David Beckworth interviews Joe Gagnon
Joe Gagnon is a former Fed economist now located at the Peterson Institute in Washington. He’s also arguably the world’s leading expert on QE. The conversation with David was excellent from beginning to end.
Here are a few highlights (from memory):
1. Gagnon’s research suggested that QE was somewhat effective at lowering long-term interest rates. That’s also the consensus view of dozens of other studies he looked at. He suggested that QE probably led to higher bond yields in the long run, which reflected the higher nominal growth of countries that engaged in more aggressive monetary stimulus. He seemed to suggest this was a rather surprising result, but I think it’s in line with the previous views of monetarists like Milton Friedman.
2. Gagnon likes NGDP targeting, and is somewhat split between growth rate and level targeting. At one point he seemed to suggest an option somewhere between those two extremes. I’d guess that reflects the behavior of the economy after 2007, when (in retrospect) a continued 5% NGDP growth rate might seem a bit too aggressive.
3. He suggested that the Fed may have been held back around 2009-10 from doing even more QE by a fear of the unknown. It was a new and untried policy instrument. Gagnon also indicated that (in retrospect) it probably would have been better to do all three or four QEs right up front. (I’m glad he said 3 or 4, as I’ve always been a bit unclear as to whether the first QE was in late 2008, or March 2009. It seems the leading expert also views the number of QEs as ambiguous. The official number was three, but it seems like there were four.)
4. David asked him about the options for monetary policy that he came up with as a researcher at the Fed during 2008-09. I kind of regret not hearing him talk about whether the Fed looked at the options for Japan that were outlined in Bernanke’s 2003 paper. There’s been a lot of criticism (from me and others) of the fact that Bernanke’s Fed did not pursue some of the more aggressive options that Bernanke recommended to the BOJ, in his famous paper that discussed the need for “Rooseveltian resolve.” (Here I’m especially thinking of price level targeting.) That’s not to say there are not good answers. Bernanke got in hot water in 2010 for suggesting we needed to raise the inflation rate (to 2%). If he had indicated a need to raise it to 3% or 4% to catch up to the trend line, the policy would have been even more controversial.
5. Gagnon gave an excellent summary of recent events in Japan. His view is similar to mine, but he’s followed things more closely and has much more knowledge of the situation. The original Abe/Kuroda push for 2% inflation was partially successful. Core inflation expectation quickly rose by about 200 basis points, from minus 0.75% to 1.25%. They needed one final push, and in early 2016 tried to do so using negative IOR. Unfortunately, the negative rate was only 0.1%, which was too little to have much effect. Even worse, there was a political backlash. That led the markets to lose confidence in the BOJ, and since then the yen has soared in value. Thus inflation expectations are now coming down. In retrospect, they would have been better off doing more QE. Gagnon even suggested buying equities as an option. He thought it was really important that the BOJ hit its 2% inflation target, and I agree. That’s not because I favor inflation targeting (I don’t) but rather because I think it’s really important for central banks to hit their targets, for credibility reasons. Unfortunately, it appears the Abe government has lost interest in this policy target.
6. Gagnon pointed out that before the Great Recession most economists thought that a 2% inflation target would be enough to keep us away from the zero bound. He also noted that early discussions of the pros and cons of a higher inflation target took account of the likelihood of hitting the zero bound. Then Gagnon said something to the effect; “Well, we now know something new”. We know that the zero bound problem can occur with a 2% inflation target. So if economists thought that 2% inflation target was optimal for the US, then that can’t possibly be the case now. Speaking for myself, I’m disappointed that so few economists are forcefully explaining why this new information suggests that we need a new target. And it does not have to be 3% or 4% inflation, it could be NGDPLT. But clearly some change is required. And yet as far as I can see the Fed seems determined to continue muddling along with a 2% inflation target, which makes their “conventional” policy tool (interest rate targeting) almost useless going forward.
7. Gagnon also did an excellent job explaining the problems with helicopter drops–it’s too effective if expected to be permanent.
