Curious? Only to non-market monetarists
Market monetarists emphasize that interest rates are an unreliable guide to the stance of monetary policy. Monetary stimulus can lower rates via the liquidity effect, but it can also raise rates via the expected inflation and income effects. Here’s the Financial Times:
BoJ deflation war begets curious results
Japanese companies that borrow money from Mizuho Corporate Bank, one of the country’s largest lenders, received a small but unpleasant surprise this week when the bank decided to raise the rate of interest it charges on some of its loans.
The move, which was mimicked by two smaller lenders, Shinsei and Aozora, sat oddly because it came just a few days after Japan’s monetary authorities announced a dramatic effort to make money cheaper and more abundant.
In its most assertive attempt yet to stimulate the economy and end years of corrosive price declines, the Bank of Japan is doubling the amount of money in circulation by sharply increasing purchases of government bonds and other assets.
The central bank’s new policy, which it began to implement on Monday, should have pushed the cost of borrowing down, not up, from what are already historically low levels. Mizuho’s decision to increase its long-term prime lending rate, by 0.05 percentage point to 1.2 per cent, reflectedgyrations in Japanese bond markets that have followed the BoJ’s announcement.
Don’t trust the media. Don’t trust the old monetarists, the new and old Keynesians, the RBCers, the Austrians, the MMTers, or the new classicals. Only the market monetarists offer a model that allows you to make sense of what’s going on in the world.
PS. Not all non-market monetarists are wrong about interest rates, but they are all wrong about something. Maybe it’s the importance of demand shocks, maybe it’s the fiscal multiplier, or the effectiveness of monetary policy at the zero bound. Or whether QE will create high inflation.
Tags:
11. April 2013 at 06:34
A few questions arise here. Was this a forseen effect? Regardless of yes or no, how could this have been predicted? What were the incentives for banks like Mizuho to raise or lower their lending rates and how could we perhaps tell which force (liquidity effects vs future inflation and income effects) would be dominant?
Does this rate increase come as welcome news? i.e. beneficial to the Japanese economy? Or offset by the positive effects of expanded monetary base / deflated yen ?
Suspecting the financial times of misinterpreting this data was clear as I become more custom to the monetarist rhetoric and usual criticisms, but being able to paint a more complete picture of the mechanism would be more helpful here.
11. April 2013 at 06:50
Scott, wondering: how significant do you think it is that BOJ is buying equities and real estate. And that the Fed can’t do that. Does the BOJ have a(n easier) way out that the Fed doesn’t have?
Also (you’ve almost certainly addressed this…): If “liquidity trap” just means that it’s much harder for monetary policy to move NGDP at the ZLB — that the monetary authority has to be much more “adventurous” than it does when interest rates are higher — would you object to anything about the concept except maybe the absoluteness implied by the word “trap”?
11. April 2013 at 07:09
Dr. Sumner:
“Market monetarists emphasize that interest rates are an unreliable guide to the stance of monetary policy.”
That’s because market monetarism holds that current short term rates are a function of past monetary policy, and current long term rates are a function of future expected monetary policy, right?
“Don’t trust the old monetarists, the new and old Keynesians, the RBCers, the Austrians, the MMTers, or the new classicals. Only the market monetarists offer a model that allows you to make sense of what’s going on in the world.”
Austrian theory doesn’t make empirical predictions of where interest rates have to go given a change to money, so you can’t say their “models” should not be trusted. You’re imagining that Austrian theory predicts lower interest rates from looser money. That’s not the case.
It actually says this: IF interest rates are lowered from looser money (which you admit is possible), THEN such and such phenomena occurring in the capital structure of the economy is not stopped by what otherwise would have been the capital structure regulating force from free market interest rates.
Please don’t delude yourself into thinking that only MM theory explains historical data. There are always many mutually incompatible theories that are consistent with history *thus far*.
11. April 2013 at 07:21
Yes, this is true in theory, but it’s inconsistent with the recent, dramatic fall in JGB yields. Even the 30-year yield has fallen from 2% to 1.5% over the last few months (though, to be fair, it has risen sharply from a low of 1.2% recently).
The question isn’t “is it theoretically possible that easy monetary policy could lead to higher long-term yields” (in which case the answer is clearly yes, and I am somewhat surprised that it has not), but “does it make sense for banks raise their long-term rates at the same time market long-term rates are plummeting”. I suppose that with segmented enough markets this is possible, but it is still quite weird.
11. April 2013 at 07:31
Andrew, Interest rate changes are hard to predict.
