The Fed doesn’t actually “control” short term interest rates

At least not in the sense that NYC controls the rent on apartments.  This post was triggered by a recent Karl Smith post:

The interest rate on T-Bills is simply whatever the Fed wants it to be. T-Bills and bank excess bank reserves are essentially interchangeable. In normal times the value of excess reserves is the Fed Funds rate. Today it is the Interest on Reserves rate. However, both of those are essentially controlled by the Federal Reserve.

This is the conventional wisdom, but I thinks it’s (approximately) wrong.  Because this is very confusing, let’s start off by discussing why controlling interest rates is not like controlling rents.  Under rent control, the government legally mandates a particular maximum rent, and a shortage often results.  The Fed doesn’t put any legal price controls on fed funds or T-bills.   There is no shortage.  Rather it adjusts the money supply as needed to keep short term rates at the desired level.  One might argue that this is merely a quibble, but in fact there is a much bigger problem with Karl’s argument, which he alludes to in this paragraph:

While its possible that the interest rate on T-Bills averages only 3.3% over the next 30 years this is – hopefully – extremely unlikely. It implies that nominal GDP growth will average 3.3% over the next 30 years, as the Fed must ultimately align interest rates with nominal growth rates or the economy will persistently overheat or stagnate.

To see why this matters, consider the following thought experiment.  Suppose the Fed was run by God, and God was considering the following options for hitting a 2% inflation target:

1.  Adjust the monetary base until God himself (or herself) expects 2% inflation.

2.  Adjust the monetary base until the fed funds rate settles at a level that God expects to produce 2% inflation.

3.  Adjust the monetary base until the euro/$ exchange rate settles at a level that God expects to produce 2% inflation.

4.  Adjust the monetary base until the CPI futures market settles at a level that God expects to produce 2% inflation.  This need not be 2% inflation in the CPI futures market, because God knows the exact risk premium in CPI futures prices.  (And I mean “God knows” literally, not in the sense of “who knows?”)

I hope it’s obvious that all 4 monetary regimes are identical.  The path of base money, fed funds, exchange rates and CPI futures prices is exactly the same in all four cases.  I think it’s also obvious that no one would call options 1, 3 and 4 “controlling interest rates.”  Since option 2 is exactly the same, it is also not “controlling interest rates.”

So what is different about our current monetary regime that leads people to think that the Fed does “control” short term interest rates?  I guess we could start with the fact that Ben Bernanke isn’t God.  But seriously, I see two differences.  I foresaw God adjusting the fed funds target continuously, and the Fed actually adjusts the target every 6 weeks.  And I saw God as being omniscient, and the Fed . . .  well let’s just say it’s not.  So obviously there is a sense in which Karl is right.  The Fed sort of “controls” (but doesn’t legally fix) the fed funds rate for 6 week periods.

But the Smith quotation on top referred to T-bill yields, not the fed funds rate.  In this case Karl’s argument is even weaker.  As the market sees the economy strengthen and/or inflation rise, it will bid up T-bill yields in anticipation of the Fed raising the fed funds target at the next meeting.  If the market sees the central bank as operating on a target-the-forecast basis, then T-bill yields will always be determined by the market, not the Fed.  They will represent the market’s estimate of what sort of short term interest rate path would mostly likely be associated with 2% inflation (or 5% NGDP growth, if that were the Fed’s target.)

Looking at monetary policy in terms of interest rate control is likely to lead one astray:

1.  It leads many people to equate low rates with easy money, and vice versa.  (Never reason from a price change.)  Karl doesn’t do that, but lots of people do.

2.  It leads lots of libertarians to assume the Fed is more interventionist that than it really is.  The Fed does have monopoly on money creation, and that’s obviously very interventionist.  It means the Fed steers the nominal economy (although like all captains, it occasionally goes off course.)  But it doesn’t make sense to argue the Fed targets inflation and is also highly interventionist in the credit markets.  If T-bill yields are X% when the Fed is successfully targeting inflation at 2%, then those are essentially free market short term interest rates.  You can’t target two variables with one tool.  It may be that 2% inflation was a bad target in the middle of the housing boom, and tighter money was desirable.  But it makes no sense to argue the inflation target was wrong and also that the interest rates were too low.  Had the target for inflation been lower, or had a 5% NGDP target was been adopted (as David Beckworth recommended), it’s quite likely that short term rates would have been even lower, as economic growth would have been weaker.  Nominal interest rates are strongly procyclical, they are highly correlated with NGDP relative to trend.

PS.  Although Milton Friedman was a great economist, he made one silly mistake.  He once argued that fixed exchange rates were a bad idea because the market should set exchange rates.  But fixed exchange rates are no more nor less interventionist than a fixed growth rate of the money supply (Friedman’s preferred target.)  In a sense all central bank policy targets are equally interventionist.  If the central bank remains in public hands, the only question is which target produces the least harm to the economy.


Tags:

 
 
 

49 Responses to “The Fed doesn’t actually “control” short term interest rates”

  1. Gravatar of Karl Smith Karl Smith
    25. March 2012 at 08:49

    I think there is some meaning in saying the Fed controls short term rates because given the current regime they can and frequently do miss their inflation targets, but they essentially never miss their interest rate targets.

    So, if they were prefect on inflation or NGDP then interest rates would be predetermined. But, because they frequently miss on inflation and NGDP we can think of interest rates as determined by the process that goes on in FOMC meetings.

  2. Gravatar of Cantillonblog Cantillonblog
    25. March 2012 at 09:10

    I understand the point you are trying to make, but isn’t it an unfortunate time to try and make it? It would have been better made a few years ago. Today the Fed sets IOER as an administered rate. Because of arbitrage sort of considerations it’s pretty hard for Fed Funds to diverge very far from this level. So the Fed basically does set Fed Funds.

    Over and beyond that, there is a big difference between option 2 and the others. There is a stickiness and a lagging adjustment process in Fed policy – target Fed Funds doesn’t go up and down intraday, or even week to week, in the way that one’s expectations about the rate required to produce 2% inflation given evolving data evolve. And there is a predictability in Fed Funds that there wouldn’t be under the other regimes. Once the Fed has started hiking, you kind of know it has a certain expected distance to go before it is done.

    Your critique of Karl Smith seems also misplaced. T Bills have some technical features – their collateral value and absolute safety allow them to trade at negative yields when credit markets are distressed, even with a positive Fed Funds rate. But abstracting away from this dynamic, the Fed does have control over the Fed Funds futures strip (from which pricing of T Bills is derives as a collateral spread). If the market starts trying to tighten policy and the Fed doesn’t like it the Fed can just jawbone the market higher in price/lower in yield. And the embarassment factor and past track record means such open mouth policy is generally effective.

    Of course if the Fed adopted your favoured “market monetarism” approach then it would lose control of short rates. Just as if it went back to the gold standard, it would lose control of short rates. Or if it decided to target growth in M2, or broad liquidity. But I don’t see why that makes Karl Smith wrong.

    And finally with regards to Friedman’s point I see things very differently. If a government targets 3% growth in M2, there is not really room for interpretation. But under a fixed exchange rate regime, when your currency is in demand, you can print a lot of money without much bad seeming to happen. Then when the pendulum swings to the other side and your currency is disfavoured, it’s very painful to tighten money enough to maintain the exchange rate and the temptation is to intervene or try to control capital flows. Under fractional reserve banking, the elastic is too loose to be an effective constraint on monetary policy.

