My preferred monetary policy would involve stabilizing the price of NGDP futures contracts. Under this regime, important variables such as the money supply and interest rates would be determined by the market, set at a level that the market expected would lead to on-target growth in NGDP.
This sort of regime is not likely to be adopted in the near future, and thus I’ve also recommended a long series of incremental policy reforms that are more politically feasible. These include:
1. A more nimble process for adjusting the policy instrument.
2. A “whatever it takes” approach to setting the policy instrument.
3. Targeting the market forecast.
4. Level targeting.
As I look back on all these posts, I see a pattern. My “pragmatic” suggestions represent an attempt to mimic the performance of a market-based regime, under the existing (discretionary) regime. I.e., sort of like when the Singapore government instructs (state-owned) Singapore Airlines to “run yourself like a private company”.
1. I’ve recommended that the target interest rate be adjusted daily to the nearest basis point, set at the median vote of the voting FOMC members (which might be emailed in to the Fed each day.) Thus on any given day, rates would be equally likely to go up or down. This is very different from the current regime, where adjustments are made by at least 25 basis points, at six week or longer intervals, with a presumption that the next change is likely to be in a certain direction. My proposal is closer to how real world asset prices move, closer to a random walk. More importantly, my proposal is closer to how the interest rate would move under an NGDP futures targeting regime.
2. When economic or financial conditions change dramatically, the Fed tends to move more slowly than the markets. Thus if the natural rate of interest declines, the Fed is likely to cut rates, but not as fast as the fall in the natural rate. I favor a “whatever it takes” approach, where the rate is always moved aggressively enough so that policy goal is expected to remain on target. When interest rates are at zero, I favor a QE program that is aggressive enough to insure that policy remains effective, even if it means buying up the entire world. (And there are worse things that owning the whole world!) Some might point to the danger of hyperinflation, but if that’s actually a risk then you don’t have a demand shortfall problem; you don’t need stimulus.
Again, this is how markets behave. When new information causes an equilibrium asset price to move sharply, market prices immediately move as far as necessary to restore equilibrium, not in a series of small quarter point moves, spaced out over time.
3. I’ve advocated that the policy instrument be set at a position where the market forecast of growth in the policy goal variable (say NGDP) was equal to the target. Here the market analogy is a bit weaker, as markets don’t have policy targets. Nonetheless, this policy idea is consistent with the spirit of efficient markets, where prices reflect the wisdom of crowds. Thus if the 2-year forward fed funds futures is trading at 97.8 (par value 100), that can be seen as a market forecast that the actual fed funds rate will be roughly 2.2% in early 2021. Current asset prices reflect the collective wisdom of traders regarding futures cash flows, interest rates, among other variables. My “target the forecast” idea relies on the wisdom of crowds in a very similar fashion.
4. There is no exact market equivalent to “level targeting”, as markets don’t target variables. However, as with level targeting of the price level or NGDP, efficient markets do not engage in “let bygones be bygones” when mistakes are made. Thus if a certain stock ends the day trading at $47, and unexpectedly bad news comes out an hour later, it will immediately become clear that the market made a mistake. Traders do not simply ignore this mistake; they try to correct it on the following day by moving the stock price to the appropriate level. Traders don’t focus on the appropriate rate of change in stock prices, it’s about returning the price to the equilibrium level, after forecasting mistakes cause a temporary deviation from that equilibrium.
Similarly, under level targeting, a deviation of the price level or NGDP from the target path calls for a later adjustment to bring it back to the pre-determined trend line.
All four ideas in this post have been discussed in previous posts, but I think it’s important to stand back and see how they are all part of a pattern. If I were invited to the Fed’s June conference on monetary policy reform ideas, I would not waste time advocating NGDP targeting. I have very little new to offer there, as the advantages of NGDP targeting are increasingly well understood. Instead I’d explain why Fed policy should continue to move in the direction of mimicking a market-based process. If I have anything both useful and novel to offer, then it is in this area.
READER COMMENTS
John hall
Apr 2 2019 at 1:37pm
I think there is room to improve #4.
