Reply to Karl Smith

Karl Smith recently commented on my post about Hume and Woodford. I had argued that during “normal times” (when rates are positive) the future path of the base is a much better indicator of the future path of NGDP than is the future path of interest rates.  (I should have added that exactly the reverse is true at the zero bound.)  Karl responded:

Are you simply saying that we can’t solve for the interest rate path and therefore it is difficult to set up an effective communications strategy?

I think that’s true though we have to be aware the problem works both ways. The actual economy consists of millions of different economic agents all of whom face an interest rate that is based off of the Funds rate or the Interest on Reserves rate. However, each of their relationships to NGDP is less clear.

To know what this policy means they – or more realistically their banker – has to solve backwards.

For example, suppose I am expanding a hot new burrito chain in East Texas. Does a 30% higher NGDP for the US in 2017 mean I – and my banker – want to expand faster or slower?

Its not clear. The overall growth path of NGDP for the US is only loosely connected the actual revenue that I am going to receive. A booming national NGDP might mean I grow at 14% nominal rate rather than a 11% nominal rate. Though it could go the other way if my customers get lured away to work in West Texas.

However, the interest rate over this period is going to determine whether dumping a bunch of money into new restaurants is a good idea or not. Is that going to be higher or lower because of your target. Again, its not clear.

So, its not clear what I should do.

On the other hand if you said – interest rates will be zero through at least 2014, then it is at least certain that no one is getting any surprises over the construction loan period and that more than likely some decent loan is going to be available when my stores are done.

I have two problems with this.  Any macro signaling will seem to have a trivial effect if discussed at the micro level.  Let’s suppose the Fed did something now that caused the representative firm to boost output by 2% more than on the old Fed policy.  That’s a really big deal for the macroeconomy, but small potatoes for the representative firm within the economy (as 2% more business on average is swamped by local real shocks.)

But my bigger problem is Karl’s claim that interest rates are of more interest to the average business owner than the monetary base.  Suppose Karl and I both looked into a crystal ball and saw that the fed funds rate would be 12% in 2014.  From Karl’s post, I infer that he’d advise that small businessman to hold off on the investment project, as the cost of that floating rate business loan would soar in 2014.  I’d have the opposite reaction.  I’d beg, borrow, and steal every penny I could get my hands on, and pour all the money into REITs.  That’s because the 12% rate in the crystal ball would tell me I am wrong and Bob Murphy is right—that the inflated monetary base is going to drive inflation and NGDP dramatically higher in 2014, forcing the Fed to raise rates sharply in order to hold inflation down.  I see that as incredibly bullish for real estate.

To summarize, not only is the interest rate path not a good indicator of whether you want to invest more or not, it’s not even clear whether low rates are a good sign or a bad sign.  It’s hard to think of any other indicator that’s worse.  Most asset values at least have a monotonic relationship with NGDP growth, but not short term nominal interest rates.

BTW, there is an interesting discussion of exactly what the Fed’s low rates until 2014 promise (or prediction?) really means.  Romer and Romer have a good PP slide that correctly explains the flaw in the Fed’s policy.  But the art of signaling in a messy real world with a divided FOMC is about as subtle and complex issue as exists in all of macroeconomics.  So after reading Romer and Romer I’d recommend Ryan Avent, who I think strikes exactly the right balance.  The Romer’s are right that the policy is not what it might seem to be, or even close to what it should be, but as Ryan points out it’s probably something.

PS.  Ryan is also the only one who seems to understand what that “silly” OECD graph really means.  BTW, people who take medication tend to be sicker than those who don’t.


Tags:

 
 
 

16 Responses to “Reply to Karl Smith”

  1. Gravatar of dwb dwb
    28. February 2012 at 15:21

    well, Karl Smith is about half right – both matter – and neither.

    first, the average analyst evaluating a project does not care a whit about the fed funds rate (or the monetary base, sorry). what matters is the all-in cost (treasury+credit+any adders) of financing (with a maturity that roughly matches the project) for the discount rate, the debt/equity ratio, and the cost of equity. …

    second, the pace of revenue growth will matter (and cost growth), each specific project will be tied to a project specific revenue/cost growth assumption. But at a high level it will still be governed by the pace of ngdp growth.

    i.e. a basic static analysis would need both (a more sophisticated analysis might look at a distribution of NPVs around various scenarios)

    And while this is right for your basic capital investments 101 taught to an MBA, this of course is all wrong for macro analysis: interest rates also determine consumptions/savings decisions of households (consume now vs later)…. which is why for macro you really need something more sophisticated where the equilibrium is endogenous.

    in other words, why do we care how much of the P*Y comes from consumption generated from existing unused capacity, and how much comes from genuine new project-level investment? I do not.

