How to test market monetarism
In his recent paper describing “market monetarism,” Lars Christensen suggests there is a need for more empirical research on the impact of monetary policy:
Finally, while the Market Monetarist method of “story telling” and case studies clearly has value, there is also a need for more hardcore econometric testing of some Market Monetarist views, such as Scott Sumner’s view that monetary policy works with long and variable leads.
In my view the best way to study the effects of monetary shocks is to look at market responses to monetary policy announcements. For example, in my monetary economics class we always begin by studying the January 3, 2001, Fed announcement, which cut the fed funds target from 6.5% to 6.0%. Here were some market reactions:
1. The S&P fell by 5% in one day. (Update: should have said rose 5%)
2. T-bill yields fell slightly.
3. Long term T-bond prices fell 2% to 3%, meaning long term bond yields soared much higher on the news.
4. Long term TIPS yields rose modestly.
The fall in short term yields represented the liquidity effect from an expansionary surprise, whereas the rise in stock prices and TIPS yields represented the income effect, as markets expected faster RGDP growth. And the much larger rise in long term conventional bond yields represented both the income and Fisher effects, as inflation expectations also increased sharply.
Not all monetary policy announcements generate this pattern. The question is why. One approach would look for correlations between market responses and economic conditions. I believe market responses are quite different when the markets are worried about the adequacy of aggregate demand, compared to periods where there is confidence about future AD growth.
If I was young I’d try to do a sort of “Monetary History of the US” using market reactions to monetary policy news, rather than economic responses to changes in the monetary aggregates, as Friedman and Schwartz did. I did try to do that for the Great Depression, and while it is certainly quite time-consuming, it also proved quite fruitful.
PS. I surveyed various other bloggers about a name change, and got 5 different answers from 6 bloggers (neo-monetarism, new monetarism, post-monetarism, market monetarism, and monetarism.) If you put 7 economists in a room . . .
Perhaps we should just let the market decide.
Tags:
17. September 2011 at 07:12
” The S&P fell by 5% in one day.”
check that
17. September 2011 at 07:13
Scott,
there is a mistake @ “1. The S&P fell by 5% in one day.” You meant rose, I take?
“If I was young I’d try to do a sort of “Monetary History of the US” using market reactions to monetary policy news, rather than economic responses to changes in the monetary aggregates, as Friedman and Schwartz did.”
Don’t you teach a class?
Assignment 10% of the grade: Pick a fed announcement between and and explain what happened using Mishkin’s four indicators of monetary policy in approximately 500 words. Tell me which you’re going to pick so you don’t pick one that someone else has picked already.
😛
17. September 2011 at 07:14
and and were supposed to be year and year in brackets.
17. September 2011 at 07:19
Scott, thanks for posting this. I continue to think there is need for econometric testing and this obviously could based empirical tests of policy announcement. I however think you can do it the other way around – look at market to identify changes in money demand and in monetary policy.
Furthermore, monetary policy is often “passive” – when the central bank fails to respond to changes in money demand. This obviously is what happened at the start of the Great Recession. How do we test that? Again market moves probably would be the best starting point.
17. September 2011 at 07:21
Scott,
Useful, the facts may not be of evidence standard but enough anecdotal evidence without much suggestig alternative explanation should be enough in political economy (not a science anyway). For policymakers it is plausibility rather than proof that counts.
I like Lars’ paper a lot, it has been very helpful. The problem with non status quo approaches is that policymakers are risk averse. One cynical explanation for the excessive use of elaborate models in policy economics is that policymakers (and their lawyers) like to rely on advice with a bit of an engineering flavor. The New Keynesian models do that and in such a way that the naked eye will have great difficulty second guessing the results.
It may be a bit analogous to the use of Latin in the medieval church. The priests were benevolent but did not like to be bothered. Your movement is dangerous, because you may be right (I would give the NDGP targeters the benefit of the doubt, it is not clear to me how it can be done effectively without throwing away lots of things politicians and the public got used to. Like the medieval Church, there were lots of people who agreed with the need for reform, but they would have prefered to stay in the Church rather than join the “heretics” (or “reformers”) and improve it from within. In hindsight, that would have saved the world a lot of trouble.
17. September 2011 at 07:27
I wouldn’t recommend the name “market monetarism.” “Market” is a bad word to many people. But seriously, it makes it sound like you have to be a libertarian or a free-market type to subscribe to the view (when it really shouldn’t matter).
17. September 2011 at 08:10
Rien Huizer mentions “the excessive use of elaborate models in policy economics.” But the actual economy is fabulously complicated. It will often seem plausible to hope that a more complex model will bring us closer to that reality.
17. September 2011 at 08:19
How did markets react to big discoveries of gold in the gold standard era? That was a big exogenous change in the money supply, a perfect natural experiment.
PS: I like the term “market monetarism”.
