A few notes on interest on reserves

It’s been fascinating to see the topic of interest on reserves (IOR) all over the news.  I was watching a CNBC video sent to me by Mike Sandifer, and they discussed two interesting tidbits around the 5 minute mark.  First, that rumors the Fed would eliminate IOR triggered a big stock market rally the day before.  And second, they actually discussed the idea of negative interest rates on reserves.  I was gratified by the stock market reaction to the rumor that IOR would be eliminated (assuming the market was in fact reacting to this rumor, and I don’t know if it was) because market responses to news are the gold standard of causality.  Even if the stock market is wrong, it might nevertheless be right for the usual “Tinkerbell” reasons.  (If I think I can fly, then I can fly.)  If the stock market thinks a Fed action is very bullish, and rises sharply in response, that market reaction can trigger more investment even if based on an erroneous theory of monetary economics

I was also pleased to see the elimination of interest on reserves mentioned in these two excellent posts by Joe Gagnon and Bruce Bartlett.

And finally, I few days ago I mentioned how Jim Hamilton and Robert Hall had noticed the contractionary nature of IOR quite early on.  Yesterday I ran across a David Beckworth post that was one of the earliest and best posts on the subject.  Not only did David notice the contractionary impact, but he also saw the analogy to 1937, at a time (October 2008) when the vast majority of macroeconomists were still completely oblivious to what was going on. 

Monetary policy can affect either the supply or demand for money.  In the short run, monetary policy tools are generally used to change the supply of base money.  Indeed most textbooks are only able to come up with a single example of the Fed targeting the demand for base money in order to change the stance of monetary policy;  the 1936-37 decision to double reserve requirements.  Futures textbooks will almost certainly mention the October 6, 2008 IOR decision as a second, and even more foolish, example of raising the demand for base money at an inappropriate time.  Shame on the economics profession for not calling out the Fed at the time. 

I like to point out that 2009 saw the biggest drop in NGDP since 1938.  But as of October 2008, the decline was still not expected to be that large.  So David’s reference to the 1937 example proved to be not only accurate, but also prophetic.  Here’s how he modestly concluded his post:

Pardon me for being skeptical, but from what I can see this approach has only encouraged banks to hoard more excess reserves. I may be missing something here–and please let me know if I am–but it seems like we are making the same policy mistakes that were made in 1936-1937.

No David, you didn’t miss anything, but the rest of the profession sure did.


Tags:

 
 
 

49 Responses to “A few notes on interest on reserves”

  1. Gravatar of Jon Jon
    24. July 2010 at 09:50

    Its interesting to put IOR in context of what was claimed when it first came into being:

    1) Fed to Congress, we need to pay interest-on-reserves to compensate banks for the ‘tax’ we impose on them by having reserve requirements. We’re so backwards, the Canadians have no reserve requirement, and the ECB already does this.

    Congress passes measure, but not to take effect until late 2011.

    2) Crisis. Fed creates new facilities that are like the discount window but are not called that. The discount window sounds so scary! The Fed sells t-bills to sterilize one-for-one. Fed sells down all of t-bill portfolio. Commentators begin wondering about about the Fed being out of ammunition. Many panic that and inflate this to be about monetary policy being stuck. Informed commentators know the question is about ‘liquidity’ interventions being sterile with respect to monetary policy. Most economics reporters in the MSM show themselves to be dumb as dirt.

    Fed is already out of sterile liquidity ammunition, actually. Selling notes and bonds to fund liquidity operations is not Fed policy (because of interest-rate losses); plus, much of the residual treasury portfolio has already been posted as or is intended to be posted as collateral.

    Fed and Treasury concoct the supplementary funding program. Tbills are sold by the treasury, proceeds held on deposit. Liquidity programs continue to expand and remain, by design, sterile with respect to macroeconomic policy.

    Fed & Treasury to Congress: let us pay IOR now. Then we can replace the treasury operation. Congress includes this in their panic legislation.

    Couple months pass…

    Fed to Public: we need to keep interest on reserves at a positive marginal rate otherwise the broker-dealers will leave the t-bill market, but trust us Policy is REEALLLLYY loose!

    Uhh yeah Mr. Chairman, but wasn’t the SFP there to sterilize the base with respect to policy. Tell me again how the new situation is different?

  2. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    24. July 2010 at 12:30

    Scott,

    1936-37 decision was very harmful, while current 25bps IOR is just a minor detail. This should be obvious if you measure the tightness of a policy stance by a difference between the Wicksellian natural interest rate on reserves and the actual interest rate on reserves. In 1936-37 increased reserve requirements have created a huge gap of something like 5 percentage points between the very negative natural rate and the actual rate on reserves. Current 25bps IOR is just peanuts compared to 1936.