  3. Gravatar of Justin Irving Justin Irving
    25. March 2012 at 09:38

    To push back a little on the Friedman (may peace be upon him) jab: I suppose an M2 growth rate/level target is no more or less ‘interventionist’ than an exchange rate target, but an M2 target gets one closer to stable NGDP than an exchange rate target, which is guaranteed to bring long term drift in the NGDP path (Hong Kong, every EMU/ERM2 country). So in a pragmatic sense, Friedman was right, better to have the market set exchange rates conditional on an M2 target, than vice versa. Maybe I am missing the point here.

  4. Gravatar of D R D R
    25. March 2012 at 09:50

    Isn’t this like saying airline pilots don’t control altitude, because altitude is determined by gravity and lift?

  5. Gravatar of Jon Jon
    25. March 2012 at 09:51

    ” And I saw God as being omniscient, and the Fed . . . well let’s just say it’s not. So obviously there is a sense in which Karl is right. The Fed sort of “controls” (but doesn’t legally fix) the fed funds rate for 6 week periods.”

    1) The IOR amount and the t-bill rate are different; they have been different for some time and are usually different. The reason is that reserve accounts are open only to certain institutions. The rest of world bids the t-bill rate lower (or higher when banks are reserve limited).

    2) While the FF target is set every six weeks; you will see that the NYFED trading desk is rather more sly than that. During Fall’08 it was routine to see the trading operation ease past the announced targeted. Indeed, the FOMC appeared to ratify the preemptory move by the NYFED which as soon as they did so the trading desk let the target slip further.

    This was in the gossip circles at the time as an open political battle within the Fed with the hawks dominating the FOMC but the Doves largely in technical control and basically bullying the FOMC into lowering the FF target.

    —-

    Once again we have a case here of a monetary policy theorist (Karl) who is insightful on theory but rather ignorant of day-to-day operational practice and history. This isn’t the worst example of this though. (For that I point to the wide-spread idea that the OMOs were primarily conducted in t-bills (before the crisis), when this was quite blatantly false in the trading operations of the Fed and in their public statements about why the Fed held t-bills (to have a pool of ready securities to sterilize a sudden draw at the discount window). Too many professors are too arrogant to admit this particular error. Particularly as they’ve rather perniciously filled the textbooks with the same nonsense.

    These kinds of mistakes are all well and minor, until they lead to some critical errors such as the kind discussed here.

  6. Gravatar of ssumner ssumner
    25. March 2012 at 09:54

    Karl, Just to be clear I’m talking about the pre-2008 period. Right now the Fed seems to have giving up targeting rates, and uses other tools such as QE2 (monetary base changes.)

    Second, my inflation targeting hypothetical was simply meant to be illustrative. They actually target a weighted average of inflation and output deviations (aka something like the Taylor Rule.) In that case it’s less obvious that they miss their target (prior to 2008) But more importantly, what matters isn’t whether they miss their target ex post, but whether rates are always set at a level expected, ex ante, to hit their target. That seems an eminently reasonable assumption to me (at least prior to 2008.) And as I said, post-2008 they clearly aren’t targeting rates–market rates have fallen to near zero, and can’t go lower.

    If rates are always set at a level expected to achieve on-target AD growth going forward, then it’s reasonable to say the Fed “controls AD growth expectations,” and the market controls short term rates (on T-bills.)

    I’m not saying you are completely wrong (there is obviously a bit of wiggle room), just that it makes more sense to see the central bank controlling nominal aggregates than controlling interest rates.

    Cantillonblog, I don’t agree with your first point. The Fed considers interest rates to have fallen to zero, and hence they’ve lost control of them (the 1/4% of IOER is just to keep money market funds afloat.) Right now the Fed is relying on QE, not control of rates.

    You said:

    “Over and beyond that, there is a big difference between option 2 and the others. There is a stickiness and a lagging adjustment process in Fed policy – target Fed Funds doesn’t go up and down intraday, or even week to week, in the way that one’s expectations about the rate required to produce 2% inflation given evolving data evolve.”

    No, there is no difference at all between option 2 and the others. There would be a difference if the Fed was targeting short term rates for 6 weeks, but that’s not part of option 2–which was continuous control of rates. In my post I indicated that in the real world the Fed does target the fed funds rate for 6 weeks at a time, and hence there was a bit of truth to Karl’s claim. But Karl’s claim referred to T-bill yields, over which the Fed has far less control than the fed funds rate. The Fed doesn’t just arbitrarily move rates around. It adjusts the money supply in such a way that free market interest rates are expected to lead to on-target aggregate demand growth.

    You said;

    “Of course if the Fed adopted your favoured “market monetarism” approach then it would lose control of short rates. Just as if it went back to the gold standard, it would lose control of short rates. Or if it decided to target growth in M2, or broad liquidity. But I don’t see why that makes Karl Smith wrong.”

    Another thing that would make Karl wrong is if they targeted expected inflation, or if they targeted a forward-looking Taylor Rule, or some other similar Svenssonian approach. I think if you asked Bernanke or anyone else at the Fed they’d claim they are already Svenssonian. I.e. if you asked Bernanke whether some alternative interest rate path would be expected to produce superior macroeconomic results, he’d say “of course not.” He’d say policy is expected to produce the results they are looking for (obviously I’m talking about before 2008.)

    So in that case a Svenssonian policy would also make Karl wrong. Now I happen to think policy isn’t exactly Svenssonian, that there is a little bit of inefficient inertia in fed funds targets, (although even less in T-bill yields) hence I said Karl was “approximately wrong.”

    We’ll have to agree to disagree on exchange rates. In my view the money supply is endogenous when you target exchange rates.

  7. Gravatar of ssumner ssumner
    25. March 2012 at 10:04

    Justin, You said;

    Maybe I am missing the point here.”

    The only point you are missing is in assuming that I disagree with what you said. Reread the last 13 words of my post. But more or less effective is very different from more or less interventionist.

    DR, Definitely not. I think readers aren’t thinking hard enough about my four options. They are exactly the same, even though one looks like interest rate control and the other three don’t. Yes, the real world is more complicated, but that example (stylized as it is) is telling us that our intuition may be a bit off.

    Also note that my argument here applies best to T-bill yields; as I mentioned Karl has a much stronger case to make for control of fed funds rates (although even there I have reservations about how “control” is misinterpreted.)

    Jon, Interesting point. But I don’t want to call Karl “ignorant” as I didn’t know that about the NY desk either. I do recall rates slipping below target in late 2008, but thought it reflected the big QE. In other words, does that occur during “normal times” prior to 2008?

  8. Gravatar of D R D R
    25. March 2012 at 10:17

    Never mind. I don’t think you understood my point, and I doubt it’s worth clarifying for your sake.

    But once again, I question whether you aren’t confusing policy and outcome.

  9. Gravatar of Cantillonblog Cantillonblog
    25. March 2012 at 10:24

    It seems most peculiar to me to assert that the Fed has lost control of interest rates. If it so pleased them, the Fed could hike rates at the next meeting to whatever level they felt was required. And for practical but powerful reasons, there are limits to how far the structure of spreads between T Bills, IOER and target Fed Funds can move. The Fed cannot set IOER at 10% and choose a monetary base in order to achieve a 0% target on Fed Funds. Furthermore, given that in practical terms the Fed seems to choose a monetary base that allows it to hit its Fed Funds target, on what grounds do you claim that the Fed is not controlling Fed Funds? Right now it thinks Fed Funds should be zero, and is using QE2 to ease further. But that doesn’t mean it doesn’t control Fed Funds. Fed Funds are exactly where the Fed wants them – at zero.