17. September 2011 at 08:22
GU, Isn’t “markets” the focus of what most economist do? How can that be bad? And you see from Scott’s 2001 example that MARKET reactions to monetary policy decisions is at the core of what Scott and the other bloggers in this new tradition is blogging about. Anyway, I will not try to make the argument too strongly as I mostly care about the content of the “movement”.
neo-new-post monetarism doesn’t really work for me. Neo/New Monetarism is already taken by Steven Williamson (are there others??). Post – I think Post Keynesianism – that might not be too bad in the sense of economists like Clower and Leijonhufvud has a lot of clever things to say about monetary theory. Clower clearly has been a aspiration for Nick Rowe. By the way what is the name of the 1986 collection of paper by Clower? I am sure Nick will tell you that you all should buy it…
The titel is “Money and Markets”…there you go…
17. September 2011 at 08:57
In footnote 11 of Lars’ paper he notes that “Market Monetarists in general have not acknowledged the work of Johnson and Keleher.”
I agree but would go further and suggest they are also ignoring or are ignorant of the monetary work of “supply-siders” Jude Wanniski(Polyconomics) and David Ranson(Wainwright Economics),
both of whom extended the work of Johnson and Keleher by building monetarist models using market price signals. Jude’s monetary thought can still be found on the Polyconomics website (but is not fully covered in his book “The Way The World Works”). In fact, when it comes to Jude’s monetary model I would argue that “supply side” economics is a misnomer. A better name might be…”market monetarism”.
17. September 2011 at 09:31
I like “Market Monetarism” and hats off to Lars for that. “Free Market Monetarists” might be a tad better, primarily for PR reasons.
Right now we have to get the right-wing on board, as the left-wing is either clueless or does not care. Slogans and rhetoric are effective tools.
“Free Market Monetarists!”
17. September 2011 at 09:40
Friedman or Greenspan might track interest rates, or announcements, but that’s not how you trade. Historically, the markets have always reacted predictably – they haven’t changed their behavior.
Monetary policy objectives should not be in terms of any particular interest rate, or range of growth of any monetary aggregate. Rather, policy should be formulated in terms of desired roc’s in monetary flows (MVt) relative to roc’s in real GDP.
17. September 2011 at 09:59
“Not all monetary policy announcements generate this pattern”
The passenger riding next to me the week before the announcement told me he just margined up on Cisco. I told him to dump it & sell it short.
It’s a simple mathematical truism. MVt=PT. Combining real-output with inflation to obtain roc’s in nominal GDP, can then be used as a proxy figure for roc’s in all transactions. Roc’s in real GDP have to be used, of course, as a policy standard
17. September 2011 at 10:10
Leo, Oops, I should have said rose 5%–I just corrected it.
Martin, That’s a good suggestion.
Lars, I agree, but passivity is hard to measure.
Rien, The church analogy may be a good one, but I think progress in macro proceeds slightly faster than progress in theology. There is at least slightly less resistance to change.
GU, You said;
“”Market” is a bad word to many people.”
I can’t imagine why someone would view it as a bad word (unless they were socialists.) Markets are an important part of my approach, so I’m not sure I could ever please someone for whom markets were a bad word.
Libertarians are certainly not the only people who embrace markets.
Philo, But these models use the wrong kind of complexity–they are naive about the areas where complexity is needed (say the way monetary policymakers respond to fiscal stimulus, and are complex in areas where it doesn’t help (like trying to model all the various sectors of the economy.)
Alleged Wisdom, I’m not sure, but I did study the way markets responded to gold market shocks in the 1930s (mostly demand shocks) and it was consistent with market monetarism.
Leo, I am very aware of supply-side views on monetary policy. This was an offshoot of the “New Monetary Economics” of the early 1980s (Black, Fama, Hall), and was influenced by people like Mundell.
This approach also goes back to the 1930s with people like George Warren.
flow5, I advocate stable NGDP growth, is that your goal as well?
17. September 2011 at 10:14
flow5, What are rocs?
17. September 2011 at 10:50
‘Market Monetarism’ does not roll trippingly off the tongue, but ‘neo-Monetarism’ does. And, as others have pointed out, it has a tradition.
Nor does it ‘mean pork chop in Arabic’ (nor ‘it doesn’t go’, in Spanish, to name a famous blunder by Chevrolet in the 60s).
17. September 2011 at 12:04
“Not all monetary policy announcements generate this pattern. The question is why.”
I’ll take a stab.
During normal times (mid 90s), a Taylor Rulesque policy roughly kept NGDP expectations relatively stable regardless of the Fed funds rate. If markets thought it was too high, they assumed it would be reduced later on to keep inflation and (in the absence of supply shocks) NGDP stable. When the economy experiences an oil shock however, the Fed seems to under react to bad conditions and focus more on inflation. Long term NGDP expectations fall, and for good reason: Every recession since the 80’s has been followed by a permanently lower NGDP trend.
If you read the January 3rd 2001 statement it sounds a lot like the Fed was shifting focus from inflation to NGDP.