  3. Gravatar of John Hall John Hall
    24. July 2010 at 13:47

    I think the folks at CNBC were overestimating how many people thought the Fed would drop the IOR. Most Wall Street economists thought it was a low probability he would do that in the testimony.

  4. Gravatar of scott sumner scott sumner
    24. July 2010 at 16:53

    Jon, Very good points. I’d add that the quickest way out of zero interest rates is the Australian way–fast NGDP growth expectations.

    123, Actually the effects on rates were almost identical, In 1936-37-38, short term rates rose from about 10 basis points to 35 and then fell to 10 again. In other words, a measely 25 basis points can have a huge effect if it creates deflationary expectations.

    John Hall, You are probably right, but I’m more interested in the stock market’s reaction, rather than whether the rumor was likely to be true. If it was believed, and if the market reacted, that’s news. Of course it is very possible there was no market reaction, I’d like to see the evidence they have.

  5. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    25. July 2010 at 01:14

    Scott, effect on rates is not important. Effect on the gap between the natural rate and actual rate is important. The doubling of reserve requirement in 1936-37 could be neutralized by negative 5% IOR. Current 25bps IOR could be neutralized by a tiny cut of 25bps.

  6. Gravatar of Lee Kelly Lee Kelly
    25. July 2010 at 07:56

    If the quantity of base money that banks wish to hold is equal to or greater than what is mandated by the Fed, then any increase in the rate of interest earned on reserves (normally zero) will have similar consequences to setting the reserve ratio slightly above whatever quantity of base banks wish to hold.

    Or at least that is how it seems to me. But why would the Fed do that!? If I didn’t know better, I wouldn’t have believed it.

  7. Gravatar of Lord Lord
    25. July 2010 at 09:43

    I don’t think it was contractionary when first implemented. At the time banks weren’t lending to each other much less anyone else and the intent was to direct reserves to the Fed who would lend through their various facilities. The problem now is the Fed who have decided to keep their powder dry. Is there any reason to think that ending this would not cause them to be less accommodative in other ways? They are happy with things as they are, and that is the problem.

  8. Gravatar of ssumner ssumner
    25. July 2010 at 18:45

    123, I have no idea where you got the 5% number. And it wouldn’t neutralize it, it would make it twice as bad.

    Lee, That’s what I’d like to know, what was the Fed thinking.

    Lord, It’s not contractionary because it discourages lending, it is contractionary because it encourage money hoarding.

  9. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    26. July 2010 at 03:04

    Scott, it was a negative 5%. Doubling of reserve requirement has sharply increased demand for reserves in 1936-37, it would take a massive negative IOR to reduce demand for reserves back to the original point. 5% is just a guess, I am willing to agree with any figure between minus 3% and minus 10%.

  10. Gravatar of ssumner ssumner
    26. July 2010 at 10:52

    123, Sorry, That was my mistake. But in general I don’t see the need for big negative IORs. Cash yields zero, which means T-bill yields can never fall below the cost of storing cash. So a negative IOR merely needs to be a bit below T-bill yeilds. Perhaps minus 1%?

    And that’s in a static model. If the policy succeeds and boost inflation expectations, then perhaps even less is needed. I think the honest answer is we can never provide a single number in isolation. It depends on other aspects of monetary policy, and how the action would change the public’s expectations.

  11. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    27. July 2010 at 02:51

    I am not discussing negative IORs in order to support it as a policy. My purpose was to find a way to compare a colossal 1936-37 error to the current 25bps IOR policy. And the only way to do that is to ask a question what amount of negative IOR could have neutralized the negative consequences of the doubled reserve requirements in a context of a short term static model And the answer I am getting is that 25bps IOR is insignificant compared to massive negative IOR that was needed in 1936. Of course such negative IOR was never feasible because of the cash limit.

  12. Gravatar of Scott Sumner Scott Sumner
    27. July 2010 at 06:09

    123, I understand what you were saying, but I don’t buy the model that comes up with “much more than 0.25%” There is no way a person can come up with any number for the fed funds target, money supply, or IOR needed to produce any given result, without knowing the context. It will all depend on how that action affects expectations of future policy. A tiny change can have a huge impact under the right circumstances.

    BTW, I actually don’t think the higher RR actually was the main factor in the 1937 recession. Rather I bring that example up because other economists do. So in that case, why not outrage over the IOR program.

  13. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    30. July 2010 at 05:15

    Scott,
    What was the main factor in 1937 then?

  14. Gravatar of ssumner ssumner
    31. July 2010 at 07:14

    123, In order of imprtance:

    1. Gold hoarding in late 1937 due to dollar devaluation fears, it made the recession dramtically worse, a la August 2008.