I don’t see it as being particularly welcome or unwelcome (never reason from a price change. But if forced to take a stand, I’d say welcome, as more often than not higher nominal rates are associated with higher NGDP growth expectations. But again, never reason from a price change–it could be the liquidity effect.
Steve, No, I have lots of other objections. In my posts on “looking though the wrong end of the telescope” I argue that Keynesians have things backwards. Low rates are not a barrier to action, but rather a symptom of a failed policy, an excessively low (implicit) NGDP target. And for the same reason QE is not a particularly effective “solution”, what’s needed is NGDPLT.
In my view the Fed should target the forecast and make the base endogenous–not simply boost the base and hope for the best. In that case, a zero bound creates a very different policy dilemma from what the Keynesians usually assume. It’s not about the zero bound, it’s about the desired central bank balance sheet. Suppose NGDP were on target and yet he public wanted to hold a stock of base money that was larger than the public debt? That’s the real issue. In my view the best solution would be some combination of higher NGDP target path and negative IOR on bank reserves. Of course this is a problem the US is not likely to face, but perhaps Japan might, if they set a very low implicit NGDP target.
The BOJ has decided to buy a few non-traditional assets as well, but this is not because they no longer have any conventional assets to buy. Thus I think they are making a mistake. Still, if pressed I’d say it’s a lesser evil than fiscal stimulus.
The key is to first think clearly about what sort of monetary policy Japan would like to have. Even today the Japanese have not really done that. There’s no reason to suppose that a 2% inflation target will raise them above the zero bound. So they actually need to think about whether they are happy having a very large central bank balance sheet. If not, then do negative IOR and/or set a higher inflation target (preferably NGDP, not inflation.)
Geoff, You said:
“Austrian theory doesn’t make empirical predictions of where interest rates have to go given a change to money, so you can’t say their “models” should not be trusted. You’re imagining that Austrian theory predicts lower interest rates from looser money. That’s not the case.”
Nope. You forgot to read my PS.
You said;
“That’s because market monetarism holds that current short term rates are a function of past monetary policy, and current long term rates are a function of future expected monetary policy, right?”
Those factors you mention are not the only way that monetary policy affects rates, the liquidity effect also affects short term rates.
11. April 2013 at 07:35
reader, Fair enough, but then the FT should have said so. I was responding to the faulty reasoning process in the article–which I presume even you admit is flawed. I’m no expert on all the factors that go into the interest rate charged in risky loans. For instance, is corporate credit demand soaring? I don’t know. And as you say, is there any segmentation involved?
11. April 2013 at 07:35
The curves I see say that rates have increased to for intermediate term borrowing (1-10 years) and fallen from long term borrowing (more than 10 years).
This would suggest that expectations greater optimisim for an economic improvement in the near term, but inflation expectations and expectations for longer term economic performance is largely unchanged.
11. April 2013 at 07:43
Dr. Sumner:
“Those factors you mention are not the only way that monetary policy affects rates, the liquidity effect also affects short term rates.”
Dang, your responses are getting more and more sophisticated, and it is getting harder and harder for me to say anything against them.
OK, would you accept the possibility that in a world of price inflation targeting, where investors expect low price inflation, that the liquidity effect can dominate the directions of interest rates for significant periods of time, say 5-10 years?
11. April 2013 at 07:46
In other words, is it possible for investors to be partially doing the Fed’s price inflation targeting for the Fed, by increasing their cash preference given OMOs in such a way that the investors are partially bringing about a liquidity effect on themselves?
11. April 2013 at 08:26
“The BOJ has decided to buy a few non-traditional assets as well, but this is not because they no longer have any conventional assets to buy. Thus I think they are making a mistake. Still, if pressed I’d say it’s a lesser evil than fiscal stimulus.”
I did do a double-take when I saw they would be buying equities. The BOJ is the first to actually do that, but many at the Fed have suggested buying equities. They assume that OMOs only work through pushing up prices directly, and since yields are “already low,” the thinking is that the central bank should go straight for lowering the risk premiums.
A question I have is how much of Japan’s stock market movements have been pure expectations (i.e. the explicit 2% target), how much has been conventional QE and how much has been unconventional QE?
One could imagine a Fed where the only asset on their balance sheet is very short-term Treasuries and they buy Treasuries at the same rate they want to increase NGDP (i.e. assume constant velocity). Then they enforce constant velocity through IOR, positive or negative, with some way of getting around cash-hoarding with negative IOR. That would seem to be the most neutral way to me of doing monetary policy, excepting some sort of NGDP futures regime.
11. April 2013 at 09:41
“Monetary stimulus can lower rates via the liquidity effect, but it can also raise rates via the expected inflation and income effects.”