There are many participants in equity markets. Short-term traders are merely one group. People with longer-term horizons do care about levels, but it depends on what levels you are talking about. The dividend discount theory of valuation is that a stock’s price corresponds to the present value of the stream of the dividends the share generates over time. If there is bad news, then either the future growth of dividends might be less or the risk premium associated with those dividends will decline, leading longer-term investors to update their estimates of fair value. When investors compare stock prices to next twelve month’s expected earnings, this is kind of a short-cut to what I described above. But it’s still a form of level targeting. It’s just that the appropriate level changes over time.
Benjamin Cole
Apr 2 2019 at 8:16pm
The more I think about it, the less confidence I have in QE, certainly within practical ranges of application. I suppose the Federal Reserve could buy all the assets on the planet, but long before that I think they run into practical limitations.
Also, capital markets are globalized. Ergo, the actions of any particular central-bank take place on a global playing field. The Fed can try QE in hopes of obtaining a hot potato effect—- but the Fed can only buy so many potatoes and there are a lot of hands on the planet.
There are additional concerns regarding banking systems. Japanese banks are complaining they can no longer make money with interest-rate stuck at zero or less. I do not really understand the argument, but commercial bankers globally cry for safe assets that they can trade around. They complain when central banks own too much of the safe assets, although my experience is commercial bankers always complain about everything.
I wonder if QE should always be married to federal deficits. We know then that the money is spent (generally) inside the United States. The Fed conducting QE alone appears to be an attempt to stimulate the domestic economy by lowering interest rates globally.
Of course, a more simple approach would be to sidestep the whole claptrap of the Federal Reserve, and allow the Treasury to print money to finance fiscal deficits.
Side note: stocks now trade premarket and post market and around-the-clock in other markets (even US equities). Traders do not wait to the next day to correct a stock price.
Matthias Görgens
Apr 2 2019 at 11:11pm
Shouldn’t global markets like you suggest make QE even more rewarding, not less?
That kind of global nature would suggest that the Fed could buy more of the world’s assets before running into inflation limitations. QE would then become a competitive endeavour between money issuers?
Scott Sumner
Apr 3 2019 at 12:38pm
Ben, You said:
“The Fed can try QE in hopes of obtaining a hot potato effect—- but the Fed can only buy so many potatoes and there are a lot of hands on the planet.”
I don’t think you understand the hot potato effect. The Fed doesn’t buy them; it creates them with a magic wand and then sells them.
Michael Sandifer
Apr 2 2019 at 11:32pm
Scott,
I’m curious as to how you see NGDP level-targeting using NGDP futures as opposed to simply having all electronic monetary transactions, particularly after paper money is eliminated, reported to the Fed daily with computer-automated daily adjustments to monetary policy. The latter would have the advantage of making the electronic payments systems more efficient to be able to facilitate daily reported.
Scott Sumner
Apr 3 2019 at 12:40pm
The elimination of paper money will change things (for the worse). As to its effect on our ability to monitor NGDP in real time, I can’t say. We already have enough non-currency transactions to do that, at least in theory. So why is it not being done?
Scott Sumner
Apr 3 2019 at 12:43pm
On second thought, perhaps the reason is that the vast majority of transactions (over 99% by value) are financial, which is not a part of GDP.
Michael Sandifer
Apr 3 2019 at 3:48pm
Scott,
We already categorize transactions by SIC codes. No reason we can’t also code them as being part of NGDP.
Scott Sumner
Apr 3 2019 at 5:40pm
OK, but then I would reiterate that this doesn’t require a cashless society; we could do this with 99% as much accuracy today, as compared to an economy with zero cash transactions.
I’m agnostic on the questions of how hard this would be, or why we aren’t doing it already.
Michael Sandifer
Apr 3 2019 at 7:51pm
Scott,
Yes, I’m not 100% for a cashless society either, but I agree with you that in think we will probably have that situation, possibly within my lifetime.
I favor a massive rollback on the assault in financial privacy.
Lorenzo from Oz
Apr 5 2019 at 9:36pm
Incremental reform, very practical Popperian of you.
Comments are closed.