    2. Much higher wages following a doubling of union strength following the passage of the Wagner Act.

    3. Higher reserve requirements and gold sterilization in early 1937.

  15. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    31. July 2010 at 12:09

    Scott,
    I guess 1 was caused by 3.

  16. Gravatar of ssumner ssumner
    1. August 2010 at 06:48

    123, Yes and no. The market’s did not react strongly to the RR increases, because it was duirng a period of inflation, and was viewed as not credible enough to stop the inflation. But when economic conditions changed unexpectedly later in the year, the same policy stance became more contractionary. Then as the economy slowed, investors began to fear another FDR “shot in the arm” of dollar devaluation. Interesting, unlike 1933, it did not happen. But the fear of devaluation led to massive gold hoarding, which turned a mild recession into a severe recession. BTW, I have a recent post that has a quotation from the Fed minutes of 1937.

    So basically you are right, but there was no rational reason for the market to expect that effect as of the early part of 1937.

  17. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    2. August 2010 at 05:41

    Scott, the market got it wrong. In early 1937 there was no certainty that increased reserve requirements will cause recession, however the economy became much more fragile. So risk premiums should have increased, driving stock prices down. I have never believed that markets process risk related information efficiently.

  18. Gravatar of ssumner ssumner
    3. August 2010 at 05:55

    123, Yes, they got it wrong ex post, but ex ante the market may have been rational. The experts also missed the boat on the RR increases. So if you are going to argue the markets are ineffeicent, and yet the exports were also wrong, then who is your benchmark for market effiecieny? God?

  19. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    4. August 2010 at 04:22

    I agree that markets are usually better than experts. But according to Fama’s efficient market hypothesis, market should be God-like.

  20. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    4. August 2010 at 11:30

    Regarding interest on reserves, I have a new post that disagrees with your take on opportunity costs:
    http://themoneydemand.blogspot.com/2010/08/should-fed-stop-paying-interest-on.html

  21. Gravatar of ssumner ssumner
    6. August 2010 at 06:36

    123, That’s confusing perfect foresight with Ratex. Fama definitely is fine with markets making lots of mistakes, ex post.

    123, If there are liquidity services provided, then the minimum op. cost is the IOR. The actual op. cost may be higher. But right now the marginal dollar of ERs provides little liquidity services.

    I don’t agree with your criticism of Beckworth and Hamilton, because there is much more to monetary policy than the fed funds target. You can make monetary policy easier or tighter without shifting the fed funds target.

  22. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    6. August 2010 at 12:54

    1. EMH. I did not say markets should have perfect ex-post foresight according to EMH. But EMH says that markets should know the perfect ex-ante distribution of probabilities according to all publicly available information (semi-strong EMH), or according to all public and private information (strong EMH)

    2. Opportunity cost of reserves. Fed has flooded the system with reserves. The actual quantity of reserves is much larger than any plausible requirement, so at this time the opportunity cost of reserves is equal to IOR. Prior to the policy of IOR, reserves were very scarce. Holders of reserves were engaging in the arbitrage in fed funds market, and in equilibrium opportunity cost of reserves was equal to fed funds rate. So IOR had no adverse effect on the opportunity cost of reserves (if we hold fed funds rate constant).

    3. Beckworth and Hamilton. The key issue in monetary policy is the Fed’s reaction function, according to which Fed sets the fed funds rate target. The October 6, 2008 IOR announcement did not contain any negative information about the fed funds rate reaction function. Even more, in the announcement the Fed Board was very explicit about the main reason for the IOR program: “The payment of interest on excess reserve balances will give the Federal Reserve greater scope to use its lending programs to address conditions in credit markets while also maintaining the federal funds rate close to the target established by the Federal Open Market Committee.”. By starting IOR, the Fed board could implement more extensive credit easing operations without violating FOMC fed funds rate target. Indeed, in the same announcement the Fed board announced the massive expansion of credit easing. On October 8, 2008 FOMC together with other foreign central banks announced coordinated unscheduled cut in the Fed funds rate. Beckworth and Hamilton have chosen the wrong target for criticism. FOMC should have cut rates even faster, as situation was deteriorating very rapidly, the Fed board should have announced even larger steps in credit support.