Tags:
5. July 2016 at 07:25
Scott,
I agree and disagree. There were four QEs or there were zero. What do I mean:
One element of the contractionary jawboning in late 2008 was the premise that the fed had exhausted its supply of short term treasuries with which to sterilize its liquidity auctions. Interest on reserves followed as the alternative tool to sterilize. so you can see the first QE here before they reached the lower bound, but you can also see the contemporaneous explanation of why QE would not be a Milton Friedman got money injection but really somethings more akin to an operation twist–reducing the term structure of the consolidated fed-federal debt.
Regarding the lack of rational follow through–that had much to do with the left wing focus on liar’s poker and subprime this or that. The fed tried to diagnosis the problem as a liquidity crisis, but the hollering of the left was that it solvency crisis–and this became way to signal you were intelligent. The fed though was also too constrained — focused on the interbank market and not the soaring value of the US dollar.
5. July 2016 at 07:35
I haven’t listened to the podcast yet, but it is probably important to remember that after the epic Sept. 2008 decision to keep rates at 2%,ior was instituted and held at 1% until mid december. It’s really strange to talk about the problem of the lower bound when after the Fed made a 100 year flood level blunder they immediately moved the floor up. That doesn’t seem like the action of an institution concerned about the lower bound.
5. July 2016 at 07:48
Jon, I thought the first QE was an attempt to add liquidity without easing monetary policy.
Kevin, I think Gagnon would agree that money was too tight in 2008.
5. July 2016 at 08:29
Basing monetary policy on inflation targets may have made sense in the 1970s but not today. With wages flat or declining, and income increasingly concentrated, how can there be a “general increase in prices”? The related problem is confusing bubbles with inflation – even Larry Summers was guilty of it in 2008. Indeed, with incomes increasingly concentrated and interest rates persistently low, we should expect recurring bubbles (as investors chase higher yields by bidding up asset prices). Attacking bubbles in a slack economy with monetary brakes designed to combat inflation is essentially Sumner’s criticism of the Fed (although his criticism is put in much more technical terms). Sumner’s NGDP targeting is gaining converts, including Robert Shiller (who especially likes the idea when coupled with an NGDP prediction market). If the Fed ends up adopting Sumner’s idea it would be the greatest achievement in economics since, well, Keynes. Whether it will work is an entirely different question. If insanity is doing the same thing again and again and expecting a different result, at least Sumner’s idea won’t be insane.
5. July 2016 at 08:34
Rayward, Do you have a link where Shiller talks about NGDP targeting and or using prediction markets? Thanks.
5. July 2016 at 08:58
What a surprise, Sumner thinks somebody who (supposedly) agrees with what Sumner thought of first (supposedly) is correct. Yet Sumner still believes in the expectations fairy: just in case he’s wrong, he can blame it on faulty expectations.
5. July 2016 at 09:25
These Beckworth podcasts are amazing!!! Funny how Gagnon was able to get copies of the different QE options (including his own work) only by submitting FOIA requests.
Ray Lopez – did you listen to the podcast?
5. July 2016 at 10:13
I didn’t save the link because Shiller’s comment was so offhand that I assumed his favorable view of NGDP targeting was common knowledge among economists – I’m not one so I wouldn’t know. Of course, the prediction market in NGDP complements Shiller’s idea of the government selling “trills” to fund its borrowing requirements. Anybody who has ever watched Shiller’s lectures in finance knows he lights up when talking about futures markets – how futures markets (beginning with commodities (e.g., rice) made asset supplies and hence asset prices much more stable. His lectures are available at itunes university for those interested.
5. July 2016 at 10:24
Bernanke the ‘good guy’ & Bernanke the ‘bad guy’
https://thefaintofheart.wordpress.com/2015/05/03/bernankes-amnesia-caused-the-depression/
5. July 2016 at 10:54
It’s about the reason why policy was tight. There are two potential reasons for policy to cause a ZLB problem: (1) the inflation target is too low to avoid interest rate fluctuations that reach zero, or (2) variance between the stated Fed target and the effective Fed policy causes a potentially sharp negative shock. It seems to me that market based NGDP level targeting mostly solves #2, and that #2 was mostly what caused us to hit ZLB in late 2008. If the Fed avoids facing up to that, then it makes it more likely to look to half-step measures that solve #1. A 3-4% inflation target would probably have helped in late 2008, but it probably isn’t as dependable a solution as NGDP targeting. And, in fact, NGDP level targeting at a higher rate could actually address both #1 & #2.
