Scott Sumner  

Sumnerian luck and monetary reform

PRINT
Weekend Grab Bag... Finding Out the Implications o...

I can't recall how many times I've visited some tropical paradise, only to be told, "that's funny, it usually never rains at this time of year." When I move to California, I can guarantee that their beautiful Mediterranean climate will suddenly become more like Morocco, due to global warming. Whenever I hear about a foolproof way to beat the market, it seems to immediately stop working the minute that I try it out.

Screen Shot 2016-06-25 at 10.45.08 AM.png
So how can we use my bad luck to make monetary policy more effective?

Commenter Zathrus recently asked me the following:

Second, moving from Brexit to NGDP futures targeting, how could it be politically sustainable to set monetary policy with NGDP futures markets if there is a perception out there, even among market commentators like the one above on Bloomberg, that prices are being set by the elites? I would think that as soon as the market started doing something politically unpopular there would be overwhelming political pressure to detach monetary policy from the futures market and go back to what we have now (discretionary policy, bound by some rules). If market participants believed this was a possibility, a monetary regime attached to even a highly liquid NGDP futures market would never have the credibility to be successful in the first place. How could these political credibility issues possibly be overcome?
It turns out that both of these concerns can be easily addressed with a monetary reform that leverages my gloomy disposition.

Under the basic NGDP futures targeting proposal, the purchase and sale of NGDP futures contracts automatically adjusts the money supply (and hence interest rates), in a way that the market believes will lead to on-target future NGDP. This is the proposal that Zathrus thinks is too controversial, and he's probably right. As a result I've recently moved over to a version of Bill Woolsey's "index futures convertibility" approach, which is more analogous to the old gold standard, except that gold is replaced with NGDP futures. Under this plan, the Fed would have unlimited discretion to set the money supply and interest rates wherever they wished, as long as they made NGDP futures contracts convertible at a fixed price. This could be viewed as the central bank providing insurance to anyone concerned about NGDP volatility. In my plan, they would implicitly charge a price, and hence earn a profit. For instance, I've suggested a "guardrails" approach where the Fed takes a long position on 3% NGDP growth contracts, and a short position on 5% NGDP growth contracts. That means the Fed profits whenever the actual NGDP growth is within those guardrails. And yet, a three to five percent range is small enough to provide decent macroeconomic stability.

So far, you might wonder what any of this has to do with the little raincloud that always hovers over my head. It turns out that the existence of that raincloud is the key to making the entire system work. Back in late 2008, I would have been rubbing my hands together with glee. It was obvious that NGDP growth for the next 12 months was going to come in far below 3%. In that case, I would have had an easy way to get rich---just put my entire 401k plan into a short position in NGDP futures, and if you assume a reasonable margin requirement, I would probably have doubled my money. That would have pushed me from upper middle class to lower rich class.

Unfortunately, there's the Lucas Critique. It's very unlikely that someone like me will ever be presented with an easy path to riches. That means that if the Fed installs a monetary regime where obvious likely policy failure provides an easy path to riches, then as soon as the regime is implemented, there will no longer be obvious likely policy failures. In other words, opportunities like 2008 (or the 1965-81 period on the high side) will no longer occur. Darn it!

Of course for this plan to work, it's essential that I stay alive, so that there is someone unlucky enough to make the entire system work. And I'm already 60. But surely among the 7.3 billion people in the world, there is at least one other person as unlucky as me.

Screen Shot 2016-06-25 at 11.23.28 AM.png
So I'm confident that even after I am dead the plan would continue working.

And if the policy fails, we would all have at least one silver lining to look forward to---it would be extremely easy to get rich!

PS. This is obviously partly tongue in cheek, but on the actual policy proposal I'm dead serious. Don't assume that you can find an obvious flaw of gimmick, just because it sounds silly.

PPS. I provided some responses to questions raised in David Henderson's recent post.


Comments and Sharing


CATEGORIES: Monetary Policy




COMMENTS (13 to date)
bill writes:

Can you walk me through the mechanics if NGDP is overshooting the 5% guardrail? I totally get what happens when we miss the 3% as the payments by the Fed also increase the money supply and should themselves push us back inside the guardrails. Thanks.

Robert writes:

Completely off topic, but your opening paragraph reminded me of Rob McKenna from So Long, and Thanks for All the Fish.

“And as he drove on, the rainclouds dragged down the sky after him, for, though he did not know it, Rob McKenna was a Rain God. All he knew was that his working days were miserable and he had a succession of lousy holidays. All the clouds knew was that they loved him and wanted to be near him, to cherish him, and to water him.”
bill writes:

Now I see it. The Fed is long on the 5% futures, so the private sector is paying it. Sorry. Makes sense.