  23. Gravatar of ssumner ssumner
    6. August 2010 at 17:02

    123, Your first point gets into some deep philosophical issues. We never directly observe probablity distributions in most real world economics problems. They are subjective. So you can’t really test whether they were accurate, ex ante. If the universe is deterministic, then of course the ex ante probability of the actual outcome is exactly one. But in the real world subjective probabilities are reflective of our ignorance. The EMH is certainly consistent with the public being profoundly ignorant of what will happen. What are the odds that a terrorist will relaease a dangerous virus in the next three years? No one knows. That doesn’t make markets inefficient

    Regarding your second point, we can’t hold the fed funds rate constant. The reason the IOR is now the op. cost is because the fed funds rate has fallen to a level below the IOR.

    3. You said;

    “The key issue in monetary policy is the Fed’s reaction function, according to which Fed sets the fed funds rate target.”

    I don’t agree at all. This view suggests that a low interest rate trajectory is more expansionary than a high interest rate trajectory. But just the opposite is true. And how do you know that the institution of IOR did not affect the expected path of interest rates over time? Suppose that w/o IOR we would have had hyperinflation. In that case the decision to do IOR almost certainly led to far lower interest rate expectations, and much tighter monetary policy.

    The market fell sharply during the first 10 days of October, every single day.

    I do agree with you on one point. If the Fed had cut the fed funds rate and IOR to zero immediately, then the IOR policy would not have done any harm. But that is the same as having no IOR policy.

  24. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    7. August 2010 at 05:16

    Scott, if there is no way to test if ex-ante probability distributions are correct, then EMH is not testable. Yet there is evidence that these ex-ante distributions are not efficient with respect to publicly available macroeconomic information. For example, Shiller has found that stockmarket fluctuates more than variation of fundamentals suggests.

    Regarding the second point, the most important thing is that both now, and before the crisis, the main driver of opportunity cost of reserves is FOMC. 25 bps IOR is within the bounds set by FOMC. Before the crisis, scarcity of reserves drove the opportunity cost of reserves up to the fed funds rate, even though IOR was zero. IOR is just a technical detail, and according to reasons suggested by Milton Friedman in 1959, it is a detail that improves the functioning of markets.

    Regarding Fed’s reaction function, nothing I said contradicts the fact that a low interest rate trajectory is more expansionary than a high interest rate trajectory. The reaction function that targets higher long term inflation will produce higher interest rate trajectory. But a reaction function that does not cut rates fast enough when the natural interest rate declines will produce excess macroeconomic volatility in a form of a deflationary shock. This is what happened in the first 10 days of October, when the financial panic (as measured by the TED spread) was exploding. By the way, TED spread reacted favourably to the Oct. 6 IOR announcement. Stockmarket was dropping because stocmarket started to understand what TED spread is, and it stopped falling when Bernanke’s credit easing started reversing the TED spread.

    Without IOR we would have even more deflationary pressure, as the size of Fed’s balance sheet that is consistent with the Fed funds rate target set by FOMC would not allow for significant credit easing.

  25. Gravatar of The Federal Reserve: “erring on the side of laxity”? « Left Outside The Federal Reserve: “erring on the side of laxity”? « Left Outside
    13. August 2010 at 01:01

    [...] Lots of other things impact monetary policy apart from the headline interest rate, for example the Federal Reserve is currently paying interest on its reserves and this is a contractionary policy which pulls in the opposite direction to low interest rates. [...]

  26. Gravatar of Scott Sumner Scott Sumner
    22. August 2010 at 14:06

    123, The EMH is testable. The test is whether excess returns earned by investors are serially correlated. Studies suggest that the EMH is approaximately true (not exactly), that mutual funds that do well one year, revert back to trend the next year. Shiller-type searches for anomalies are no way to test the EMH.

    Rregarding the elimination of the IOR, I agree that if it was accompanied by a huge drop in the monetary base, it might be contractionary. But I am pretty sure the stock market would react positively to an announcement by the Fed that the IOR was going to be eliminated, because they would (correctly) interpret it is a signal that the Fed was determined to ease further.

    There is much more to monetary policy than the TED spread. The key is expectations of NGDP growth, which fell from mid-2008 to March 2009.

  27. Gravatar of malavel malavel
    22. August 2010 at 14:14

    Luck versus Skill in Mutual Fund Performance

    http://www.dimensional.com/famafrench/2009/11/luck-versus-skill-in-mutual-fund-performance-1.html

  28. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    22. August 2010 at 15:14

    Scott,
    The test you mentioned measures the efficiency of the mutual fund industry, and not the efficiency of the stockmarket. According to Shiller, sometimes stockmarket is too cheap, sometimes it is too expensive. Since most mutual funds are 99% invested all the time, it is no wonder serial correlation tests do not pick up this anomaly.