    “But Karl’s claim referred to T-bill yields, over which the Fed has far less control than the fed funds rate. The Fed doesn’t just arbitrarily move rates around. It adjusts the money supply in such a way that free market interest rates are expected to lead to on-target aggregate demand growth.”

    It would have been better if he had referred to the Fed Funds futures curve, since T Bills have some aspect of collateral value to them. But in practice the Fed can get the first few fed funds contracts wherever it wants them. If it thinks they are pricing rate hikes that shouldn’t be there it can announce it isn’t going to hike, or say something similar in a softer manner.

    You assert that the Fed just adjusts the money supply so that its models imply it will hit desired aggregate demand growth, but that’s just an assumption and is contingent on the particular set of people and policies currently in place. Bernanke could quit tomorrow, and we could find once things have settled down that we have another Fed Chairman who is much more pliant towards what politicians want and sets policy easier ahead of the election. Or we could see growth and inflation surprise to the upside and eventually see a bond market selloff to the extent that Fed feels obliged to answer for the third component of its mandate and stabilize yields by buying bonds directly (risking price stability). In practice the Fed controls Fed Funds (and can choose to pursue unconventional measures if it considers them appropriate). The target may be for now as you describe it, but this can certainly shift over time. The Bank of England shows us how creative central banks can be when choosing how to interpret its target so as to do what it feels it ought to.

    “They actually target a weighted average of inflation and output deviations (aka something like the Taylor Rule.)”
    That’s your guess, based on their behaviour they followed in the recent past. At times before that they pursued different policies. And maybe they will continue to behave the way they do today. But it’s not like it is set in stone, or anything.

  10. Gravatar of Cantillonblog Cantillonblog
    25. March 2012 at 10:28

    “I do recall rates slipping below target in late 2008, but thought it reflected the big QE. In other words, does that occur during “normal times” prior to 2008?”

    I’m not sure what ‘normal times’ means, but you saw this (realized fed funds slipping below the target) happen more than a few times in previous episodes of stress. See March 2005, for an example.

    I agree that academic economists seem to often be very confused about exactly how markets work, but I didn’t think this question of plumbing detracted from Karl Smith’s point at all. Nitty gritty details about the collateral spread are really not material to it, and his point would stand if more pedantically phrased.

  11. Gravatar of Cantillonblog Cantillonblog
    25. March 2012 at 10:34

    D R I don’t think your point was understood, and I think there is indeed a confusion of policy and outcome. I think we have a very left hemisphered confusion between our representation of the world and the world as it truly is.

    If the Fed were a machine, and economic modelling were objective and economic forecasts were any good then Sumner would be right about everything being determined by its unarticulated objective function. Given welfare function of Fed, economic model, and some market inputs then the Fed has no discretion about what to do. Target for Fed Funds and money supply just fall out of the model.

    But having been an institutional speculator and investor in interest rate markets for eighteen years (having to get direction of markets right before the moves happen, not just make up a nice story to explain it after the move is over), that’s not in my view actually a good description of how the world works. Economic models are horrible; economic forecasts are no good; the Fed has a great amount of discretion in practice about how to act; its welfare function isn’t well defined. I could go on.

  12. Gravatar of CA CA
    25. March 2012 at 10:56

    I love it when anonymous people online talk about their successful investing careers.

  13. Gravatar of D R D R
    25. March 2012 at 10:58

    Cantillionblog,

    I don’t see the significance. For whatever reason, the Fed selects a target interest rate. Maybe they select a rate which is more or less consistent with the dual mandate, or maybe they select a rate which they falsely believe is consistent with the dual mandate. It doesn’t matter. If they are able to hit their selected target, then they are controlling the interest rate.

    Now, perhaps, given the target, the market determines the money supply, inflation, unemployment… so maybe they aren’t in direct control of those (or do a poor job of controlling them) But that seems a different question.

    It seems to me that if the Fed is not in control of those interest rates, then they are awfully good at predicting what they will be– at least in the short run.

  14. Gravatar of Cantillonblog Cantillonblog
    25. March 2012 at 11:14

    D R – I am very much in agreement with you. The significance of all those loose joints I point out (the inability to forecast, the practical freedom to act, etc) is that the Fed has free will and therefore can choose whatever Fed Funds target it wants. If it were a machine and there were not all these loose joints, then Sumner’s point would be spot on – in effect it would have no freedom to act since the target rate would just spill out of the model. It’s just that this does not characterize reality.

    C A – I do not much love the snark of anonymous people on teh interwebs. For regulatory reasons I do not wish to be perceived as marketing myself, or as talking up my position. My point is not to talk about my success in investing, but to explain a difference in focus. It is one thing to make up stories after markets have moved to make sense of the moves. It is another to try to anticipate moves before they happen. To my way of looking at the world, a somewhat unpolished framework that is successful in helping one anticipate must surely be preferred to one that gives one a perfect rear view mirror narrative but that has no ability to anticipate changes before they happen.

  15. Gravatar of cthorm cthorm
    25. March 2012 at 11:29

    As the market sees the economy strengthen and/or inflation rise, it will bid up T-bill yields in anticipation of the Fed raising the fed funds target at the next meeting.

    I had to read this part a couple of times to get your logic, even though it’s a simple point. One does not “bid up” T-Bill yields, one “bids up” the price of T-bills, pushing yields down; alternatively, one “bids down” the price of T-bills, pushing yields up. So what you’re saying is that a bullish outlook would make investors value T-bills less.

  16. Gravatar of D R D R
    25. March 2012 at 11:35

    “If it were a machine and there were not all these loose joints, then Sumner’s point would be spot on – in effect it would have no freedom to act since the target rate would just spill out of the model.”

    Perhaps… but if the Fed had no discretion that it wouldn’t be in control of anything at all, would it?

  17. Gravatar of Bill Woolsey Bill Woolsey
    25. March 2012 at 12:22

    I disagree.

    Suppose that this week the short term interest rate needed to clear credit markets is 3% and next week it needs to be 1% to clear credit markets.

    The Fed, wanting to keep short term interest rates stable, creates an excess supply of money this week to keep the short term interest rate 2%. Next week, it creates an excess demand for money to keep short term interest rates at 2%.

    The excess supply of money this weeek would, if it persisted, eventuually cause nominal expenditure and the price level to rise to a higher growth path. But it doesn’t persist. In fact, next week, there is an excess demand for money that offsets this week’ excess supply of money.

    The Fed is able to keep the short term interest rate at 2% and because the excess supplies or demands for money are temporary, it also keeps nominal GDP on target.

    Long term interest rates aren’t much effected either.

    If you don’t like a week, go for a day or an hour. Now, how long can we go? Keep the itnerest rate from falling with an excess demand for money for a month, and then an excess supply of money to keep it from rising for another month? Quarters? Six months?

    Suppose the central bank has great credibility in its comittment to inflation. How much excess supply or demand for can it generate before firms start setting prices in a way inconsistent with the target? That is, how long must an excess supply or demand for money persist before people start doubting that it will be reversed in the future?

    How long can the Fed keep market interest rates from rising before people believe that they won’t offset the inflationary pressure?

    It is true, of course, that if the short term interest rate needed to clear credit markets starts to fluctuation around 3%rather that 2%, keeping it at 2% would creat problems.

    Implicit in my analysis is the alternative of having the quantity of money adjust to the demand to hold money. Failure to make such adjustments can result in a liquidty impact on interest rates. However, if there is a shortage or surplus of credit at the current interest rate, creating an excess demand or supply of money can create a liquidity effect that keeps the interest rate from changing.

