficit spending, pumping out new treasuries like no tomorrow and China has had enough. Good luck selling junk toxic mortgages. In the meantime make sure you pay your banker tax which will quickly swell to tens of billions of dollars of your taxpayer dollars being funneled directly into private banks as pure profit. Bernanke is a agent of private banking with the private Federal Reserve being the biggest private bank of them all. The private banks and their king, Bernanke, are merely looking out for themselves at your expense. And now Bernanke has set up a situation in which you have no choice but to pay or suffer a hyperinflation holocaust. One man, one private banker, now has a red button to implode our economy any time he wishes, and the only way to keep him from pushing the button is paying the ransom.