5. July 2016 at 12:30
Leaving the LT aspect aside for a minute, NGDP growth rate as a target instead of inflation has the benefit of being a lot more politically palatable. So much easier to say that “the economy” should grow by 4% or 5% than that prices aren’t rising fast enough. And politically palatable ends up meaning more credible.
5. July 2016 at 12:32
My thoughts on the podcast – shamed by SG I listened to it, skipping just a few minutes – a massive waste of time!!!
Opening slow, entire podcast slow. Joe sounds young. Interested as a youth in Time magazine (LOL). Sounds like a history major. “Thinking about science” but became a economics major. Likes macro since likes “big picture” (like Time?). Boston based so bigoted Bumner likes him no doubt. 3:00 mark and still a biography of Joe. Taught for a year in Berkeley (resume builder). Sounds gay. Boring “day in the life”, prefers research not management. 4:50 boring. Speaks generalities, nothing juicy. Boring Meta picture. 5:20 QE mentioned. Joe speaks interest rates hitting zero. Did not think of negative interest rates, so logically wanted QE. 6:40 still stalling. Still “meta”. Buy longer term and riskier assets and that’s QE1. TAL meant to be 1T$ junk buying program by Fed, but this “buy junk” never took off because Fed charged premium over what market charged. 9:00 EMT says no effect of QE, yet (implies) it does. Regression analysis used to prove QE. Small results shown, non-zero (which if anything shows EMT is right), 25 basis points for $1T assets bought. Japan’s (10:22) QE disagreement about effect. Again, money is neutral–RL. Japan’s QE was too short term maturities, was an asset swap, unlike US’s QE. Now Fed claims QE lowers long term bond yields (though not explained why this claim made). 12:19 ‘portfolio’ and ‘signaling’ channels mentioned (metaphysics Sumner would enjoy). US/UK results not fully believable (12:55), consistent with metaphysics. 13:06 two channels defined. Metaphysics. Signaling: supposes Central Banks care about profits (they don’t), CBs can’t forecast more than 2 years anyway, so this signaling is bunk 15:08. Joe speaks too slow. I’m getting bored. Portfolio channel: 15:31 still not explained. Signaling channel cannot explain things Portfolio channel can (16:00) David Beckworth does a credible job summarizing Joe (16:47). David mentioned early monetarists as precursors to QE. Paper 17:44 by Joe: days Fed announced (19:00 mark) compared to other days, and changes in bond yields in these two windows compared, and QE delta estimate was 100 basis points (1%, basically noise). Then another analysis found 50 basis points (0.5%). 20:36 Joe says ‘yes-and-no’ in QE’s diminishing effects. 0.5% is ‘non-diminishing’ (I would argue this is just noise). Regression analysis ‘proves this’ (data mining), 21:32. boring boring boring 24:00 long term yields go negative anyway (Beckworth) so what’s the deal with QE? Good point. Downward 30 year trend in bond yields, 25:00. Boring history everybody knows. C’mon man, move! 26:00 Lower interest rates happen regardless: demographics, less capital needed, no dissavings in most people’s retirement, productivity weak thus less investment, developing countries now lend to advanced economies (Rodrick), 27:00, so yields low. Beckworth talks about a ‘massive wound’ in Great Recession–28:19 –but fails to make his case for low yields. Neo-Fisher mentioned by Joe (23:30), more metaphysics. Higher rates, bigger rebound if aggressive early in recession easing money. 30:00 US > EU. US suffered less due to better monetary policy (!–you can also say US’s labor market more dynamic, US has better R&D, US more integrated than EU, etc etc etc) Skip ahead 33:35 “QE tried but muzzled by inflation fears” – Fed payed little attention to this, but afraid of ‘new policy’, ‘big numbers scary’ (shows Fed is conservative, but they don’t realize that money is neutral so it doesn’t matter IMO). Fed sounds like a bunch of rubes “geez my balance sheet is so big” 35:35). Krugman et al 35:48 – QE is irrelevant, temporary, since not credible. Beckworth speaks with molasses in his mouth. 36:41 Krugman right in theory, and apparently in practice. 