    Perhaps thinking too much about what happens if there is an exogenous change in the quantity of money–say, for example, that involves government creating money and spending it or giving it away–hides some interesting issues.

  18. Gravatar of D R D R
    25. March 2012 at 12:39

    “I disagree.”

    With whom?

  19. Gravatar of Major_Freedom Major_Freedom
    25. March 2012 at 13:10

    ssumner:

    Although Milton Friedman was a great economist, he made one silly mistake. He once argued that fixed exchange rates were a bad idea because the market should set exchange rates. But fixed exchange rates are no more nor less interventionist than a fixed growth rate of the money supply (Friedman’s preferred target.)

    This is an excellent point. This kind of “intervention” is really just setting a definition, and then sticking to it, and letting market forces act on the side of supply and demand utilizing this definition. It is very much like the “intervention” in defining “the dollar” to be “1/20 ounces of gold”, or whatever, and then sticking to that definition by not inflating more dollar bills than there is gold. It’s like defining a foot to be 12 inches, and then sticking to that definition by not changing the distances between the tick marks on the rulers, and then noticing that one foot has a fixed “exchange” rate with yards, miles, centimeters, kilometers, etc.

  20. Gravatar of Major_Freedom Major_Freedom
    25. March 2012 at 13:16

    Karl Smith:

    I think there is some meaning in saying the Fed controls short term rates because given the current regime they can and frequently do miss their inflation targets, but they essentially never miss their interest rate target.

    This is the correct interpretation IMO.

    A quick glance at this graph that charts the targeted fed funds rate along with the effective fed funds rate, can show the Fed does have “control” over short term rates.

    I mean, what else can we call a situation where the Fed publicizes a particular fed funds rate, and that is the rate that actually occurs?

    For better or for worse, short term rates are tied to the fed funds rate, so while the Fed doesn’t “control” short term rates on things like t-bills, they nevertheless strongly influence them. I can only see the Fed failing to influence short term rates like t-bills, if they lose control of the currency.

  21. Gravatar of Bill Woolsey Bill Woolsey
    25. March 2012 at 13:30

    DR:

    With Sumner’s post.

  22. Gravatar of 123 123
    25. March 2012 at 13:33

    In fact, there was a period when the Fed wanted to control the FFR, but failed. The result was the Great Recession:
    http://themoneydemand.blogspot.com/2010/11/collapse-of-fed-funds-rate-peg-and.html

  23. Gravatar of Jon Jon
    25. March 2012 at 13:34

    I do recall rates slipping below target in late 2008, but thought it reflected the big QE. In other words, does that occur during “normal times” prior to 2008?

    Which rate are we talking about? Historically, the NYFed has maintained very precise control of the FF rate, tracing the target very exactly. You can debate whether the behavior in fall’08 was due to the QE operations. As I said, the gossip at the time was that this wasn’t a technical result.

    The t-bill rates have almost always been under the FF rate.

  24. Gravatar of Bill Woolsey Bill Woolsey
    25. March 2012 at 13:37

    The Fed could control T-bill rates–or at least the rate on one particular T-bills. Just buy and sell them at the target price. It is true, of course, that the federal funds rate might vary a bit, but not too much.

  25. Gravatar of Louis R. Woodhill Louis R. Woodhill
    25. March 2012 at 13:37

    Anyone who has ever drive on an icy road knows that you don’t control the direction of the car. You *control* the position of the steering wheel and you *target* the direction of the car.

    Right now, the Fed truly *controls* three things: the size of the monetary base, the announced Fed Funds target rate (which has a psychological impact), and the IOR rate. It targets the Fed Funds rate, which in, in turn, influences other interest rates.

  26. Gravatar of 123 123
    25. March 2012 at 13:39

    Jon, NYFed has completely lost control after Lehman, FFR was too high. See my comment above.

  27. Gravatar of Jim Glass Jim Glass
    25. March 2012 at 17:21

    But, because they frequently miss on inflation and NGDP we can think of interest rates as determined by the process that goes on in FOMC meetings.

    Milton Friedman said until the day he died that the Fed controls the money supply and pretends (his word) to control the interest rate. Many others say the Fed controls interest rates. Yet others seek to square the circle by saying the Fed targets the interest rate but as the interest rate is a price reflecting supply and demand, the Fed reaches its rate target by actually controlling the supply of money through open market operations — “that’s what Friedman means”, Lucas says at Econtalk.

    I once heard Greenspan explain it this way: A given level of money supply corresponds with a given interest rate. However there is significant very short term noise/variation in the relation. Thus if the money supply is held fixed at a desired amount there will be a lot of short term interest rate variation around the average, while if the interest rate is kept set there will be variation in the money supply around the average. As a practical matter, business/finance operations are very sensitive to interest rate changes, while nobody cares about short-term money supply variations, so the Fed stabilizes the interest rate and varies the money supply as needed to keep it stable. (Apart from the 1981 episode when the Fed politically couldn’t set a “target” rate high enough, and had lost credibility with the markets not believing it would control inflation, so Volcker did the reverse, controlled the money supply directly and let the rate go where it would.) That’s how it determines “things” in the short term, of course in the longer term it changes the entire level of the supply-rate correlation in line with its changing policy objectives and economic conditions. That’s what he said.

    FWIW.

  28. Gravatar of Major_Freedom Major_Freedom
    25. March 2012 at 17:35

    Jim Glass:

    I once heard Greenspan explain it this way: A given level of money supply corresponds with a given interest rate.

    Greenspan said that? Was he drunk?

  29. Gravatar of Bonnie Bonnie
    25. March 2012 at 22:51

    Here’s a paper by James Bullard of the St. Louis Fed describing sort of how they derive the stable state. It’s written in the context of how to deal with ZLB and how to not end up like Japan when they can’t use interest rates as a policy tool. I came away with the impression that they definitely do not want interest rates at zero, just can’t agree on the best way forward.

    “Seven Faces of the Peril”
    http://research.stlouisfed.org/publications/review/10/09/Bullard.pdf

  30. Gravatar of Browsing Catharsis – 03.26.12 « Increasing Marginal Utility Browsing Catharsis – 03.26.12 « Increasing Marginal Utility
    26. March 2012 at 04:08

    […] Sumner corrects Karl Smith interest rate myopia. How many times is it gonna happen until this is going to sink in for Keynesians? I guess I need to post this again: […]

  31. Gravatar of ssumner ssumner
    26. March 2012 at 05:38

    Everyone, This is turning out to be such a complicated issue that I’m going to stay away from the last three years (zero rates and IOR) and focus on whether they have control in “normal times.” (prior to 2008)

    Catillonblog, You said;

    “Economic models are horrible; economic forecasts are no good; the Fed has a great amount of discretion in practice about how to act; its welfare function isn’t well defined. I could go on.”

    Obviously I agree with all that, which makes me think we are talking past each other. Let me try a different tack. No one would argue that the Fed controls the price of toasters sold at Walmart. But the Fed policy does influece the price of toasters, as it influences all nominal variables in the economy. What makes the fed funds rate different is that the Fed actually pegs that rate for periods of 6 weeks. No one would dispute that they control the fed funds rate for that six week period, and no one would dispute that they could extent that period longer, if they wished. But there is a tension there. The more (arbitrary) control they have over the fed funds rate the less control they have over inflation, or their other goal variables. If they are serious about keeping your own internal forecast of inflation on target, the interest rate become endogenous. And it makes no different whether your goal is different from inflation, and it makes no difference how bad your models are. There is a tradeoff between control of the fed funds rate and control of inflation, or whatever the goal variable actually is.

