Wednesday, July 18, 2012

The ZLB is floor not a ceiling

Again and again I see the economy's problem described along these lines:

"At the ZLB (zero lower bound), the real interest rate is too high to get us to the optimum. The nominal interest rate cannot fall any further by definition. So to get to the optimum the expected rate of inflation must rise."

Those are Simon Wren-Lewis' words (they appear in a comment at the link), but Krugman and many others tell roughly the same story.

As always, I have questions.

In the IS/LM framework many (not Wren-Lewis) are using, doesn't this mean that we are getting "growth" by firms investing in projects with a negative NPV now made profitable by an even more negative discount rate?

Second, how is that inflation expectations rise and the nominal interest rate remains unchanged?

From Fisher, we think of the nominal rate as the required real rate of return plus a premium to offset expected inflation. So it's hard for me at least to think about expected inflation doubling (from 1.5 to 3 percent) or tripling (from 1.5 to 4.5 percent) without the nominal rate rising. For that to happen the required real return would have to fall one for one with the rise in expected inflation.

In other words, the ZLB is a floor, but not a ceiling.


John Thacker said...

I think that for many of these people, what they really mean is that the nominal interest rate (and nominal GDP, i.e., MV) must rise.

What makes even less sense than those people are people who keep insisting that we're going to (or are getting) get a huge amount of inflation. The market rate for long term debt should also be, following fisher, inflation plus the real growth rate. Yet the 30 year bond is yielding 2.6%, and CPI measures of inflation are well below the 2% target in recent years.

M has definitely increased, but the assumption that V is constant seems definitely untrue.

In any case, the argument for nominal money supply growth is that wages need to be cut, but that that can't happen because wages are sticky at the zero lower bound of growth. People hate 0% pay raises (and hate even more pay cuts) than they do 3% pay raises in a time of 4% inflation. They also will hate 0% pay raises if there is 1% deflation more than 4% pay raises in a time of 3% inflation.

People also hate the idea of paying a bank in nominal to safeguard their money, instead of getting below inflation but positive interest return.

Anonymous said...

For that to happen the required real return would have to fall one for one with the rise in expected inflation.


If the required RR is negative,
the nominal rate at zero (given say expected inflation =2) may be "too high". In that case higher inflation expectations can be consistent with i=0 and not require the required rr to fall.

a higher required real rate would be great news for the economy.

Unknown said...

I'm not exactly a fan of the IS/LM framework, but can attempt to offer insight from a different view for the second question. Nominal interest rates are controlled by the Federal Reserve. Rising inflation expectations may cause the Fed to raise rates or, as many hope, the Fed could maintain the ZLB regardless. In the latter case (assuming expectations are correct), the real rate would become even more negative.

I presume those supporting this policy expect the lower rr to lead to greater spending and hence GDP growth. I remain unconvinced.

Yuriy said...

1) It really does not have any connection with IS-LM; however, you seem to be right about NPV (if you think NPV to be a consistent model of a firm). Discount rate is obviously negative, so (an innovative negative-rate project of) using a pile of president-faced notes as a piece of decoration has positive NPV. The dynamics of financial corporation reserves is generally a worthy example of the idea.

2) You have strange ideas about Fisher's equation. One can't state that nominal rate must rise because of inflation increase: it's about EITHER the nominal rate rising OR real rate falling.

Really, it has some (very) vague connection with IS-LM: if the inflation is IS-side, nominal rate rises (investment increase causes lack of money, rate is bound to expand); if the inflation is LM-side, nominal rate should even decrease and the real rate is thus bound to plummet (more money means rates falling). However, the connection seems more like a bad intuition example than a real scientific thing (it's 3 o'clock here).

Then, Fed comes into the story. Even if inflation is caused by the budget deficit being up (so it's IS-side inflation) and the nominal rate may rise, Fed will not allow nominal rates to do so until the inflation hits the target. If the inflation is caused by monetary policy (Fed further expanding the monetary base), nominal rates can't rise; furthermore, they could even fall if not ZLB.

P.S. I personally fancy 3-step plan most:
1) Fed gets the inflation target up and then does nothing unusual: while the traditional monetary stimulation seems inefficient, the obligation NOT to restrict economy monetary seems the better choice.
2) Government temporary stops the austerity policy and spends lots on liquidity constraint people. Theoretically, fiscal policy is unnaturally efficient now in ZLB.
3) Everybody is happy: Fed and government are popular, rGDP is up, government income increases and so on. PROFIT!

Anonymous said...

Good comment. Let me contribute:

1) When we hit the lower bound, possibly, because the real interest rate is negative, monetary policy becomes innefective: we cannot further decresae interest rates and money and bonds become perfect subtitutes. Everybody here seems to get that.

2) Is there any policy that Central Bank can do? Actually yes, and Krugman 98 paper got this right (later formalized by Woodford) -- we can promise inflation. However, for this to work, the promise has to be credible (If not, announcing higher inflation will not affect expectations). To be credible, commitment is crucial -- and at the moment there is no technology for commitment.

3) But, assuming that the announcement is believed, my impression is that the channel that most economist have in mind is through higher consumption all the future periods, that initiates consumption know. I get what you have in mind for NPV, but long run rates dont seem to affect dicounting that much -- if not, we shoud see stock market booming.

Good entry!

Unknown said...

Stocks are a big predictor. They valued based on a ratio called the price earnings ratio.

Watch "Inflation Nation" by Ed Butowsky