A cop out statement that renders your views “right” no matter what happens.
Certainly you can not argue that these falling rates are a sign of successful policy under your own terms. Without raised expectations of NGDP, none of the successes in terms of “expectations based policy” you’ll want to lay claim to will be of any meaning.
Which part of the story did MMTers get wrong exactly? From how I understand their views they would argue that aggressive monetary stimulus will lead to debt monetization, thus driving down interest rates and guaranteeing the solvency of the government. This counteracts the disinflationary effects of Japan’s debt overhang (see: http://michael-hudson.com/2012/09/incorporating-the-rentier-sectors-into-a-financial-model-3/), thus not leading to unmanageable inflation. In fact it’s hard to come to any conclusion other than MMT’s based on the data seen so far.
11. April 2013 at 12:06
Rademaker:
“A cop out statement that renders your views “right” no matter what happens.”
A snarky way of arguing your own underhanded presumption that is “right” no matter what happens, namely, that the only valid knowledge of the world are through considering falsifiable arguments and only falsifiable arguments.
If I said that two things added to two things makes four things, then that will be true “no matter what happens.” But does that mean that my statement is not saying anything true about the world? No.
Understanding and presenting the factors that can affect interest rates, the extents and interactions of which are not predictable in advance, isn’t a “cop out”. It only seems like a cop out if you believe the duty of the economist is to make predictions of people’s future knowledge.
11. April 2013 at 13:37
“Monetary stimulus can lower rates via the liquidity effect, but it can also raise rates via the expected inflation and income effects.”
Good observation. But is there a way to predict which consequence will follow from any particular monetary stimulus?
One variable that might indicate the consequences would be who receives the monetary stimulus. If the new money ends up in the hands of investors, i.e. banks, sovereigns, money market funds, etc., then the increased demand for liquid, income generating assets will drive interest rates down. If the new money ends up in the hands of consumers (either directly or by increased lending) than the expectation of price inflation will drive yields up.
But short of sending everyone a check for $50,000, how does the central bank allow stimulus money to reach the consumer economy if the banks won’t lend it out?
Buy securitized consumer debt? (In addition to government bonds and MBS)
11. April 2013 at 13:41
I lost some esteem for Lars Svensson. He´s still worried about the value of parameters in the Taylor-Rule interest rate setting!
http://thefaintofheart.wordpress.com/2013/04/11/denmark-and-sweden-in-deflationary-hell/
11. April 2013 at 13:58
Scott, you are unfortunately not as good at snark as Krugman. What people seem to miss is that you are saying that interest rate changes can’t tell you anything and should not be used to reason what path the economy is on right now or in the future. This shouldn’t be surprising because any change shouldn’t be surprising and we can never really understand the fundamentals of why interest rates move until some point in the future.
It was actually reading the MMT/post-Keynesian types that elucidated this for me when they showed that the gov’t long term budget constraint (no ponzi) is inherently unknowable.
11. April 2013 at 14:12
The confusion here seems to be over real interest rates and nominal interest rates. Nominal interest rates went up by 5 BPs but with an extra 1% worth of inflation coming down the pike, real interest rates dropped by 95 BPs. That result shouldn’t be curious to anybody, market monetarist or otherwise.
11. April 2013 at 16:22
Hi Scott,
Just wanted to let you know I posted some historical break even spreads for Japan bonds relative to there inflation adjusted counter parts. You might find it interesting/useful and I haven’t seen the data many other places.
http://badoutcomes.blogspot.com/2013/04/inflation-expectation-in-japan.html
11. April 2013 at 16:53
Scott,
First some facts on Japan interest rates…..
Changes to nominal short term rates (TIBOR) and long term rates (30 year JGB) have been minimal.
Mizuho prices their prime rate off the 5 year JGB.
2 to 5 year JGB yields have gone up, but this is largely due to liquidity issues in the market and the fact that the BOJ is lengthening the maturity of it’s portfolio. I don’t think you can really read too much into the change in mid-term yields as they relate to inflation expectations or NGDP growth expectations because of these market anomalies
It seems to me the real question is whether BOJ action and announcements are effective, i.e. whether there is evidence to suggest that it will actually result in increased NGDP growth.
If we assume the triggering/transmission mechanism for higher NGDP growth is higher real prices of financial assets and higher expectations of NGDP growth then the evidence and answer are very clear.
If we ignore nominal pricing and think of real pricing as 1/(1 – expected risk adjusted real return), then what do we know.