37:00 Joe against helicopter drop. Amount of QE that you do in recession would lead to hyperinflation (my words not his, which is my point on this forum) “Money cyclical in ways we don’t understand” “inflation unacceptably high” (38:28) “price level target” NGDPLT? 38:43. Joe: with level target cannot do helicopter drop (good point) if promise never to mop up money. Level target has small benefits in recession, not as powerful as helicopter drop. Joe likes: not pure level, not rate of change, somewhere between: 40:10 Joe: “I DON’T KNOW” – yet Sumner seizes on this as an endorsement? I’m out of here…. Japan mentioned 40:40 as ‘frustrating’. Japan as exemplar. Boring slow Joe. Japan does US style QE 41:31, bigger than US. Jp inflation “soars” from -0.75 to +1.25 in two years “huge success” (keep in mind money neutral–RL). Diminishing returns in JP 42:45. Tiny moves into negative rates in JP. Lots of complaining about negative rates. Bad PR 43:40. “Markets think” – LOL, Joe knows what markets think, like Sumner I guess. “Safe box sales” 44:10 up in JP. Equity should be bought by JP CB 44:32. Credibility fairy. Fed cannot buy stocks, JP CB can. Skip ahead. 47:00 negative rates cannot go below -1%, -2% Behavior fairy 47:35. Metaphysics. These clowns have hubris. Miles Kimball 48:00 proposal. Money illusion 48:35 at work. Joe: we need 3-4% inflation. 49:10 “contagious” “chain reaction” on all long term rates. Domino theory about as sound as Vietnam. 49:50 outlook: global economy is OK says Peterson Inst. China is OK (50:14), won’t drag world down. “China determined” not to let economy go down (foolish Joe believes in a planned economy), Fiscal Capacity 51:11 massive in China. Brazil: 51:46 some good reforms coming. India, Indonesia doing well, 52:00. Jp need central banks to step up, EU CB committed, Joe don’t see collapse 52:40. Hard to have deflationary spirals 52:55. Strong dollar (53:00) – bad? (Beckworth) Joe: strong dollar a problem but helps Fed. Doesn’t explain why, just concludes. Entire podcast a waste of time. 54:51 BrExit a big deal says Joe, weaken dollar, investment drops in UK. No good from BrExit. Goodbye, what a waste of time!
5. July 2016 at 13:59
Given Friedman’s assertion about low interest rates signalling tight monetary policy…would it be accurate to state the following?
Monetary policy is actually relatively tight when the liquidity effect dominates the Fisher effect.
Or rather:
Monetary policy is actually relatively loose when the Fisher effect dominates because higher inflation expectations produce higher nominal interest rates.
This runs counter to the standard logic: A liquidity effect suggests an expansionary monetary policy.
What’s a market monetarism take on this line of reasoning?
5. July 2016 at 16:06
Nice post.
Yes, if a central bank targets inflation, why has 2% become sacred (in the West). China, India target 4% +- 2%.
An IT band around a higher target makes sense.
The Bank of Japan does buy equities. ETFs.
When Takahashi Korekiyo, Japan’s central banker, used money-financed fiscal programs in the Great Depression, it did not result in too much inflation, though it did end the depression in Japan.
Yes, Ray, money is neutral but it can end a depression. Oh, that. In theory, ending the depression by money-financed fiscal programs is not important.
5. July 2016 at 19:08
@ScottSumner you said: “He suggested that QE probably led to higher bond yields in the long run, which reflected the higher nominal growth of countries that engaged in more aggressive monetary stimulus. He seemed to suggest this was a rather surprising result, but I think it’s in line with the previous views of monetarists like Milton Friedman.”
That could be true, or maybe not if you run into a shortage of the bonds. Wouldn’t supply and demand determine whether it is true or not true, Scott? Bond shortages come from massive demand in the new clearinghouses and those shortages push yields down.
Also, helicopter money disbursement cannot be permanent, according to Eric Lonergan. Only the increase in the money supply can be permanent, but the window of disbursement must be finite.
5. July 2016 at 20:18
So, I was so excited when I found out what monetarism could do. Helicopter money was the culmination of Friedman’s genius, and yet non of the monetarists, especially at the Fed, even would dare consider it.