    Because fed funds rate is 0-0.25%, eliminating IOR now would not cause the drop in the monetary base, because interest rates will still be on the target set by FOMC. This is why I supported 0 IOR as one of the policy options:
    http://themoneydemand.blogspot.com/2010/08/should-fed-stop-paying-interest-on.html
    But if Fed funds rate was 1%, zero IOR would crash both the monetary base and NGDP expectations.

    Bank of England started IOR in May 2006. Where was the NGDP crash? Why Fed was different?

    NGDP expectations were falling from June 2008 to March 2009.
    But the speed of the fall was highest from September 15 2008 to October 10 2008, and the Ted spread was the key indicator to watch if you were interested in NGDP expectations during that time. Stockmarket was reacting to falling NGDP expectations with a lag.

  29. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    22. August 2010 at 15:19

    Scott, I have excerpted to disturbing quotes by Kocherlakota here:
    http://themoneydemand.blogspot.com/2010/08/really-good-links-deficit-economics.html
    Krugman is angry too, but Stephen Williamson is very satisfied.

  30. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    22. August 2010 at 15:23

    Scott,
    at David Beckwort’s blog I have found a very useful monthly NGDP data series:
    http://www.macroadvisers.com/content/MA_Monthly_GDP_Index.xls

  31. Gravatar of Scott Sumner Scott Sumner
    23. August 2010 at 16:20

    Malavel, Yes, I was relying on that paper in my comments.

    123, There is no law against market-timing funds, is there? If they could do it successfully, I am sure they would. Many mutual funds are not fully invested, at least many I have owned. (Which angers me.)

    You said:

    “But if Fed funds rate was 1%, zero IOR would crash both the monetary base and NGDP expectations.”

    I don’t follow. This is exactly what happened in 2003, and things didn’t crash. We had a 1% ffr, and a zero IOR.

    You said:

    “Bank of England started IOR in May 2006. Where was the NGDP crash? Why Fed was different?”

    I have never said IOR is necessarily going to result in a crash. But it can be bearish if it leads to expectations that future QE won’t work. Nobody in Britain in 2006 was worried about future QE, and hence I would not have expected a big impact. I do understand that in normal times IOR can work well, and has.

    You said:

    “Stockmarket was reacting to falling NGDP expectations with a lag.”

    Then why didn’t investors get rich by selling S&P futures immediately after NGDP expectations fell?

    123#2, Oddly, I just wrote a post that includes the Kocherlakota quotation, but won’t post it until tomorrow. Ditto for the Beckworth post.

    I love the George Bush quotation, where did you get it?

  32. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    24. August 2010 at 08:10

    Scott, you said:
    “There is no law against market-timing funds, is there? If they could do it successfully, I am sure they would. Many mutual funds are not fully invested, at least many I have owned. (Which angers me.)”
    You should be angry, as mutual funds hold more cash when stocks are cheap, and they are more fully invested when stocks are more expensive.
    The amount of market timing activity that we are seeing is tiny compared to what is required for efficiency of aggregate level of stock market value.

    You said:
    “I don’t follow. This is exactly what happened in 2003, and things didn’t crash. We had a 1% ffr, and a zero IOR.”
    If FOMC decided that 1% ffr is needed (with zero IOR), Fed would have to sharply contract the monetary base in order to hit the ffr target. In 2003 1% ffr was implemented without such a sharp contraction in monetary base.

    You said:
    “Then why didn’t investors get rich by selling S&P futures immediately after NGDP expectations fell?”
    Those investors who like Krugman noticed that Ted spread is incompatible with the valuation of stock market got richer, but this process was not instantaneous.

    You said:
    “I have never said IOR is necessarily going to result in a crash. But it can be bearish if it leads to expectations that future QE won’t work. ”
    IOR shouldn’t lead to expectations that future QE won’t work. QE with 25bps IOR works 95% as well as QE with zero IOR, just like QE with -25bps negative IOR would work a little bit better than QE with zero IOR. As a plus, IOR increases expectations that QE/credit easing will be started before zero bound is hit.
    QE fan Roger Farmer says IOR is bullish:
    http://themoneydemand.blogspot.com/2010/08/really-good-links-interest-on-reserves.html

    You said:
    “I love the George Bush quotation, where did you get it?”
    The source is a book called “Too big to fail” by Sorkin. Quotation sounds nicer that it really is, as Bush was probably referring to TARP.

  33. Gravatar of ssumner ssumner
    24. August 2010 at 14:36

    123, Yes, I am angry because my mutual funds held cash. They can’t time the market (as you say), and that’s because the EMH is true, or truish.

    Good for Krugman, but there aren’t enough people like him to make the EMH false. Buffett wasn’t smart enough to sell in 2008, and he’s pretty smart.