1. Let’s make the assumption that real prices for equities have not dropped, i.e. real prices are either flat or have gone up. That tells us that the increase in nominal pricing means the market expects higher profits which I think we can only conclude means that the market expects higher NGDP growth.
2. Unfortunately, there is not a liquid TIPS market in Japan so it’s not possible to directly gauge inflation expectations, which is problematic for determining real rates/prices for the debt market. However we do know from 1. above that the market expects higher NGDP growth. I think it’s safe to assume that the market does not expect all of this growth to be real and that part of the growth will be in higher prices. Therefore we do know that the market expects higher inflation and therefore since nominal yields (at least in the short and long end of the curve) are basically flat, we also know that real debt yields have fallen (the real price of debt securities has risen.)
So in summary, we know based on reasonable assumptions that a) the real price of financial assets has risen, and b) expectations of NGDP growth have risen.
IMHO the presumption that higher real prices of financial asset and higher expectations of NGDP will growth will trigger and translate into higher actual NGDP growth is obvious and fundamental to the basic tenets of economics.
But again, Scott, you’re using a different model (HPE), which I think makes it much harder for you to definitively show that market movements prove that the BOJ strategy is working.
11. April 2013 at 17:00
Charlie,
The only problem with the inflation adjusted bonds is that the government stopped issuing them a long time ago and has bought back most of the outstanding paper so the market is very, very illiquid. That said that data is very consistent with what one would expect.
Can you get any data out of Bloomberg on trade volume for the inflation adjusted bonds?
11. April 2013 at 18:42
Matt, It’s very clear that most of the stock market rise has reflected the QE itself, plus the 2% inflation target. I doubt anyone would claim the stock purchases had major effect. I seem to recall Hong Kong tried that once.
Rademaker, You said;
“A cop out statement that renders your views “right” no matter what happens.”
That’s better than being wrong, which describes 95% of pundits. More seriously, you obviously are new to my blog, I’ve been screaming that interest rates are not a reliable indicators of monetary policy at least once a month for 4 1/2 years. It’s NGDP expectations that matter.
Bernard, It doesn’t matter who gets it, it depends on the change in the expected future path of policy.
errorr, Yup, He’s better at snark.
Randomize, Yup, no one should have been surprised. So why were they?
dtoh, Thanks for that info–it all sounds reasonable to me.
Out of time . . . more tomorrow
11. April 2013 at 20:21
dtoh,
Good to know. I will look for volume numbers tomorrow. That may explain why the two year break even rate looks so strange and out of line with the others.
12. April 2013 at 06:00
The latest from Plosser: http://www.bloomberg.com/video/won-t-cut-qe-if-jobs-market-weakens-plosser-Dn3J5fPRRU6QdwmYfFvSSg.html
And Eichengreen on the econ textbooks of 2033: http://www.project-syndicate.org/commentary/how-economics-will-change-in-the-next-20-years-by-barry-eichengreen
12. April 2013 at 07:37
[…] h/t Saturos […]
12. April 2013 at 09:30
Dr. Sumner, et. al.,
Does anyone know where to find rates on Japanese Inflation Protected Securities? It would be much, much more surprising if rates on those have risen as well.
12. April 2013 at 11:05
Saturos, Wiki-text?
Randomize, Check out dtoh’s comment.
12. April 2013 at 11:20
Were the inflation-adjusted bonds linked to PCE, CPI, or something else? It seems like a great way to arbitrage the differences between how economic players experience inflation.
12. April 2013 at 11:21
…er, the Japanese equivalents, I mean.
13. April 2013 at 03:25
I find it quite interesting that you seem to believe that the Fed has the potential for extraordinary control over NGDP yet it is mostly helpless when it comes to controlling interest rates, that all it can do is “manage” expectations. How is that?
Interest rates ARE a policy prescription of the fed, NGDP is the result of numerous actions taking place in a market of over 300 million nationally and billions worldwide.
13. April 2013 at 04:41
Gasman, The problem is that monetary policy affects NGDP in only one direction (stimulus tends to raise NGDP.) In contrast, it affects interest rates in two directions; stimulus tends to both raise and lower rates, so it’s unclear how it will net out.
15. April 2013 at 13:41
I see Scott didn’t mention Monetary Realists on his list of people not to trust. Whew!
15. April 2013 at 13:43
Shoot! I guess you caught them in your “PS” … “all non-Market Monetarists.” Ha! 😉
15. April 2013 at 19:58
Also notice that Scott didn’t explicitly say not to trust the Marxists! … Hmmmmmmmmmmm 😉
22. April 2013 at 10:53
Tom, Who are the Monetary Realists?