  2. Gravatar of Major_Freedom Major_Freedom
    28. February 2012 at 15:21

    Ah, Karl Smith gets it. He knows that NGDP isn’t something that individual investors and sellers actually invest or sell into, like I’ve been saying repeatedly here.

    A growing NGDP could accompany a decrease in sales of Edsels and hoola hoops and season three of Wings. Investors contemplating where and what to invest in, and sellers contemplating where and what to sell, get influenced more by prevailing interest rates and individual market demands, than they would NGDP.

    ssumner:

    I have two problems with this. Any macro signaling will seem to have a trivial effect if discussed at the micro level. Let’s suppose the Fed did something now that caused the representative firm to boost output by 2% more than on the old Fed policy. That’s a really big deal for the macroeconomy, but small potatoes for the representative firm within the economy (as 2% more business on average is swamped by local real shocks.)

    Wow. What confusion. You’re not even addressing micro-economics when you talk about a macro concept called “the representative firm.” A micro-economics discussion would talk about “this particular firm here, in this location, in these circumstances, with these demands, and these costs of production, etc.” Saying “the representative firm” and imagining all firms collectively, is still in the macro world. You’re just talking about the output of N firms increasing by 2%, which is N x 1.02 total output = Y, which is a macro concept.

    This statement: “Any macro signaling will seem to have a trivial effect if discussed at the micro level” is totally off the wall wrong. If macro policy did not have any effect at the micro level, it couldn’t even qualify as an economic policy. The economy IS the micro level. That’s where ALL the activity takes place. If macro policy didn’t signal anything to the micro level, then the Fed’s policy, and the Treasury’s policy, would be vacuous. Clearly that’s all wrong.

    But my bigger problem is Karl’s claim that interest rates are of more interest to the average business owner than the monetary base.

    You just straw manned Smith and deftly moved the goal posts away from NGDP and to the monetary base. Nice.

    Now address the comparison between interest rates and NGDP, you know, what Smith was actually talking about.

    Suppose Karl and I both looked into a crystal ball and saw that the fed funds rate would be 12% in 2014. From Karl’s post, I infer that he’d advise that small businessman to hold off on the investment project, as the cost of that floating rate business loan would soar in 2014. I’d have the opposite reaction. I’d beg, borrow, and steal every penny I could get my hands on, and pour all the money into REITs.

    [Facepalm] That isn’t an opposite reaction, that is you granting that the small business investment should be postponed, and that you should CHANGE YOUR INVESTMENT BEHAVIOR, and invest in something else instead!

    You’re proving Smith’s point, and you don’t even know it.

    Smith is saying that individuals are more influenced by interest rate policy than they are by NGDP policy. In response to this, you criticize Smith, and say that he’s wrong, on the basis that your investment policy would be…. HIGHLY affected by interest rates. I mean, you even added the words “beg, borrow and steal” to boot. You could not have made a stronger case in favor of Smith’s point!

    This is the funniest thing ever.

    Now contrast your incredibly forthright investment reaction to interest rates, to you having a crystal ball that says “NGDP will rise by 12% in 2014.”

    Now what will you do? Would you “beg, borrow, and steal” to do something that you wouldn’t have done if the crystal ball said “NGDP will rise by 5% in 2014″?

    Where will you invest? What will you invest in? What will be different?

    I think this is the most revealing post you’ve ever made on the confusion of fetishizing over NGDP targeting. Not only did you completely grant Smith’s point, but you actually believe you successfully refuted him. This is too funny.

    To summarize, not only is the interest rate path not a good indicator of whether you want to invest more or not, it’s not even clear whether low rates are a good sign or a bad sign.

    First off, low interest rates is what led to the boom in housing. It’s why people who couldn’t have otherwise afforded a home with higher borrowing costs, could afford them with lower borrowing costs, thus creating the housing boom.

    Secondly, you just admitted that interest rates will have a huge impact on your behavior. You just said that you will “beg, borrow, and steal” on account of certain interest rates!

    Why didn’t you give an example of how NGDP has a greater affect on what you do compared to interest rates? Is it because you deep down know what I and now Smith knew all along, which is that NGDP is something no individual investor or seller seriously takes into serious consideration when making business decisions, and that they take into account interest rates far more seriously?

    It’s hard to think of any other indicator that’s worse. Most asset values at least have a monotonic relationship with NGDP growth, but not short term nominal interest rates.

    Absolutely false. Asset values are affected by interest rate changes far more than consumer goods. If there is an asset boom, and the REAL savings rate hasn’t increased, then chances are it’s because of low interest rates and credit expansion.

