When I engage in monetary policy research, I try to write papers that will seem persuasive in the future. That may not always be a wise decision, and indeed back in the 1980s, some of my colleagues suggested that I might be better off doing more conventional research, instead of writing papers advocating the targeting of NGDP futures prices.
What was conventional monetary economics during he 1980s? I recall lots of papers such as “Money Demand in Turkey”, or “Money Demand in South Korea.” Today there is a lot less interest in studying money demand, perhaps because the topic doesn’t seem that useful in a world where central banks do not target the money supply (nor is the money supply determined by factors such as the gold stock.) Today there are lots of papers that use interest rates as an indicator of monetary policy, perhaps because central banks target interest rates.
A few weeks back, Tyler Cowen linked to a NBER paper by Claudia M. Buch, Matthieu Bussiere, Linda Goldberg and Robert Hills that tried to measure the international transmission of monetary policy. But what is monetary policy?
We also consider two commonly used alternative measures of the monetary policy stance. The first is the actual short term policy rate, defined in first-difference form in order to account for changes in the policy stance. For the US, the effective federal funds rate (FFR) is a typical measure. The rate used should represent an average funding cost for banks. The second measure is the degree of quantitative easing (QE) capturing the volume of central bank liquidity provision such as the change in the central bank’s balance sheet relative to GDP.
Apart from the choice of the proxy for the monetary policy stance, the issue of identification of monetary policy shocks arises. Otherwise, the sensitivity of banking variables to the policy measure is polluted by the effects of the variables to which policy rates respond. Identification of monetary policy shocks has, arguably, been one of the most contentious issues in the macroeconomic literature, and we cannot settle this debate here.
While the authors admit that identification is a tricky subject, it does not stop them from pursuing their research on the international transmission of monetary policy.
I believe it should have stopped them. The problems are so great that any study of monetary policy using interest rates to identify policy shocks is unreliable. Interest rate-oriented monetary studies are today’s version of the money demand studies of the 1970s and 1980s.
I would encourage younger academics to avoid doing research in areas that are currently popular, and instead try to focus on areas that will be popular in about 30 years. In 1986, I thought that the futures targeting approach would be the wave of the future, and so far I’ve been wrong. But hope springs eternal, and I continue to believe that this approach will catch on at some point in the future.
How about the identification of monetary policy? If there are any readers who still believe that interest rates are a good way of identifying changes in monetary policy, I encourage you to read the literature on NeoFisherism. Not because I believe the NeoFisherians are correct (I think they are wrong), but rather because the rise of the NeoFisherian view clearly suggests that there is something profoundly wrong with the tendency of mainstream economists to equate low interest rates with easy money.
If interest rates are a dead end, what will replace this variable? I can’t be sure, but my own preference is to use expected NGDP growth (currently 4.6%) as an indicator of the stance of monetary policy (currently expansionary). I’m not confident that NGDP growth will be the accepted indicator of the 2040s, but I am confident that the money supply (widely used as a policy indicator in the 1980s) and the nominal interest rate (widely used today) will both be thoroughly discredited within 30 years.
If you plan to do research on the impact of monetary policy on some other variable(s), then please pick a monetary policy indicator that other economists are not using. Better to aim high and end up an unconventional academic failure like me, rather than have a modest success writing papers that no one will care about in 30 years.
PS. Note that the traditional interest rate policy indicator has led some pundits to conclude that US monetary policy has recently tightened, whereas the NGDP growth indicator more accurately points to a recent easing of policy.
READER COMMENTS
Michael Sandifer
May 1 2018 at 1:38am
Changes in NGDP growth expectations can be gleaned from changes in broad stock indexes, Treasuries, commodities, and exchange rates with more precision than is commonly recognized. I think it is more likely that if market forecasts for NGDP are ever targeted, an approach using existing markets will prevail. I’ve made significant progress developing such an approach and am launching a website soon that will offer streaming, market-based NGDP forecasts.
An AI component is also being developed to improve the approach, adding forecasts for other variables in the process.
The NGDP futures market targeting idea seems a good one, but unfortunately it seems much like many ideas of Shiller for market securities that seemingly would make the world a better place. For some reason, the ideas don’t sell.
It’s unfortunately hard to imagine something as widely misunderstood and potentially controversial as a subsidized NGDP futures market being adopted by a major central bank any time soon.
On the other hand, central banks could start targeting market forecasts like mine without even necessarily initially admitting they’re doing so. They can also just initially study the forecasts and determine whether they expect targeting such forecasts to improve monetary policy.
I think Lars Christensen also has a market-based approach to estimating changes in NGDP growth, so I’m not the only one on this path, but my approach is unique.