    We are going around in circles on the IOR. We both agree it is contractionary for any given monetary base. You say if they hadn’t done the IOR, they wouldn’t have done the QE. I say the IOR prevent future QE (in 2009) from having the desired effect.

    I don’t agree with your “work 95% as well” comment. We can’t know that. There are knife edge equilibria in interest rate targeting. For instance, if you peg nominal rates, the difference between a peg of 4.25% and 4.5% can be the difference between the price level spirally toward zero or infinity. It all depends on where the Walrasian equilibrium rate is. I’m not saying it was definitely a big factor, but it might have been.

    Thanks for the Bush info–I’m not surprised.

  34. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    25. August 2010 at 03:41

    Scott, it is worse than that. Mutual fund cash is a good contrarian indicator, by accumulating cash when stocks are going down, mutual funds are creating the excess volatility that according to Robert Shiller is a sign stock market is not 100% efficient.

    TED spread – S&P 500 disconnect in September-October 2008 did not last long, as arbitrage activity steadily increased. In late September the information processing lag was something like a week, but by mid-October TED spread has ceased to be a reliable leading indicator for stocks.

    All markets have useful information about NGDP expectations, starting from the market for treasuries, ending with the market for the orange juice in the local Wal-Mart (that’s why analysts are paying money for Wal-Mart parking lot occupancy data from space). Yet at the turning points different people have different NGDP expectations, and you could not get an accurate picture of economywide NGDP expectations just by looking at what stocks are doing. The best indicator was TED spread, and it reacted positively to IOR announcement.

  35. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    25. August 2010 at 05:09

    Scott, you said:
    “We are going around in circles on the IOR. We both agree it is contractionary for any given monetary base. You say if they hadn’t done the IOR, they wouldn’t have done the QE. I say the IOR prevent future QE (in 2009) from having the desired effect.”

    I feel that IOR is a very important topic, as there should be some deeper issue behind the disagreement (Woodford, Farmer and Bernanke vs you, Hamilton and Beckworth), but I still don’t get what that deeper issue might be.

    At this time the Fed can cut IOR to zero without decreasing the monetary base, that’s why I think it is one of the good policy options. But cutting IOR to negative -25bps would be a bad idea unless FOMC cuts ffr to negative -25bps too. And anyway, holding monetary base constant was a wrong assumption before 2009.
    When IOR was announced, nobody had any idea what is the lower bound of ffr (0, 0.25 or 0.5 or 1.0 as ECB has done), what would be the ultimate size of QE etc. So I don’t see how IOR announcement could have communicated anything negative about the future QE. In the end the Fed has stopped at 0.25 in December 2008 in order to preserve the money market fund industry (as Andy Harless explains here: http://blog.andyharless.com/2010/08/what-will-happen-if-fed-stops-paying.html), but there is always a possibility to reconsider.

    You said:
    “I don’t agree with your “work 95% as well” comment. We can’t know that. There are knife edge equilibria in interest rate targeting. For instance, if you peg nominal rates, the difference between a peg of 4.25% and 4.5% can be the difference between the price level spirally toward zero or infinity. It all depends on where the Walrasian equilibrium rate is. I’m not saying it was definitely a big factor, but it might have been.”
    These knife edge equilibria apply when there is a permanent peg of interest rates, and they are unimportant with flexible central bank reaction function. With flexible monetary policy and with interest rate flexibility, 0.25 mistakes cause very little damage. ECB has de facto tightened by 50 bps this year, and so far markets are not signalling anything worse for the eurozone compared with the USA, as ECB can always reopen the liquidity gates.
    And I would say that only -4% ffr would have been too stimulative in 2009, but of course -4% ffr is not possible because of zero yield on cash.

  36. Gravatar of ssumner ssumner
    25. August 2010 at 16:56

    123, You said;

    “Scott, it is worse than that. Mutual fund cash is a good contrarian indicator”

    Never say something is a good indicator. Say it has been a good indicator. There is never any reason to think something will continue to be a good indicator, as the market is unforcastable.

    NGDP is different, I agree that Walmart parking might be a predictor.

    I don’t buy the argument about preserving the MMMF industry. But that’s really a symptom of the deflationary monetary policy. Which does suggest that there are better ways to solve deflation that trying to drive IOR ever lower. Obviously setting a higher NGDP target would be better for both the economy, and the MMMF industry than simply eliminating IOR. BY the way, hasn’t Japan often had zero rates? Maybe that’s why they hold so much cash.

    I don’t follow your last argument. I thought the whole point was that interest rates were stuck at 0.25%, and were no longer adjusting as in the past. In that case the rate chosen does create long run price level indeterminacy. Or am I missing something?