    Asset values and NGDP growth are both influenced by the same prior cause. One doesn’t cause the other. They are parts of the same process of increased money and spending.

    Why do you try to shoehorn every last economic event into what’s happening with NGDP?

    No investor even takes it seriously, not even the Fed. Even if the Fed adopted NGDP targeting, investors would still be more influenced by interest rates than they would NGDP.

  3. Gravatar of Morgan Warstler Morgan Warstler
    28. February 2012 at 16:14

    “Any macro signaling will seem to have a trivial effect if discussed at the micro level.”

    Not under 4% NGDP level targeting.

    The “signal” predicted by futures markets etc. will specifically rile up the East Texas Burrito maker IF he sees NGDP going up because public employees got a raise.

    He and all the other likely voters will become NGPD future watchers – they will want ALL growth to come from REAL GDP in the private sector driven by productivity gains.

    Fake public sector growth and inflation will be hated

  4. Gravatar of ssumner ssumner
    28. February 2012 at 17:13

    dwb, I mostly agree.

    Major Freedom, So low interest rates cause assets price booms? Have you looked at real estate prices since the Fed adopted a low interest rate policy in 2008? And how did real estate prices do during the high interest rate 1970s?

    Morgan, They may pay attention, but local shocks will still dominate for the individual firm.

  5. Gravatar of Mark A. Sadowski Mark A. Sadowski
    28. February 2012 at 18:42

    But the art of signaling in a messy real world with a divided FOMC is about as subtle and complex issue as exists in all of macroeconomics.”

    This is about as close as I can get to stupidity without letting it rub off on me.

    The key to getting out of this mess is to stop this mindless focus on INTEREST RATES!

  6. Gravatar of Morgan Warstler Morgan Warstler
    28. February 2012 at 20:26

    If all you are banging on is local shocks, I’ll trust the Internet, and only 55% US home ownership to smooth those out.

    Thanks for punting on the big stuff.

    You’d think you’d say out loud more often that under NGDP level there won’t be much in the way of “shocks.”

  7. Gravatar of Mark A. Sadowski Mark A. Sadowski
    28. February 2012 at 20:30

    Morgan,
    55%? Try 15%.

    Most Americans aren’t mortgage free (like me).

  8. Gravatar of Major_Freedom Major_Freedom
    28. February 2012 at 22:08

    ssumner:

    Major Freedom, So low interest rates cause assets price booms? Have you looked at real estate prices since the Fed adopted a low interest rate policy in 2008?

    Yes. Home prices would have been a LOT lower had the Fed not imposed low interest rates. Prices would have collapsed even more.

    Besides, low interest rates don’t always affect the same exact assets the same exact way every time. Humans aren’t robots. After the biggest housing collapse in history, it should be expected that mortgage lenders would not be as eager, especially with so very many unoccupied houses on the market, keeping a lid on those prices.

    One cannot predict exactly where the new money will go, but one can know that it will gravitate towards those investments most susceptible to interest rates, including houses (which is why prices are no longer falling), commodities, stocks.

    Real estate is not the only asset. That Bernanke’s low interest rates have been producing a boom in asset prices doesn’t mean one has to see a temporal nominal increase. It means higher than it otherwise would have been in the absence of low interest rates! There was an inflation financed boom of the 1920s despite the fact that nominal consumer prices were more or less stable. Much of the new money went into the stock market.

    More recently, the S&P500 alone has almost doubled since 2008. That’s something like $9 trillion. You think the real savings rate justifies that kind of spike? It’s Bernanke’s printing.

    Just look at producer prices since 2008 as well as commodity prices since 2008.

    And how did real estate prices do during the high interest rate 1970s?

    When the fed funds rate fell, a year or so later (since there is always a time lag between change in interest rates and the effects) commercial loans to real estate heated up, and when the fed funds rate increased, a year or so later commercial loans to real estate cooled down.

    The same thing occurred in the 2000s

    I used Case Shiller for 2000-2010 because the data is available, but it clearly shows consistency with the theory that when Bernanke lowered rates, a year or after home prices increased, and when he increased rates, a year or so after home prices peaked and then fell. Notice that after a year or so of decreasing rates once again after 2007, home prices stopped declining. They would have kept declining had Bernanke not decreased rates.

  9. Gravatar of ssumner ssumner
    29. February 2012 at 06:02

    Mark, That’s right.

    Morgan, You are wrong.

    Major Freeman, No with 5% NGDP targeting both interest rates and house prices would be much higher. High nominal interest rates are strongly correlated with high NGDP growth rates.