    But it’s even worse, as the less effective the control of inflation, the more the T-bill yield will diverge from the fed funds target. In December 2007 there was a contractionary surprise out of the Fed, and yet 3 month T-bill yields actually fell. That’s because the market saw that the Fed had blown it, by cutting rates by only 1/4%, instead of the 1/2% hoped for. So the market anticipated the Fed would see its error and cut rates much more sharply in the near future. And that’s what happened, as the Fed cut rates by 125 basis points over the next 6 weeks. I don’t think anyone in their right mind would call that panicky cut in rates “Fed control” over short term interest rates–the markets forced the Fed’s hand. Sure, if they wanted to be stubborn they could peg rates forever, at the cost of hyperinflation or hyperdeflation.

    Regarding your debate with CA, I certainly agree that changes in the fed funds target are forecastible, at least to some extent. In 2004-06 the rate marched steadily higher–everyone saw what was happening.

    Cthorm, Yes, that’s what I meant.

    Bill, No one disputes that the Fed can peg the fed funds rate for 6 weeks, or longer–see my reply to Cantillonblog.

    The other issue you raise is the indeterminacy problem. There is more than one interest rate path consistent with the same inflation outcome. I agree, but that certainly doesn’t indicate “control” in the usual sense of the term, as the Fed still needs to make sure the average level of rates is consistent with their inflation target. The difference between 2% and alternating between 1% and 3% is rather trivial. And the impact on T-bill yields (the issue discussed in this post) is more trivial still.

    MF, You said;

    “For better or for worse, short term rates are tied to the fed funds rate, so while the Fed doesn’t “control” short term rates on things like t-bills, they nevertheless strongly influence them. I can only see the Fed failing to influence short term rates like t-bills, if they lose control of the currency.”

    That correlation breaks down when the Fed fails to successfully target it goal variable. See my discussion of the December 2007 meeting above.

    123, Good example.

    Jon, I agree that’s normally true, but see 123’s comment.

    Louis, I agree.

    Jim Glass, I agree that rates are controlled via changes in the money supply, but also through signaling of future money supply changes.

    Thanks Bonnie,

    Everyone, I think Bill got at the most important issue here–indeterminacy. There is more than one interest rate path consistent with 12 month forward inflation and/or NGDP expectations remaining on target. I would add that the longer the term of the interest rate, the less control the Fed has, as those paths will tend to average out. Thus 2% fed funds rate for the next 6 weeks is very similar to 1% for 3 weeks and then 3% for the next three weeks. I agree that the Fed can choose between those two paths. In that sense they have control. But those two paths imply almost identical 3 month T-bill yields. And if you redo the problem with 1% for 3 months and then 3% for the next 3 months, I’d argue the 6 months T-bill yield is the same. In other words the argument for control has some validity for the fed funds rate, far less so for 3 month T-bill yields, and almost none fore 6 month T-bill yields. That’s the sense in which Karl was “approximately” wrong in my view.

  32. Gravatar of Bill Woolsey Bill Woolsey
    26. March 2012 at 06:22

    Scott:

    OK.

    But I don’t like interest rate smoothing, and your argument seemed to igore it.

    Also, as you often point out, temporary changes in base money have no effect. That is the other side of the coin of an ability to make interest rate smoothing more persisent.

    I think it is also related to Austrian Business Cycle Theory.

    I am not sure that “indeterminacy” is the way I would frame the issue, but I certainly agree that interest rate smoothing has little effect on long term interest rates.

    Your framing is that the central bank might manipulate short term rates between 1 and 3 percent or leave them at 2%. My framing is that the central bank might manipulate the quantity of money–create shortages or surpluses of money–to prevent fluctuations of short term rates between 1 percent and 3 percent, so that short term rates stay 2%. Longer term rates, however, would stay 2% under either scenario.

    To me, the short term rates give useful signals. Mostly it is about when do you start the investment project. With the status quo, everyone starts when they like, and the Fed creates enough money to make sure they can. With short term rates fluctuating, it creates an incentive to smooth the start of the projects. If everyone wants to start now, then short rates rise, and those for whom it is easier to wait a bit, do so, and start later when short term rates are low. If short term rates are low, then lets get started now.

    This could involve consumer durable borrowing as well. Short rates are high, wait to buy the car. Short term rates are really low, lets start looking at ads today. We do need a new car.

    Interst rate smoothing ruins this. If everyone wants to start the project now, the central bank creates enough money for them to do so. It works because people will accept the money even though they don’t plan to hold it. If the demand is for capital goods, the direct impact on their prices does not have anything to do with the CPI. Produce and sell the capital goods now when demand and prices are high. If this will ultimately lead to a higher CPI, then before that time, the central bank will tighten, and sales will be bad. Make hay while the sun is shining!

    Maybe I am missing something, but I think central banks control interest rates and _screw_ _them_ _up_.

    Now, if there is no change in the demand for credit, and instead, there is a change in the demand to hold money, say, people sell off short term securities to accumulate money (or maybe just don’t buy as many out of current money receipts,) then the increase in short term rates would be inappropriate. There is no need to pospone projects to free up resources, because the people accumulating money aren’t spending it. They are just changing the form in which they hold wealth. Having the central bank “smooth” that increase in interest rates would be appropriate. But the point isn’t to smooth the interest rates, but rather to increase the quantity of money enough to match the added demand to hold money.

    As I said before, (or maybe hinted) I think you are still doing exogenous changes in the growth rate of the money supply thought experiments. They have temporary effects on short term interet rates, but this effect disipates quickly.

    Well, what if the central bank is changing base money however much is needed to keep short term rates where they want them, And they are very much depending on people expecting an offsetting contraction if necessary sometime before inflation rises?

    Oh, and with inflation targeting, if the inflation rate rises above target, that just means they will slow it down in the future so it will no longer continue to rise so fast.

  33. Gravatar of dwb dwb
    26. March 2012 at 07:05

    this is indeed a very complicated issue.

    the only part i do not agree with is that “Fed interest rate smoothing has little effect on long term interest rates.” it definitely does via expectations.

    Econ desks world wide have some version of an “inertial fed funds model” where they attempt to forecast the change in fed funds at the upcoming meeting conditional on the change in fed funds the previous meeting . {in normal times when the ZLB does not apply of course, and this is used alongside other tools like fed funds futures, taylor rule regressions, etc). There is an 80-90% chance that the change in fed funds at the upcoming meeting is the same as the previous meeting.

    In terms of policy, the Fed could: a) announce a 250 bp level for real rates (which may mean more than 250 bps in nominal rates) that it expects to sustain for 3 years; or b) announce a 25 or 50 bp hike, with the expectation they will keep hiking 25 or 50 bp as needed until real rates are whatever they need them to be (vague!)

    under both policies long term real rates rise at least 250bps.

    policy b) is what they have chosen in the past because the FOMC is loathe to tie themselve down to a fixed period or conditional statement {I am not agreeing with this policy, just stating this is how they have behaved in the past; there are a variety of reasons postulated like policy acts with lags, they dont know the state of the economy, and they want to control long rates}.

    How does policy b) with an indefinite period affect longer rates? expectations.

    the market knows the tightening cycle (based on past behavior) is 2-3 years, with 25 bps increment each meeting, with an “on-hold” period after that (i.e. they will hike 25 bps at a time until real rates hit 2%-3% and then stay on hold).

