Friday, June 17, 2011

Money and Inflation

Since an FOMC meeting is coming up next week, and we have some more information on prices, it would be useful to review the current situation. In the first chart, I plot some price indexes, i.e. PCE deflator, headline CPI, and core CPI, relative to a 2% trend (the more-or-less explicit Fed inflation target) starting at the NBER-dated onset of the recent recession in December 2007. The inflation situation, by this measure, is not dire. Currently the PCE deflator is 1% below the 2% trend, headline CPI is 0.6% lower, and core CPI is 1.6% lower.

However, the recent history from mid-2010 is troubling. The Fed should clearly be concerned with the first derivative here, and the last CPI report should tell even those die-hard core-CPI-adherents (e.g. Dave Altig) that the inflation we are seeing may not be just some temporary relative-price phenomenon. In case you hadn't noticed, seasonally adjusted core cpi increased by 0.3% in May, and headline inflation increased by 0.2% in spite of a decrease in gasoline prices. Inflation is indeed a monetary phenomenon, and monetary policy is primarily about controlling inflation, though some of our pseudo-macroeconomist blogosphere types seem to think otherwise.

Now, suppose we look at how some monetary measures have been behaving, over the same sample as for the price indexes we looked at in the first chart. In the second chart, I show monthly seasonally adjusted M1, M2, and the currency component of M1, all from the beginning of the recent recession. I have also shown a 5% trend, the notion being that, if velocity were constant, if real GDP were growing at 3% per year, and if prices were growing at 2% per year, then money would grow at a 5% annual rate.

Clearly, M1 stands out here, in that it is currently 19.3% above the 5% trend, with currency 6.8% higher and M2 1.7% higher. Now, on the one hand we have to adjust for a financial crisis effect here, i.e. an increase in the demand for liquid assets due to the financial crisis, including an increase in the demand for US currency in the world at large. On the other hand, some of that demand has gone away, and ultimately we expect it to disappear. Further, it's possible we are too optimistic in expecting that we will soon, or ever, return to a 3% real GDP growth trend.

Therefore, if we are Old Monetarists, New Monetarists, or just sensible macroeconomists who understand that the Fed ultimately controls inflation by controlling growth in Fed liabilities used in exchange, we have to be worried if the Fed does not move to tighten. Barring some game-changing information, that tightening should at least happen by early fall.

In the current context, monetary policy should not be about futile attempts to manage the labor market. As I discussed here, it's best to view QE2 as irrelevant, in which case the fact that this asset purchase program concludes at the end of this month is also irrelevant. Even if we think that there are significant nonneutralities from monetary tightening, gradual increases in the interest rate on reserves in this context are certainly preferable to a large, quick tightening, which would entail potentially greater sacrifice.

18 comments:

  1. I’m not convinced that any QE, no matter the size, can’t lower long-term rates and thus increase GDP and employment. I still think my short term supply and demand diagram holds sway, at:

    http://newmonetarism.blogspot.com/2011/05/trends-doom-and-gloom.html?showComment=1306698821077#c3331592498584774163

    You say not in a general equilibrium. But your general equilibrium has frictions and will take time to reach, maybe years; in the meantime the short term supply and demand diagram basically holds.

    ReplyDelete
  2. "You say not in a general equilibrium. But your general equilibrium has frictions and will take time to reach, maybe years; in the meantime the short term supply and demand diagram basically holds."

    There is no evidence for this statement. It is just a blind faith in a discredited model because it gives you the answer you have an ideological predilection for.

    ReplyDelete
  3. The evidence that there are frictions is mountanous.

    ReplyDelete
  4. I don't find the claim that QE is "irrelevant", or the liked post, very convincing. It seems to me that, if we accept the Fed can influence short-term rates by purchasing short-term securities, the same logic should apply to long-term securities.

    Other than that, I tend to agree with the analysis. One could say that, since M2 seems to be under control, the Fed can afford to delay tightening somewhat. One might also argue that the economy needs an increase in inflation, because the market-clearing real interest rate is negative. However, it would appear that inflation is becoming a bigger threat than deflation, so perhaps the Fed should begin to tighten. It could always change course if inflation starts falling.

    ReplyDelete
  5. Most WS analyst estimate a 40-50 bps reduction in yields for the 5-10 year sector of the UST curve relative to what future path of short term interest rates imply.

    40-50 bps isn't exactly irrelevant.

    ReplyDelete
  6. Hi Richard,

    "You say not in a general equilibrium. But your general equilibrium has frictions and will take time to reach, maybe years; in the meantime the short term supply and demand diagram basically holds."

    That statement tells me that you don't know what "general equilibrium" means. You seem quite interested in economics, but you need to learn more.

    ReplyDelete
  7. "It seems to me that, if we accept the Fed can influence short-term rates by purchasing short-term securities, the same logic should apply to long-term securities."

    Exactly. Right now, with a positive supply of excess reserves, a swap of reserves for T-bills is irrelevant. The same is true if you swap reserves for T-bonds.

    ReplyDelete
  8. "The evidence that there are frictions is mountanous."

    But not the ones you seem to be proposing, since there is no way for an economy to be out of equilibrium.

    ReplyDelete
  9. "Right now, with a positive supply of excess reserves, a swap of reserves for T-bills is irrelevant. The same is true if you swap reserves for T-bonds. "

    But T-bills and reserves are equivalent, because they pay the same interest (approximately, at least) whereas long term bonds pay higher interest. So buying T-bonds should push long-term interest rates towards the near zero rate of short-term debt.