    Stocks are still well below 2007 or well below 2000. Yes, the market is more optimistic than in the dark days of early 2009.

    But the level isn’t that high.

  10. Gravatar of Ron Ronson Ron Ronson
    29. February 2012 at 07:38

    I’m a bit confused by the apparent comparison of expectations of interest rate and monetary base as good indicators of future economic conditions.

    In a healthy economy where the money supply matches the price level then interest rates will reflect the demand and supply in the loanable funds market and this will lead to an optimal use of those funds for investment.

    If there is a demand shock that leads to a fall in NGDP then if prices do not adjust quickly we will get a situation where the money supply does not match the price level, expectations of future demand will fall. The supply of loanable funds will rise just as the demand for them falls and we hit the zero bound.

    How should the monetary authority react ?

    If it says “we will hold interest rates at zero for the next 5 years” It is admitting failure before it has even begun. Zero interest rates are a symptom of an economy in recession.

    If it says “We will double the monetary base in the next years” it is making monetary policy indeterminate. No one can say what effect on NGDP or RGDP (or interest rates) this policy may have.

    The monetary authority needs to communicate its commitment to optimal money supply – the level that leads to stable NGDP growth. This will lead to interest rates at their natural level – the level that will allow banks, business and individuals to choose the best savings and investment decisions unhampered by the uncertainty caused by indeterminate monetary policy.

  11. Gravatar of Major_Freedom Major_Freedom
    29. February 2012 at 10:28

    ssumner:

    Major Freeman, No with 5% NGDP targeting both interest rates and house prices would be much higher. High nominal interest rates are strongly correlated with high NGDP growth rates.

    No, you’re ignoring the fact that the Fed, intentionally or not, lowers “market” interest rates by its inflation into the banking system, in order to influence the banks in lending and thus affecting NGDP. If the Fed only does one round of inflation, then we might see a temporary decline in interest rates (as usual), but then once the new money is spent and respent, profits will rise and that will pull interest rates up. If the Fed wants to boost NGDP, then it may well have to inflate so much that short term rates fall to zero.

    When they decrease/increase interest rates through inflating faster or slower, then intentionally or not, there is a delay in changes to aggregate spending. Once it affects aggregate spending, that’s when profits rise, and once profits rise, nominal interest rates will then rise too. That’s why you see a positive correlation between interest rates and NGDP growth.

    Look at the charts I linked to again. You see a positive correlation between interest rates and commercial real estate lending and home prices because of the delay in monetary policy changes. It’s why the market crashed in September 2008 rather than June 2008, the latter month in which M2 was already growing at a snail’s pace compared to earlier in the year.

    Stocks are still well below 2007 or well below 2000. Yes, the market is more optimistic than in the dark days of early 2009.

    It’s not so much about “optimism”, or “animal spirits”, as it is about pure inflation induced nominal increases in spending.

    You’re just looking through your rear view mirror. You should be looking at aggregate monetary growth. That is how you can glean future trends.

    But the level isn’t that high.

    Isn’t as high as what? I only said it almost doubled since 2008. That is true. The S&P500 went from around 700 to almost 1400. Just because it is not at previous highs, it doesn’t mean Bernanke isn’t printing enough. We ought not be trapped to the past. If we were living in Zimbabwe, then should we consider present monetary policy as “not that high” because it isn’t matching previous rates?

    I don’t know how high the S&P will go, because it is of course contingent on the Fed, but if current monetary trends continue, then soon the S&P will be breaking nominal records.

    Of course it will positively correlate with NGDP growth, but that’s only because both NGDP and the stock market are effects of the same prior cause.

  12. Gravatar of david stinson david stinson
    1. March 2012 at 15:48

    Hi Scott.

    The fact that monetary authorities commit to NGDP targeting in no way reduces the information content of interest rates for purpose of capital investment planning. Central bankers are not the only source (if they are a source at all) of valid information about future interest rates.

    I would have thought that the market monetarist response to Karl Smith would have been something like the following:

    a) NGDP targeting is a means for maintaining monetary equilibrium.

    b) By avoiding discoordinating monetary disequilibrium, NGDP targeting would improve the information quality of prices (including interest rates) and permit continued real growth at the “natural” trend.

    c) If actually successful (and expected to continue to be successful) in avoiding monetary disequilibrium, NGDP targeting would reduce the potential for future monetary error, the need for any related risk premia (such as increased corporate default risk in the case of excess money demand) in interest rates and the potential damage to future (and thus current) investment environments such errors can cause. Entrepreneurs can thus focus on market sources of risk which are both their area of expertise and easier to diversify against than the (systemic) risks of monetary policy error.

    d) Under conditions of current and expected future monetary equilibrium, presumably the term structure of market interest rates would reflect, among other things, the future path of NGDP and RGDP growth at trend, and would provide useful information to entrepreneurs.