    Since longer bonds are merely the average of shorter rates, the curve will steepen (with the steepness a function of how long the market expects to tighten, inflation expectations out the term, and so on).

    Of course, by the time the Fed actually gets around to hiking, the curve has already anticipated this so the curve will already be pretty steep and only the timing of the first hike is usually in question (Fed policy forecasting is a cottage industry).

  34. Gravatar of Jon Jon
    26. March 2012 at 07:12

    Scott, 123:

    In fact, there was a period when the Fed wanted to control the FFR, but failed. The result was the Great Recession:
    http://themoneydemand.blogspot.com/2010/11/collapse-of-fed-funds-rate-peg-and.html

    Taylor had a nice early paper on the LIBOR-OIS spread “A Black Swan in the Money Market”.

    Here are a few points to consider:
    1) The spread of CP rates (of both financial and non-financial institutions) over OIS was substantially muted compared to the LIBOR spread. While LIBOR is advertised as representing bank funding costs; it hasn’t been an accurate index since the 90s. The CP rates are considered more reliable indicators.

    2) The decoupling of LIBOR from actual costs arose because of LIBOR+% adjustable rate debt. Here bank customers were given a rate tied to LIBOR plus a premium. By 2007, these contracts were exclusively seen in the consumer credit space as it was widely understood by ‘smart money’ that the LIBOR was not a market-rate but rather manipulated to scam unsuspecting borrowers. Actually there is a recent prosecution relating to this owing to a little too much overt manipulation.

    3) When we discuss whether the Fed is steering rates, surely we’re talking about the risk-free rate. Even if you believe LIBOR was a good proxy, the accepted explanation for the spread is risk premia.

  35. Gravatar of Bill Woolsey Bill Woolsey
    26. March 2012 at 08:10

    The claim that the Fed was unable to control the Fed funds rate really meant that the Fed was unable to control the LIBOR rate?

    I wasn’t aware of any claim that the smart money consider the LIBOR rate a scam. But one thing was clear–it was a pretty limited sample.

    If you assume that a handful of money center banks borrow at the lowest interest rate, and count that as some kind of “base,” then when many of those money center banks happen to have difficulties at the same time, and they are borrowing at relatively high rates, then you just need to recognize that conditions of have changed.

    The measured Federal funds rate didn’t rise during 2008. (Not that I remember.) But then, lots of boring little commercial banks with mostly FDIC insured deposits continued to borrow and lend to one another.

    The “important” money center banks, who have a god given right to borrow at approximately the same rate as the U.S. Tresury, and whose borrowing costs make up LIBOR, had to pay remarkably higher rates than the U.S. Treasury or even the tiny little commercial banks. Heaven forbid! Crisis!

    Perhaps this is unfair, but it was that episode that led me to wonder to what degree interest rate targeting isn’t about making sure that the major money center banks, including investment banks, can borrow short term at stable interest rates.

    Also, I do understand that interest rate smoothing usually refers to making step adjustments in interest rates. I have the crazy notion that a central bank should not set an interest rate at all and should let them fluctuate however much is needed to clear markets.

  36. Gravatar of 123 123
    26. March 2012 at 08:55

    1. I was aware that LIBOR is a scam. Actual unsecured interbank interest rates were much higher. LIBOR was a convenient proxy that was biased against my argument. My argument is stronger because LIBOR is a scam.
    2. Taylor rule “worked” because FFR was close to the interest rate of M2.
    3. FFR targeting has nothing to do with “a god given right to borrow at approximately the same rate as the U.S. Treasury”. The output of Taylor Rule formula refers to FFR, not to the T-Bill rate. A decent Taylor Rule central bank should let T-bill rates adjust freely, and simply focus on keeping the FFR peg working. Trichet understood that (even though he was more monetarist than taylorist), and as a result the damage to labour market was smaller in the eurozone in 2008-09. If Ben Bernanke had kept FFR at 2%, and T-bill rate was at minus 1%, the damage caused by Taylor Rule would have been smaller.

  37. Gravatar of Jon Jon
    26. March 2012 at 09:02

    Bill,

    The LIBOR situation has been known about for some time. Investigations from the SEC and others have been ongoing and just a week ago there was a civil action filed against DB, although those are a dime-a-dozen.

    http://online.wsj.com/article/BT-CO-20120320-707783.html

  38. Gravatar of Jim Glass Jim Glass
    26. March 2012 at 13:58

    Jim Glass, I agree that rates are controlled via changes in the money supply, but also through signaling of future money supply changes.

    Sure — but the signaling effect depends on the credibility of the one sending the signal. In 1981 the Fed’s credibility was shot so Volcker had to go “back to basics” so to speak. After he scorched the nonbelievers, markets started reacting to the Fed’s announced intentions with ever less actual buying and selling by the Fed needed.

    Taylor, IIRC, tells a story of the crisis days in 1981 having drinks in a Washington bar after work with Volcker. They overheard some big money manager nearby them saying “The Fed will never do what they just said, I’m going to loot all the fools who believe it will.” Volcker said to Taylor, “I wonder where he’s going to be sending his resume three months from now.”

  39. Gravatar of Cantillonblog Cantillonblog
    26. March 2012 at 18:52

    After every financial crisis there is a mass pscychological tendency to make those seen to be the culprits pay. These days we do not see an angry mob chasing and partially eating the Prime Minister, as once happened in Holland. But the emotional dynamic is similar. One needs to consider the LIBOR witch-hunt in that context. In practice one problem was that there as no market for unsecured term funding between banks. How is one supposed to quote a rate when that instrument doesn’t trade? You have to make it up somehow, and it’s normal that you talk to your peers to see how other people are thinking about the problem. Was there a degree of gaming of the number to set it according to what would be best for the firm? Of course, and this is the nature of banks who are no longer permitted by the prevailing consciousness to be “long-term greedy”, but are forced to act to appear to do be doing whatever it takes within the letter of the law to maximize short-term value at all times. Calling it a scam when you have never traded the market and weren’t there at the time is really just adding to the mindless, narcissistic and reductionist chatter that characterizes our time.

    Scott, I think you misunderstand the relationship between implied market pricing and Fed actions. When market participants starts pricing rate cuts more aggressively than the Fed itself expects to move (as revealed by the Fed’s rhetoric at the time, and its after the fact minutes), it is because the market sees further into the future and more clearly than the Fed. Nobody forces the Fed to do anything – it’s just that the market – as a dead white guy called Hayek pointed out – tends to incorporate many dispersed pieces of information, much of which could not be made known even in principle to a single mind, or group of minds such as the FOMC.

    It is well-known to old hands that the stock market is much better at discounting future economic fundamentals then today’s fundamentals are at predicting the stock market in the future. Well the same is true of the bond market.

    I suggest the following exercise for those who may be interested. Pick a currency zone, and pull up a weekly chart of the 5 year swap or high quality government bond yield for that market. Identify the major highs and lows. Then go back and read what central bankers were saying at the time – both in public, and in their minutes (when released with a lag). You will see that the market always turns ahead of the central bank, and the central bank always tends to resist believing changes that are starting to be discounted by the market. Central bankers are always trying to drive the car by looking through the back windscreen. Hence Bernanke’s concern about deflation in 2003 (as the world economy was about to recover), and the ECB’s concern about inflation in July 2008 (just as the wheels were about to fall off). So it’s just not right that the market bullies the Fed – the market is smarter than the Fed, and the Fed eventually has to recognize the market’s insight.