Even Scott Sumner generally defines it incorrectly and then blasts the wrongly defined HM.
Why don’t monetarists have more courage? Seriously. We have arrived at negative real rates. Isn’t it time to implement something better than QE?
I almost think monetarists would rather turn their science into an academic exercise, hoping they will be chosen to work for the Fed rather than clamor for real change.
I apologize beforehand if I am misinterpreting things here, just trying to make sense of it all.
5. July 2016 at 21:16
Scott:
You write, “I thought the first QE was an attempt to add liquidity without easing monetary policy.”
Quite so. The first QE was an attempt to add liquidity without easing policy. The tool to do that sterilization was IOR. If IOR sterilized QE in round one, isn’t there a reason to argue by extension that IOR sterilized QE in each subsequent round. Ergo… if the later QE rounds mattered, it was because of a second order effect arising from a perceived risk that QE would be difficult to exit rather than directly 1:1 from the injection of high-powered money.
6. July 2016 at 05:32
I hate to say it, but Gary Anderson is right about the verve and intrepidness of monetary economists.
A dance choreographed by monetarists would be a military march…in a circle.
The Fed still jibber-jabbers about inflation at FOMC meetings….
6. July 2016 at 07:20
Scott,
Longtime reader, first time questioner. Apologies for the modestly off-topic question, but this seemed the best available forum to ask it:
Something you’ve mentioned in passing in past posts (both here and in Econlog) is the difference between “monetary policy” and “credit policy”. You’ve said the two have nothing to do with each other. This stands in some contrast to most financial market participants (as well as a number of monetary policymakers), who seem to believe the two are intertwined by definition. One reason a number of my colleagues both believe money is easy right now and that higher rates would be optimal is by looking at credit creation. “Debt issuance is through the roof in an unsustainble way, thus money is easy and should be tightened through rate hikes”, they would say.
In a comment you left on an Econlog post from summer 2015 (http://econlog.econlib.org/archives/2015/06/the_fed_is_not.html), you said:
“No, monetary policy has nothing to do with credit in my view. It’s about keeping NGDP growing at a stable rate to avoid unnecessary labor market and financial market shocks.”
This seems as good a brief definition as we could offer for monetary policy. Could you offer a similar definition for “credit policy”? This would help me fight the good fight. Anything you could offer would be much appreciated.
Thanks and keep up the excellent work.
6. July 2016 at 07:27
“The singularity” will come…
Political situations are too difficult to implement NGDPLT ,at least within Japan.
Maybe AI will solve the all problems at 21XX.
6. July 2016 at 17:33
Ray, I’m guessing no one will read that long comment, hope you had fun writing it.
Alex, It’s not a question of which effect dominates, it depends on whether NGDP growth is fast or slow.
The Fisher effect dominated in the early 1930s, yet money was tight.
Chris, A good credit market is one that tries to remove artificial barriers and subsidies. Excessive regulation may be a barrier to credit, and tax breaks for mortgage interest and moral hazard from FDIC is a subsidy to credit. Get rid of those, and you’ve improved credit policy.
7. July 2016 at 04:32
Prof. Sumner
If QE works, why bother with higher inflation targets and “get away” of the zero lower bound? It seems to me there is a contradiction here: interest rates don’t matter and QE works, but we still need to get away of the zero lower bound, because, so people say, we have “room to lower rates when needed”. It seems to me that people are searching for an excuse to just choose higher inflation targets. To me it is very simple: getting away from the zero lower bound is important a) if interest rates matter after all, and b) QE does not work ….
7. July 2016 at 13:12
Jose, The main reason to get away form the zero bound is that central banks are rather clumsy with QE. In addition, it allows for a smaller central bank balance sheet, which many people prefer.
8. July 2016 at 06:00
Bankrupt U Bernanke should be in the Federal Prison. He caused the world-wide GR all by himself. All economists are vacuous – as the GR aptly proves. The GR actually started March 31st 1980 (when pass-thru account regulations removed any reserve requirement restrictions). The DIDMCA turned 38,000 NBs into 38,000 CBs. After the the dust settled during the predicted S&L crisis, the Fed lowered legal reserve requiremments by 40 percent. Thus the money multiplier exploded after c. 1995 – and the stage was set.