Scott Sumner writes:

Robert, I feel his pain.

Bill, The Fed is actually short the 5% position. If lots of investors take the long position, then the Fed gets nervous about losing lots of money. It tightens policy until NGDP expectations are again below 5%.

Benjamin Cole writes:

If I pick a checkout line in a supermarket it will slow to a halt.

I like NGDPLT targeting. Just please select a robust target and err on the high side.

bill writes:

I'd be nervous with this set up. Hopefully the Fed would be competent enough to tighten and not have to pay out at times of NGDP growth in excess of 5% because the payouts would add to the growth. But the Fed's long position at 3% could be catastrophic, no?

Bill Woolsey writes:

Bill:

The futures transactions have no monetary effect and the payoffs don't necessarily have any such effect and almost certainly should not.

The automatic version that Scott has advocated involves parallel open market operations in bonds along with futures trades. It is the open market operations that cause changes in the quantity of money. It is only after NGDP statistics are calculated that futures are settled. At that time, there will be futures transactions for the next period's NGDP report, driving changes in the monetary base. Any monetary impact of payoffs for the last period's transaction would need to be offset.

I don't favor _requiring_ parallel open market operations, though open market operations would almost certainly be necessary to keep expected nominal GDP on target. I think the central bank should choose these open market operations. If the central bank ends up with a net position on the futures contract, it will be taking a risk of loss (and have an opportunity for gain.) The conservative approach of avoiding risk is to adjust open market operations so that the shorts and longs in the market match. While I think that is probably the best policy, I don't think it should be mandated.

If the central bank does take a position, siding with either the longs or shorts, and it loses or earns money, then it would be foolish to allow that to impact the quantity of money. The quantity of money needs now needs to be consistent with keeping expected nominal GDP target. It shouldn't be based on the errors the Fed made in the recent past.

Bill Woolsey writes:

Bill:

The futures transactions have no monetary effect and the payoffs don't necessarily have any such effect and almost certainly should not.

The automatic version that Scott has advocated involves parallel open market operations in bonds along with futures trades. It is the open market operations that cause changes in the quantity of money. It is only after NGDP statistics are calculated that futures are settled. At that time, there will be futures transactions for the next period's NGDP report, driving changes in the monetary base. Any monetary impact of payoffs for the last period's transaction would need to be offset.

I don't favor _requiring_ parallel open market operations, though open market operations would almost certainly be necessary to keep expected nominal GDP on target. I think the central bank should choose these open market operations. If the central bank ends up with a net position on the futures contract, it will be taking a risk of loss (and have an opportunity for gain.) The conservative approach of avoiding risk is to adjust open market operations so that the shorts and longs in the market match. While I think that is probably the best policy, I don't think it should be mandated.

If the central bank does take a position, siding with either the longs or shorts, and it loses or earns money, then it would be foolish to allow that to impact the quantity of money. The quantity of money needs now needs to be consistent with keeping expected nominal GDP target. It shouldn't be based on the errors the Fed made in the recent past.

Lorenzo from Oz writes:

I misread the title when it turned up on my feed and thought you were going to write about early Mesopotamian money :)

bill writes:

So the Fed has to pay if NGDP falls outside the target range.

Scott Sumner writes:

Bill, You said:

"But the Fed's long position at 3% could be catastrophic, no?"

I sure hope so, I'd love to be filthy rich.

Seriously, I think you missed the point of the post.

Lorenzo, And let's not forget that the Iraqis invented money.

Jose Romeu Robazzi writes:

@Bill
Sumner/Woolsey proposal is a short NGDP volatility position by the FED.

If you are familiar with financial markets, short volatility positions are winners in the long run if you have unlimited deep pockets (in a single strategy), or if you have a multitude of uncorrelated strategies that earn a premium (casino) or both.

Guess what?! The Fed is both the house AND has unlimited deep pockets when NGDP is concerned, apart from having an unlimited time position holding horizon, so....

The fact that there is academic opposition to this idea is not only surprising, it is almost preposterous....

Zathrus writes:

Hi Scott, thanks for answering my question. I like this modification. It's a great way of introducing market discipline while while avoiding the controversy of having market-triggered OMOs. Hopefully the prospect of massive losses would be enough to focus the minds of central banks that believed they were "out of ammunition" or thought they were right when everyone else was saying they were wrong.

I have to admit the cynic in me says that the the political process would end up dismantling the system if central banks made big, persistent errors, lost lots of money, and managed to convince the public that it wasn't their fault. But I guess that will always be the case in a democracy. Hopefully the guardrails would stop central banks from going down the wrong path in the first place (eg. on a downward spiral), and at the very least it would give them another hurdle to jump over before they could give up on the target.

Comments for this entry have been closed
Return to top