    Plus, I assume T-bonds are regarded as less liquid than T-bills, so exchanging the former for reserves should encourage banks to lend more. In other words, if banks are holding T-bills in a situation of excess reserves then they obviously value liquidity, and will thus prefer to hold on to the reserves rather than lend them out. But if they are holding less liquid assets in search of a higher return, then exchanging those assets for reserves might cause them to lend them out in search of a better return.

    ReplyDelete
  10. Read these two posts:

    http://newmonetarism.blogspot.com/2011/04/qe2-is-irrelevant.html

    http://newmonetarism.blogspot.com/2011/04/attempt-at-de-bafflement.html

    The basic argument is that central bank actions work if and only if the central bank is performing a financial intermediation function that the private sector cannot replicate. A central bank swap of reserves for long maturity Treasuries is irrelevant, as turning long-maturity Treasuries into an overnight asset is something that is easily accomplished by private sector intermediaries. That is, for example, the Bank of America can create a special purpose vehicle that holds T-bonds as assets, and finances those purchases with repurchase agreements, using the T-bonds as collateral. The special purpose vehicle then just rolls over the repos each day. That's exactly the same thing the Fed is doing. It's like the Modigliani-Miller theorem, if you know that. The Fed does something, the private sector undoes it.

    ReplyDelete
  11. "A central bank swap of reserves for long maturity Treasuries is irrelevant, as turning long-maturity Treasuries into an overnight asset is something that is easily accomplished by private sector intermediaries."

    "Can replicate" doesn't mean "is replicating."

    The private sector can change short-term interest rates as well, by changing its demand for money. Does that mean that ordinary monetary policy is irrelevant?

    While we are at it, there are lots of good reasons why the MM theorem doesn't hold either.

    ReplyDelete
  12. I have a BA in econ, an MBA, and I’m ABD in finance. For the ABD I had about half the Econ PhD courses and passed the econ pre-lims. And I’ve done a lot of independent reading in a lot of areas of econ. I’m strongest in econometrics and micro, and of course finance and personal finance. I would like to learn a lot more macro, but my wife has other ideas, so it’ll be slow for the foreseeable future.

    I’m interested in your blog because you’re the most credible publicly writing critic of Krugman I’ve found. People like Taylor and Mankiw are accomplished, but from what I’ve seen very willing to intentionally mislead to advance their extreme libertarian beliefs.

    You say that the private sector could do the same thing as the Fed does with quantitative easing if they want, but, (1) They don’t want to. They only consider private profit and not the externalities and suffering of unemployment. (2) They can’t create new high powered money, so I don’t know if it’s exactly the same.

    With regard to Miller-Modigliani, that I know very well from finance. They always teach you it’s far from holding in the real world because the assumptions are too strong. You always go over what happens when you add taxes, financial distress and bankruptcy costs, and other things.

    In any case, any large enough player can distort a market, and if there are market problems, then that distortion can increase total societal utility. The government is a player in the markets. If it buys enough long term bonds it will push the price up and the rates down, the same way as if any large player in any market increases their demand enough.

    ReplyDelete
  13. It is all wonderful and probably impressive to your peers to have a model that is internally consistent. But when it doesn't gibe with reality you have to question your model. How useful can your work be if it isn't remotely representative of the real world we observe?

    Any knowledgeable observer of the yield curve will tell you that the second round of QE influenced it.

    ReplyDelete
  14. I mean if the Fed wanted to push up the price of new cars, I guarantee you that if they printed up $200 billion in new money and purchased 10 million new cars, they would push up the price of cars, at least in the short-run. In the longer run auto companies would increase production and bring back down prices, but that would take years. It wouldn’t happen instantly like in some models.

    ReplyDelete
  15. "Any knowledgeable observer of the yield curve will tell you that the second round of QE influenced it."

    Well, they might tell me that, but that doesn't make them right.

    ReplyDelete
  16. "Well, they might tell me that, but that doesn't make them right."

    True. There are no guarantees in life.

    But I guess that I'd tend to believe market participants on the subject of factors that affect the markets. Call me crazy.

    Your comment reminds me of that old joke about the economist who asks, "Yes, I know it works in practice but does it work in theory?"

    ReplyDelete
  17. Stephen,

    You say that:

    "if real GDP were growing at 3% per year, and if prices were growing at 2% per year, then money would grow at a 5% annual rate. "

    The formula that I have used for the relationship between money, GDP and inflation would indicate that the money supply should have grown by 7.2%.

    The definition of money that I use is the bank credit and the formula is shown in the following link.

    http://www.buoyanteconomies.com/Inflation.htm

    The formula explains inflation in the US since 1995. There is a chart comparing the US CPI with the results generated from the formula.

    I have similar data and charts for New Zealand and Australia, although trade plays a larger part in these economies so that exchange rates also has an effect.

    There are links to the charts of Australian and New Zealand in the link above.

    I would be interested in your comments on the relationship.

    ReplyDelete
  18. I am having problems understanding your theory of prices, inflation, and the Fed.

    You say:

    "the Fed ultimately controls inflation by controlling growth in Fed liabilities used in exchange"

    You also say that QE2 is irrelevant.

    But Fed reserve balances are liabilities used in exchanges between private member banks. And QE2 created many more of these liabilities. How can any mechanism that creates more Fed liabilities be irrelevant?

    ReplyDelete