In Sept. 1996 Ed Fry discontinued the G.6 release, i.e., stopped publishing the transactions velocity of money (whereas Vi is a contrived figure). DD turnover reflected both new and existing property sales. As such, the turnover figures would have stuck out like any rural water tower.
Notwithstanding Greenspan’s errors (he NEVER tightened), when Bernanke took over, the rate-of-change in money flows (proxy for inflation), was persistently squeezed (@ negative roc’s), for 29 contiguous months (ending July 2008).
As if lowering Yale Professor Irving Fisher’s price-level (and housing’s prorata share), for 29 consecutive months wasn’t enough, Bankrupt U Bernanke squeezed the roc in money flows (proxy for real-output), until it dropped below zero (again, negative roc’s). This recessionary trajectory was self-evident as early as Dec. 2007.
But we are not done. Bankrupt U Bernanke introduced the payment of interest on excess reserves thereby destroying all non-bank lending/investing (as the economy was already collapsing). Bernanke should be exectued by a firing squad for economic treason.
I.e., nothing’s changed in over 100 years. And we knew this aready (via, the 1931 committee on member bank reserve requirements, which the Fed finally declassified in 1983).
– Michel de Notredame
8. July 2016 at 06:51
QE, when reserves are remunerated, is a joke. I.e., the Keynesian economists have finally achieved their objective: that there is no difference between money and liquid assets (as if Alton Gilbert’s “Requiem for Regulation Q” was not enough and that Milton Friedman never knew the difference between “stock and flow” see: Dec. 1959, -Leadenham Archives).
Whereas the probable distribution of the FRB-NY’s purchases between the banks and their customers is unpredictable didn’t matter before Oct. 2008 (as the CBs minimized their non-earning assets, thereby always countercyclically expanding the money stock), afterwards money expansion became problematic (and even more so considering Basel requirements to increase bank capital – which negates FOMC policy by destroying the money stock – dollar-for-dollar).
I.e., the paymment of interest on excess reserve balances inverted the wholesale money market (which funds the capital market), which created the 2008 GR’s credit crunch (when R-gDp was already collapsing). The 1966 S&L credit crunch is the paradigm.
By using the wrong criteria, interest rate manipulation, rather than IBDD volumes, the Fed emasculated its “open market power”.
The money stock (& commercial bank credit, e.g., CBs, MSBs, CUs, and S&Ls loans and investments), can never be managed by any attempt to control the cost of credit (something Paul Vocker never tried, see: Paul Meek’s 3rd edition of “Open Market Operations”, published in 1974).
To wit: All CB time/savings deposits, rather than being a source of loan-funds, are the indirect consequence of prior bank credit creation. And TDs are derived from DDs, directly or indirectly via the currency route, or thru the CB’s undivided profits accounts…and the growth of DDs can largely be accounted for by the expansion of bank credit (as anyone who has applied double-entry bookkeeping on a national scale should already know – but that anaylsis was eventually made impossible by the DIDMCA).
8. July 2016 at 06:59
Thanks prof. Sumner
Yes, I a lot of people really freak out with large balance sheets, despite the fact that inflation is nowhere to be seen, and there is some evidence that perhaps richer societies will just hold more wealth in cash or vault cash anyway.
I liked the part above where Joe Gagnon suggested the BoJ should buy equities, not because of the event itself, but to show that it is an option to continue to buy assets, even private assets, whenever monetary policy is seem as less powerful…
8. July 2016 at 07:07
Rates-of-change in monetary flows, M*Vt, or our means-of-payment money supply times its transactions rate-of-turnover = roc’s in all transactions in Yale Professor Irving Fisher’s truistic “equation of exchange”: where M*Vt = P*T; and not M*Vi = P*Q (as on Milton Friedman’s license plate).
There are 6 seasonal inflection points each year. And these seasonal factors are determined by monetary policy. These seasonal factors are scientific proof that the “Member Bank Reserve Requirements — Analysis of Committee Proposal” published 2/5/1938 was correct.
The 5th seasonal inflection point this year is 7/20/2016. But there is an anomaly this year at Nov. month-end. Inflation crashes for one month in Dec. And then the trajectory for inflation is lower in 2017 than 2016.