“These actions were taken in light of further weakening of sales and production, and in the context of lower consumer confidence, tight conditions in some segments of financial markets, and high energy prices sapping household and business purchasing power. Moreover, inflation pressures remain contained. Nonetheless, to date there is little evidence to suggest that longer-term advances in technology and associated gains in productivity are abating.”
From an inflation targeting perspective, higher energy prices merit tighter, not easier, policy. Could this be why the announcement affected markets more than usual? Sure, the Fed should have done more, but overall the recession was quite mild.
17. September 2011 at 12:51
Consumer Goods +3.8%
Finished Goods +6.5%
Intermediate Goods +10.3%
Crude Goods +18.4%
With inflation numbers like these, all AD based economic theories about unemployment should be completely discredited by now. Stop the bullshit, cut regulations and spending and quit printing money if you want economic health.
17. September 2011 at 17:45
John:
Spending on final output remains 14% below the trend of the Great Moderation. Real output and employment are about 10% lower.
What is your explanation for the increase in intermediate goods, finished goods, and crude goods over the last year?
17. September 2011 at 18:48
Patrick, I guess we can rule out “Nova monetarism.”
Cameron, It might have been NGDP, but the Fed also talks about core inflation a lot, so that may be why they put less weight on oil prices.
I agree about the lower NGDP trend, the Fed calls it “opportunistic disinflation.”
John, You said;
“With inflation numbers like these, all AD based economic theories about unemployment should be completely discredited by now.”
Except those AD-based theories that say pay no attention to inflation, like one prominent AD-based theory I can think of (market monetarism.)
18. September 2011 at 01:37
John:
“With inflation numbers like these, all AD based economic theories about unemployment should be completely discredited by now. Stop the bullshit, cut regulations and spending and quit printing money if you want economic health.”
I saw these numbers too and I was thinking about it too. I’ve concluded now, as the news reported that profits were up earlier, that margins are being reduced to normal.
Think about it, if volume (NGDP) went down and profit went up that means that margins must have gone up. Now that volume is up, margins are being squeezed and that is what these numbers indicate.
If Keynes was right, that the classical case was just a special case when the economy is at full employment, then when the economy is short of full employment, firm market power should go up because resources are idle.
Why should it go up because resources are idle? It is a self-enforced quantity restriction. It is a NE where all players are better off by restricting quantity. Why? Wages are too high in terms of output. Reduce wages in terms of output and resources will be mobilized by competitive pressure and that same competitive pressure should lead to a margin squeeze.
18. September 2011 at 01:38
@John,
If anything, therefore, these numbers vindicate AD-based theories.
18. September 2011 at 12:47
Just came to think about 2001 and market reactions to “events”. As a market participant in 2001 I think (other than the horrors of that year) I remember best the “reverse” reaction to macroeconomic numbers. During a period during the first half of the year weak US macroeconomic lead to market reactions, which on the surface seemed quite odd. Weak US data lead to increases in the US stock markets. I am not sure whether anybody has studied that but I have the feeling that this was a pretty clear tendency during the first months of 2001.
One can only speculate about the reasons for this reverse reaction to negative surprises. At that time the financial media often explained strong equity market performance on the back of weak numbers as an increased likelihood of monetary easing. That would be consistent with a credible “LAW” regime (“Lean-Against-the Wind”) where monetary policy stabilises NGDP or some other nominal variable. I am far from sure that this can be proven empirically, but it could be interesting to test – and compare with the market reaction to weak number during the Great Recession.
The hypothesis could be that during the Great Moderation the LAW regime was credible so negative macroeconomic numbers should not make the stock market drop as market participants would expect monetary policy action to make up for the negative surprise. However, during the Great Recession market participant no longer expect monetary accommodation in reaction to weak macro data.
So here is an empirical test. Looking at large macroeconomic surprises positive and negative during the Great Moderation and the Great Recession. Look at the market reaction to these surprises on the day of the day release. On the dollar, S&P500, 30 yields and TIPS breakeven inflation expectations. If there is a significant difference between the two periods in terms of market reaction then it is an indication that the Market Montarists are on to something…
19. September 2011 at 10:03
Lars, A similar test has already been done, Glasner did a test (cited by Krugman) that showed TIPS spreads were positively correlated with stocks since 2008, but not before.
19. September 2011 at 13:17
“roc’s” just means rates-of-change (not absolutes).
Stable ngDp growth rates (roc’s) is the best target for the Fed’s technical staff. It’s one they can actually hit.
If they were a little more sophisticated they would target real-gDp + 2-3% roc’s in inflation. But then you’d have to eliminate their employment mandate (which they can’t hit anyway).
20. September 2011 at 03:18
Scott, I know of David’s study. I think a more event oriented might be warranted with a focus on market reaction to macro releases.
20. September 2011 at 12:31
flow5, I’d rather they ignore inflation, and just stick with NGDP targeting.
Lars. Yes, that would be useful to look at.