    If Karl Smith’s point was that committing to future interest rates somehow gives entrepreneurs more useful information than committing to an NGDP path, I would suggest that the most useful commitment central bankers could make would be that which ensured the lowest risk of monetary disequilibrium.

    Added to which, the analysis of capital budgeting depends not only on market interest rates but on expected cash flows. Expected cash flows are obviously influenced (often, but not always, explicitly) by expected economic growth.

  13. Gravatar of ssumner ssumner
    1. March 2012 at 17:46

    Ron, First of all, if the Fed does 5% NGDP targeting, level targeting, it’s extremely unlikely that the nominal interest rate would ever fall to zero.

    But let’s say I’m wrong. Then just issue enough base money to peg the price of NGDP futures contracts at the target NGDP. Let the market decide how much base money is needed. But I can’t emphasize enough that with level targeting this “problem” would never occur at all. I put “problem” in quotes, because in a way it’s good if the government can borrow the entire national debt at zero interest cost.

    Major Freeman, You said;

    “No, you’re ignoring the fact that the Fed, intentionally or not, lowers “market” interest rates by its inflation into the banking system”

    You have it backward–inflation affects interest rates, and in a positive direction. High inflation leads to high interest rates. Look at the 1970s. Or Latin America in the old days.

    David, You said;

    “The fact that monetary authorities commit to NGDP targeting in no way reduces the information content of interest rates for purpose of capital investment planning. Central bankers are not the only source (if they are a source at all) of valid information about future interest rates.”

    I completely agree, and favor that policy. But in that case the Fed has no control over interest rates (which I favor.) But Karl was discussing a case where the Fed was controlling interest rates. So he was assuming they weren’t controlling NGDP.

  14. Gravatar of Major_Freedom Major_Freedom
    1. March 2012 at 21:51

    ssumner:

    “No, you’re ignoring the fact that the Fed, intentionally or not, lowers “market” interest rates by its inflation into the banking system”

    You have it backward–inflation affects interest rates, and in a positive direction. High inflation leads to high interest rates. Look at the 1970s. Or Latin America in the old days.

    No, you are still confused. High inflation does lead to high interest rates, yes, but only if the money is entering the economy in a way other than through the loan market.

    The “high” interest rates in the 1970s would have been even higher if the Fed wasn’t inflating bank reserves and was instead just printing money and spending it in the greater economy. Because the Fed was inflating into the banking system, and from there into the loan market, it pulled interest rates down from where they otherwise would have been.

    Your view of interest rates and inflation is all messed up because again you believe NGDP is the be all and end all of everything. You are not taking into consideration how the money is entering the economy and how that affects interest rates. You’re just thinking “inflation premium on loans.”

    If I kept printing and spending money in the greater economy, then interest rates will rise. But, if I printed and LOANED the money out, then the supply of loans will rise, and that will push down interest rates. If I keep inflating more and more in order to “boost aggregate demand”, but I put the money into the economy via the loan market first, by lending it, then sorry, but interest rates will NOT rise. They will be pulled down. Once I stop printing and lending, or once I print but then spend instead of lend, then interest rates will rise.

  15. Gravatar of ssumner ssumner
    2. March 2012 at 18:14

    Major Freeman, You said;

    “No, you are still confused. High inflation does lead to high interest rates, yes, but only if the money is entering the economy in a way other than through the loan market.”

    No, even if money enters through the loan market, it causes higher inflation and higher interest rates.

  16. Gravatar of Major_Freedom Major_Freedom
    4. March 2012 at 18:16

    ssumner:

    “No, you are still confused. High inflation does lead to high interest rates, yes, but only if the money is entering the economy in a way other than through the loan market.”

    No, even if money enters through the loan market, it causes higher inflation and higher interest rates.

    Inflation that enters the loan market first increases the supply of loans. An increased supply of loans (temporarily) decreases interest rates on loans. It doesn’t increase them.

    Once the credit expansion is spent and respent in the economy, raising revenues and profits, then interest rates rise. But by that time, the Fed’s interest rate targeting will have already generated more credit expansion, thus maintaining the low interest rates.

    This is why to keep a particular low interest rate, the Fed has to accelerate inflation over time, but because it can’t make consumer prices rise too much, it stops inflating so much after some time, and that’s what reveals all the malinvestments that depended on the prior easy money.

    In the US, aggregate demand from 2001-2005 rose, and yet the Federal Reserve System brought about lower interest rates via credit expansion.

Leave a Reply