    It’s funny and says a great deal about the prevailing mentality that people seem incapable of correctly typing the name of Cantillon – one of the most insightful writers on the non-neutrality of money.

    “Obviously I agree with all that, which makes me think we are talking past each other.”
    I believe that I understand you very well, and that if the world were as you think it is then your argument would hold perfectly and it would be astonishing that nobody had ever recognized that point previously. However, I think the world is messier in a way that doesn’t really conform to your model and I regret to say that although I am pretty confident your doctrine is perfectly suited to become tremendously fashionable, I think its effects were it to be adopted would be most noxious, which is why I am determined to oppose it. I think the reason why it is likely to be professionally successful is that it is very much in tune with the mentality of our time, one that tends to be very abstracted, analytical and oriented towards control. Iain McGilchrist treats this at book length in “The Master and His Emissary”.

    ” Let me try a different tack… No one would dispute that they control the fed funds rate for that six week period, and no one would dispute that they could extent that period longer, if they wished. But there is a tension there. The more (arbitrary) control they have over the fed funds rate the less control they have over inflation, or their other goal variables. If they are serious about keeping your own internal forecast of inflation on target, the interest rate become endogenous.”

    Yes, exactly. As I said above, if the Fed is a machine and the economic facts are given, and its forecasts are okay then it has no real freedom as to how to set the interest rate if its mandate is given and it takes it seriously. This is a lovely engineering negative feedback model. The question is not whether the model is lovely, but whether it usefully describes the world we inhabit. I think it is dangerously wrong.

    And certainly I would agree that if the Fed wants to target a particular inflation rate, it cannot set Fed Funds capriciously (as your President once did to the gold price over breakfast). “Zero is so boring; seven is my lucky number. Lets move the rate up to 0.07% today”. The Fed cannot do whatever it wants given an unambiguous mandate. I remember reading this cybernetic argument in my macro textbook more than twenty years ago, and it was old then.

    But recognizing this aspect in which your argument is well-founded – an aspect that ought to be trivially obvious – one must move on to consider the broader case.

    “And it makes no different whether your goal is different from inflation, and it makes no difference how bad your models are. There is a tradeoff between control of the fed funds rate and control of inflation, or whatever the goal variable actually is.”

    To me this is a**-backwards. You seem to be implying that the Fed has a taste for a particular Fed Funds rate independent of its goals. Bernanke doesn’t strike me as a keen numerologist. The rate is just a number the Fed can twiddle up and down in order to influence the economy so as to try to achieve its objective.

    “But it’s even worse, as the less effective the control of inflation, the more the T-bill yield will diverge from the fed funds target.”

    I don’t mean to be unscholarly, but the simplest way to describe this argument is that it is plain silly. The Fed doesn’t intend to control T-bill yields, any more than most of the time it aims to control yields on the 2 year note or the 30 year bond. (We are talking about the regime pre and post QE, right). It all but determines the effective fed funds rate for overnight money. (Deviations are rare and non-material to my argument). If it wants to manipulate the short term yield curve it can jawbone the market or use open-mouth policy.

    ” In December 2007 there was a contractionary surprise out of the Fed, and yet 3 month T-bill yields actually fell. That’s because the market saw that the Fed had blown it, by cutting rates by only 1/4%, instead of the 1/2% hoped for. So the market anticipated the Fed would see its error and cut rates much more sharply in the near future. And that’s what happened, as the Fed cut rates by 125 basis points over the next 6 weeks.”

    Yes – exactly. I wouldn’t say the market told the Fed it had blown it – that’s implying a degree of normative assessment that was not present in prices. But yes the market saw that the Fed would need to be cutting rates.

    ” I don’t think anyone in their right mind would call that panicky cut in rates “Fed control” over short term interest rates-the markets forced the Fed’s hand. Sure, if they wanted to be stubborn they could peg rates forever, at the cost of hyperinflation or hyperdeflation”.

    That’s like your teenage son telling you that he is not truly in control of the car because he does not have the freedom to steer the car according to his whim because the need to stay on the road constrains his autonomy and freedom.

    Does the steering wheel control the car, or does the road? If you have a perfected cybernetic robot driving the car, it may make sense to speak of the road as driving the car. But alas our state of monetary knowledge is not quite at that level (and never will be).

    It’s not the market that literally forces the Fed’s hand. It is that the market anticipates reality and the Fed must eventually see the same thing too.

    I have no idea what you mean about a debate with CA, but the opportunity cost of time being what it is, I am not at this stage particularly keen to engage this gentleman further. The point isn’t that it was possible to see that the Fed was in 2004 in a hiking mode – thats not that interesting a point given that this is what was priced in to fed funds futures. The point is that based on the right framework one could see in early 2008 that the Fed would have to cut rates loads and keep them low nad in July 2008 that the ECB would have to reverse its final hike and cut rates lots. Most observers of financial markets don’t have that focus in mind, being more concerned to explain the past than to anticipate the future. When one tries to think about what is likely to happen that changes the way one thinks about markets and the economy in a way that tends to lead to the development of insights that don’t come so naturally using the backward looking perspective.

  40. Gravatar of ssumner ssumner
    27. March 2012 at 05:16

    Bill, You said;

    “But I don’t like interest rate smoothing, and your argument seemed to ignore it.”

    First, I think most people who believe the Fed “controls” rates have in mind more than smoothing. Second, doesn’t smoothing actually make rates more unstable? If you are targeting inflation, and are slow to raise rates when needed, then later you must raise them by even more than otherwise. Or am I missing something?

    Regarding the rest of your comment, it seems to me that it comes down to this;

    If the expected future inflation or NGDP is always right on target, then rates are always at the proper level and are set by markets. To the extent that the Fed goes off course, the rates are off course, and can be thought of as “controlled” by the Fed. Maybe my argument isn’t exactly true for fed funds rate, but surely it’s roughly true for 6 month T-bill yields.

    dwb, See my response to Bill (above). This is my best guess, but it’s an open question.

    Jon, Thanks for that info. I’ll take your word for it as I haven’t following the Libor issue closely.

    Cantillonblog, You said;

    “Scott, I think you misunderstand the relationship between implied market pricing and Fed actions. When market participants starts pricing rate cuts more aggressively than the Fed itself expects to move (as revealed by the Fed’s rhetoric at the time, and its after the fact minutes), it is because the market sees further into the future and more clearly than the Fed. Nobody forces the Fed to do anything – it’s just that the market – as a dead white guy called Hayek pointed out – tends to incorporate many dispersed pieces of information, much of which could not be made known even in principle to a single mind, or group of minds such as the FOMC.”

    With all due respect I still think you are missing my point. I agree that the Fed can thumb it’s nose at the markets, but only at the expense of allowing AD to drift off course. My point is that if they don’t want AD to drift off course, they pretty much have to set rates where the market thinks they should be. I certainly don’t claim the fed funds rate is always exactly where the market thinks it should be, but rather over time an inflation targeting central bank will move them in that direction, and T-bill yields will reflect that expectation.

    Let’s use the famous “low interest rate policy” of 2002-04 as an example. The conventional view is that the Fed set interest rates at a very low level. That’s what people mean by “control.” I say the market drove interest rates to a very low level because there was little business demand for credit in the years immediately after the tech bubble burst. Those are two very different ways of explaining the low interest rates. Do you think it was the market or the Fed? That will help me understand where we disagree.

    BTW, I agree the Fed could have kept rates at 5% during that period, but only at the cost of Great Depression II.