    In any case, my point was broader. It isn’t just a fixed rate that creates indeterminacy, it is any pre-determined interest rate path. You talk about reaction functions, but if the Fed was reacting we wouldn’t have this problem in the first place. I’m not trying to strike any sort of dogmatic point here, just throwing out factors that are at least ambigious, and need to be considered as possible supporting factors. Obviously the price level is not going to zero or infinity. But the process that would cause that, if allowed to go on a few months too long, could cause a cyclical problem, before the Fed got around to correcting it.

  37. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    26. August 2010 at 07:46

    Scott, I am not so optimistic about the mutual fund cash anomaly. If the source of this anomaly is agency costs that lead to excessive short-term performance chasing, the anomaly might remain, as I don’t see any new technologies or market structures that could reduce agency costs.

    TED spread was a good indicator during the crisis because it reflected pessimistic NGDP expectations of an important group of economic agents who were not trading stocks or shopping at Wal-Mart.

    You said:
    “There is never any reason to think something will continue to be a good indicator, as the market is unforecastable.”
    My point was a bit different. There are different markets – stock market, treasury market, eurodollar market and market for orange juice. All of them contain information about the economywide average NGDP expectations, and there is no reason to prefer stock market. Treasury market and eurodollar market will always remain one of the most important markets needed to gauge NGDP expectations, so TED spread will always contain useful information. On the other hand, TED spread – S&P 500 arbitrage is now very efficient, so you will gain virtually nothing by trading stocks taking your signals from TED spreads.

    I have lower estimate of the net incremental social value of money market fund industry than Fed has. So I am comfortable with 0 and slightly negative fed funds rates. You are right that the only thing that can really save money market fund industry is expansionary monetary policy.

    You said:
    “I don’t follow your last argument. I thought the whole point was that interest rates were stuck at 0.25%, and were no longer adjusting as in the past. In that case the rate chosen does create long run price level indeterminacy. Or am I missing something?”
    Rates stuck at 0.25% does not mean that the monetary policy is passive. There is always possibility of new QE, the expected date of the first tightening is also always shifting. Yes, with passive monetary policy we have long run price level indeterminacy and fiscal policy should be recommended in a case of passive monetary policy. But fortunately we do not have passive monetary policy.
    If rates we stuck at 0% instead of 0.25%, the markets would expect that FED tighten earlier (say by 2 months), and the FED would prefer smaller QE. No big net difference in the end.

    You said:
    “In any case, my point was broader. It isn’t just a fixed rate that creates indeterminacy, it is any pre-determined interest rate path. You talk about reaction functions, but if the Fed was reacting we wouldn’t have this problem in the first place. I’m not trying to strike any sort of dogmatic point here, just throwing out factors that are at least ambigious, and need to be considered as possible supporting factors. Obviously the price level is not going to zero or infinity. But the process that would cause that, if allowed to go on a few months too long, could cause a cyclical problem, before the Fed got around to correcting it.”
    Fed has got a flawed reaction function (level targeting would fix it). But a cyclical difference that is caused by 25bps bias in a central bank reaction function is too tiny to notice.

  38. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    27. August 2010 at 12:39

    Scott, I’ve got good news :)
    After Krugman has expressed his strong support for Nick Rowe’s idea of resurrecting Milton Friedman, Brad DeLong has included a direct link to your blog when he discussed Friedman (previously he was always trying to link to some left-wing blog that cited you when he felt the need to discuss your ideas.)
    So you can remove old DeLong’s over-the-top headlines from your office door, and you can attach “Scott Sumner. Liberal’s favorite reincarnation of Milton Friedman” label instead.

    BTW Cowen has cited my recent bond bubble post, and Delong has added a link to his daily linkfest.

  39. Gravatar of ssumner ssumner
    27. August 2010 at 17:18

    123, You said;

    “Scott, I am not so optimistic about the mutual fund cash anomaly. If the source of this anomaly is agency costs that lead to excessive short-term performance chasing, the anomaly might remain, as I don’t see any new technologies or market structures that could reduce agency costs”

    Then let’s you and I go invest the other way, and get rich.

    I agree policy isn’t passive at 0.25% if they do other things. But we’ll have to agree to disagree on the 0.25% question. I have seen huge stock swings on less than a 1/4% fed funds surprise.

    I’ll be on a panel with DeLong in Denver at the AEA meetings next year. I believe Cowen and Galbraith and Quiggen will also be there, but am not sure.

  40. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    28. August 2010 at 03:55

    Scott, you said:
    “Then let’s you and I go invest the other way, and get rich.”
    That’s too optimistic. But it is possible to save ~10% over 10 years by not investing in mutual funds, and earn another ~10% by doing some market timing.