Keenan
May 1 2018 at 9:36am
Scott,
I have 1 issue with NGDP futures targeting. This market will be much less liquid than other markets like the S&P 500 Index Futures or Eurodollar Futures. It doesn’t matter how liquid they are so long as they are less liquid than other markets (new markets are always less liquid, and less volatile markets will be less liquid – not as interesting to trade)
There will be situations where agents will move the NGDP market for the sole purpose of moving other markets. Let’s say that the current Fed Funds Rate is 2% and is priced to stay that way for a while. NGDP is exactly at 4.5% and expected to continue there. Monetary policy is humming along. One could place a bet in the eurodollar market that the Fed will cut at the next meeting, and get 50:1 odds. Let’s bet $20mm to win $1b.
Now, NGDP market is currently at 4.5%, exactly what the fed is targeting. All of a sudden, over the course of a day, that goes down to 3.4% as this agent sells futures, not because he has any useful information, but precisely because he wants the Fed to cut at the next meeting. $20mm worth of selling might be enough to move this market a dramatic amount, it will depend how liquid it is. The Fed will see the market moving and maybe think twice about keeping the rate constant at the next meeting.
I agree with the principle of NGDP Futures Targeting, and especially with targeting a market forecast, but if you explicitly target a market forecast, the market forecast may stop being as useful.
What are your thoughts?
Brian Donohue
May 1 2018 at 10:03am
Good post.
Hey Scott, the Fed has reduced its balance sheet about $100 billion in the past several months. Do you agree that further “tightening” should continue this trend, rather than further increases in the FFR.
Also, isn’t IOER, and particularly above-market IOER, a massive subsidy from taxpayers to banks at this point which also discourages banks from getting out there and lending money?
Hunter
May 1 2018 at 11:39am
Does NGDP include the sales of stocks? If not how would adding that to NGDP affect it?
Scott Sumner
May 1 2018 at 1:28pm
Michael, That sounds like an interesting project.
I don’t need to sell NGDP futures to the public, I need to sell the idea to the Fed. It doesn’t matter if the public is not interested.
Keenan, Those are common criticisms, but none have any bearing on the “guardrails” approach I am now advocating. Check out my latest Mercatus paper on the topic:
https://www.mercatus.org/publications/promise-nominal-gdp-targeting
Brian, They should unwind the QE first. I don’t view IOER as a major subsidy, as they pay close to market rates. But perhaps it’s a small subsidy, as rates are slightly above T-bill yields. I oppose IOER.
Hunter, No, financial transactions (except commissions) are not and should not be a part of NGDP. They are not income.
Michael Sandifer
May 1 2018 at 1:54pm
Scott,
Yes, I mean selling the idea of the Fed subsidizing an NGDP futures market and then level targeting the market forecast will be a tough sell to the Fed itself, in part because I think they’d open themselves up to controversy, as nonsensical as it would be. It’s easy to imagine some fringe elements on both the right and left charactering such a market as a subsidy to the rich, nasty speculators, etc. Once again, this would all be nonsense, but we’re increasingly governed by nonsense.
And I like Shiller’s trill idea, but he’s gotten nowhere in promoting it to the government. It doesn’t seem his even higher profile now is helping him.
Brian Donohue
May 1 2018 at 6:09pm
Thanks Scott. Do you agree that unwinding QE makes more sense than inverting the yield curve?
On the issue of IOR, banks hold more than $2 trillion in reserves at the Fed, earning 1.75%. That’s $36 billion per year. The 0.25% “subsidy” above market rates is worth more than $5 billion per year by itself.
In 2016, banks enjoyed aggregate earnings of $170 billion. Proceeds from IOR are not insignificant, and taxpayers are paying for it.
Scott Sumner
May 2 2018 at 10:01am
Michael, Controversy doesn’t stop the Fed from paying IOR, which helps big banks.
Brian, I’d do both–unwind excess reserves, and raise rates.
Michael Sandifer
May 2 2018 at 5:29pm
Scott,
It’s probably safer to say that if a good enough market forecast is already available, there’ll be less incentive to create an NGDP futures market.
Roger Sweeny
May 6 2018 at 11:08am
Scott, after I read this, I was reading Herbert Stein’s Presidential Economics (first edition, 1983). In chapter 9, “Toward a New Consensus”, he seems to endorse NGDP targeting (p. 338):
“First, the Fed would aim only at nominal targets–price level and nominal GDP. … Second, it would derive its money-supply targets from its nominal GDP targets and a prediction of velocity …”
Not exactly the same as what you have proposed, and I’m sure you know about it, but I was surprised to find someone saying this 35 years ago.
Comments are closed.