    The problem with your steering wheel analogy at the end is that the Fed’s steering wheel isn’t rates, it’s the money supply (base). It targets rates just like it targets inflation. You’d be on stronger ground claiming they control the monetary base, or reserve requirements. That is the steering wheel.

  41. Gravatar of Major_Freedom Major_Freedom
    27. March 2012 at 05:29

    ssumner:

    “For better or for worse, short term rates are tied to the fed funds rate, so while the Fed doesn’t “control” short term rates on things like t-bills, they nevertheless strongly influence them. I can only see the Fed failing to influence short term rates like t-bills, if they lose control of the currency.”

    That correlation breaks down when the Fed fails to successfully target it goal variable. See my discussion of the December 2007 meeting above.

    The Fed doesn’t fail to target its goal variable though. That’s the point. The correlation between targeted fed funds rate and effective funds rate is pretty much 1.00.

  42. Gravatar of dwb dwb
    27. March 2012 at 06:22


    Second, doesn’t smoothing actually make rates more unstable?

    I define “stability” as volatility, the standard deviation of the daily changes in rates over given time period.

    In fact, long rates are much more volatile than short rates, so yes this is in fact what you see. Only some of this is policy driven, though. The rest is driven by economy-wide supply and demand for investment.

    The flip side of discrete jumps (no smoothing) is that the FOMC has to commit to a time period with which to hold that level.

    If I understand your argument, there is a given (equilibrium) real rate of return associated with a given nominal income growth rate, the Fed has an inflation target, so there is a given level of nominal rates to hit that target. Since long rates drive the economy, there are many paths for short rates (fed fund) that work out, on average, the correct level of nominal rates. The Fed does not “control” real rates, so the fed does not “control” interest rates. I cannot opine on what people mean when they say the Fed controls rates, because “control” is clearly not a well-defined term. If someone says they control rates, they have set out for me a policy path, terminal condition, and policy goal. maybe they “control” it, maybe they don’t depending on what they actually mean.

    Getting back to rates instability, I think of the total volatility in rates as the sum of a)Fed policy uncertainty; b) real rate volatility (a function of real investment supply/demand); c)inflation volatility (even if the Fed has a 2% target, we know they’ll let it wiggle a bit).

    Whether “smoothing” reduces or increases the sum of a+b+c depends on the correlation between Fed policy reaction to unexpected events and the other two, so its really a function of Fed policy reaction itself and how they set expectations.

    At a higher level, though, “smoothing” and fed funds are more about signalling IMO than anything else. Fed Funds does not drive the economy (longer rates drive investment decisions). So the whole policy path issue is more about signalling that the FOMC is (un)comfortable with the nominal growth path.

  43. Gravatar of Dhruv Dhruv
    27. March 2012 at 13:38

    In Friedman’s defence, he absolutely recognized that fixing exchange rates or managing interest rates are equivalent and equally interventionist. However, he felt that it was better to allow exchange rates to float as exchanges rates impacted principally the export-import industry, which in 1960s was less than 10% of GDP.
    Managing interest rates (and preventing wild fluctuations due to fixed exchange rates) would benefit wider economy.

  44. Gravatar of ssumner ssumner
    28. March 2012 at 10:30

    MF, I was talking about T-bills, not the ffr.

    dwb, That sounds reasonable, although we’d have to pin down a precise model to fully resolve this issue.

    Dhruv, I wonder if he recognized this after someone pointed it out to him. He was certainly criticized by Mundellians for seeming to overlook that point.

  45. Gravatar of Jim Glass Jim Glass
    2. April 2012 at 18:18

    I once heard Greenspan explain it this way:

    Krugman just said it the exact same thing. The Fed can set the money supply and have the corresponding interest rate pop out, or set the interest rate and have the corresponding money supply pop out. It chooses to set the interest rate rather than the money supply (in normal times, not in 1981 or at the zero bound today) for the practical reason that a stable interest rate with the money supply bouncing around due to short-term variations in demand is less disruptive to business than the reverse, the interest rate bouncing around with a stable money supply.

    I mention this now in a thread nobody is reading any more in case somebody comes onto it via a search engine or reading the archives or whatever, since I didn’t give a link to Greenspan saying it.

    http://krugman.blogs.nytimes.com/2012/04/02/a-teachable-money-moment/

  46. Gravatar of ssumner ssumner
    2. April 2012 at 18:37

    Thanks Jim Glass.

  47. Gravatar of Shining Raven Shining Raven
    19. December 2012 at 11:59

    Whoohoo, one more post on which I completely missed the discussion as it was occurring. Still adding my two cents:

    I am firmly in the camp that believes that, short term, what a central bank does is setting short-term interest rates. But alas, ‘t was not always so, and some people seem to have missed the boat that things have changed.

    I highly recommend a paper by Ulrich Bindseil of the ECB who gives a nice account of where this confusion comes from. Seems to be mainly a disease of American economists, since the Fed frequently changed what it actually did target over the course of the 20th century.

    Ulrich Bindseil: The operational target of monetary policy and the rise and fall of reserve position doctrine

    http://www.ecb.int/pub/pdf/scpwps/ecbwp372.pdf

    “Today, there is little debate, at least among central bankers, about what a central bank decision on monetary policy means: it means to set the level of short-term market interest rates that the central bank will aim at in its day-to-day operations during the period until the next meeting of the central bank’s decision-making body.” …

    “It appears that with Reserve Position Doctrine, academic economists developed theories detached from reality, without resenting or even admitting this detachment. Economic variables of very different nature were mixed up and precision in the use of the different concepts (e.g. operational versus intermediate targets, short-term vs. long-term interest rates, reserve market quantities vs. monetary aggregates, reserve market shocks vs. shocks in the money demand, etc.) was often too low to allow obtaining applicable results. The dynamics of academic research and the underlying incentive mechanisms seem to have failed to ensure pressure on academics to ensure that models of central bank operations were sufficiently in line with the reality of these operations.”

  48. Gravatar of Shining Raven Shining Raven
    19. December 2012 at 12:18

    Scott writes: “The problem with your steering wheel analogy at the end is that the Fed’s steering wheel isn’t rates, it’s the money supply (base). It targets rates just like it targets inflation. You’d be on stronger ground claiming they control the monetary base, or reserve requirements. That is the steering wheel.”

    That is unambiguously wrong. In fact, the Fed funds rate is the “steering wheel” in the analogy (the operational target). The inflation rate is the actual target. There is no tradeoff between setting the Fed funds rate and controlling the inflation rate. The Fed funds rate is a policy tool that is set to achieve an inflation target in the long run.

    It’s not true that the Fed “targets rates just like it targets inflation”. Conceptually this is on a very different level. The Fed does care about inflation, it does only care about the Fed funds rate insofar as it lets it achieve the inflation target. It does not at all care about the monetary base.

    This is exactly the kind of confusion of long-term and intermediate targets with operational targets that Bindseil talks about in the paper I linked above.

  49. Gravatar of Major_Freedom Major_Freedom
    19. December 2012 at 12:57

    Shining Raven:

    It’s not true that the Fed “targets rates just like it targets inflation”. Conceptually this is on a very different level. The Fed does care about inflation, it does only care about the Fed funds rate insofar as it lets it achieve the inflation target. It does not at all care about the monetary base.

    Agreed. Using A to target B means A and B are on different conceptual levels.

    But when the Fed funds rate is at zero, the Fed can’t use A to target B anymore, and has to resort to alternative means.

Leave a Reply