    You said:
    “I agree policy isn’t passive at 0.25% if they do other things.”
    The main reason fiscal policy doesn’t work well is because the Fed is constantly revising their forecast of the time they will increase rates to 0.50%.
    And by the way, there were no net new asset purchases by the Fed since December 2008, and rates have been stuck at 0.25% from December 2008.

    You said:
    “But we’ll have to agree to disagree on the 0.25% question. I have seen huge stock swings on less than a 1/4% fed funds surprise.”
    I agree that 0.25% can move markets by 2-3%. But if you have 15-30% in mind, we have to agree to disagree.

    What is your interpretation of market reaction to Bernanke’s speech on Friday? I think a threat of more QE = higher inflation expectations, S&P 500 up and higher real interest rates.

  41. Gravatar of ssumner ssumner
    28. August 2010 at 15:30

    123, I forgot to congratulate you on the Cowen link.

    I doubt that market timing is possible, but since I have done it successfully, I suppose I should keep an open mind.

    I agree with your comments about fiscal policy.

    On the day of the March 2009 “QE” announcement, the dollar fell about 4 or 5 cents, I recall. So that must have meant something to the markets.

    I agree with your final point. I think Bernanke hinted that the Fed would not allow inflation to drift below 1%, and I think that was a small plus to the market.

  42. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    29. August 2010 at 03:51

    Scott, you said.
    “123, I forgot to congratulate you on the Cowen link.”
    Thanks. This wouldn’t have been possible without all those debates with you.

    “I doubt that market timing is possible, but since I have done it successfully, I suppose I should keep an open mind.”
    You should think about what is the source of your success:
    a) pure luck
    b) different risk preferences
    c) better valuation and macro models
    d) low agency costs

    “I agree with your comments about fiscal policy.”
    But if fiscal policy cannot move us to the other side of a knife-edge equilibrium, IOR also can’t.

    “On the day of the March 2009 “QE” announcement, the dollar fell about 4 or 5 cents, I recall. So that must have meant something to the markets.”
    So what was it:
    QE view – markets did not know the pace of the exit from other liquidity programs and mistakenly thought that fed’s balance sheet will expand, or correctly understood that an undesirable contraction will be prevented.
    credit channel view – the Fed has assumed lots of duration, convexity risk and has started treasury-MBS arbitrage. This has a beneficial effect on household and financial institution balance sheets.
    I guess that market move was a 100%-50% result of credit channel view, and 0%-50% result of QE view.

  43. Gravatar of ssumner ssumner
    29. August 2010 at 10:24

    123,

    1. Pure luck.

    2. you said;

    But if fiscal policy cannot move us to the other side of a knife-edge equilibrium, IOR also can’t.”

    I never said fiscal policy can’t work, I said it was unlikely. All macro policy works primarily through expectations. That means many outcomes are possible.

    3. I don’t have strong views about March 2009, you might be right.

  44. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    30. August 2010 at 03:20

    Scott,
    do you mean that fiscal policy should be expected not to work?

  45. Gravatar of ssumner ssumner
    30. August 2010 at 07:56

    123, Yes, but it could work if it forced the Fed to adopt a policy that raised the expected future NGDP.

  46. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    30. August 2010 at 13:08

    Scott, could you give an example?

  47. Gravatar of Scott Sumner Scott Sumner
    1. September 2010 at 10:58

    123, Suppose it led people to expect the Fed to monetize the debt, because it was so large.

  48. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    2. September 2010 at 13:37

    Kling has linked to a self-recommending interview with Sargent:
    ” Far from taking the “efficient markets” outcomes for granted, important parts of modern macro are about understanding a large and interesting suite of asset pricing puzzles, brought to us by Hansen and Singleton and their followers—puzzles about empirical failures of simple versions of efficient markets theories. Here I have in mind papers on the “equity premium puzzle,” the “risk-free rate puzzle,” the “Backus-Smith” puzzle, and on and on.2
    These papers have put interesting new forces on the table that can help explain these puzzles, including missing markets, enforcement and information problems that impede trades, difficult estimation and inference problems confronting agents, preference specifications with novel attitudes toward the timing and persistence of risk, and pessimism created by ambiguity and fears of model misspecification.”

    I believe some of the forces Sargent mentions above explain big part of your stock market outperformance.

    source: http://www.minneapolisfed.org/publications_papers/pub_display.cfm?id=4526

  49. Gravatar of ssumner ssumner
    3. September 2010 at 06:30

    123, I suppose that’s possible. BTW, keep in mind that some of that isn’t addressed at EMH broadly defined, but rather specific versions of EMH that assume things like stable discount rates. I agree those models are too rigid.

Leave a Reply