The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« The Bank of England On Hold | Main | Yuan Appreciation Watch »

November 24, 2005

More Ado About M3

The fallout continues from the Federal Reserve Board's recent announcement that they will soon discontinue publication of the M3 monetary aggregate, and many of its components.  Tom Iacono -- who must surely need a new name for his blog any day now -- has a good review of recent commentary on the subject. (A tip of the hat to Dave Iverson.)  I'm still somewhat surprised by the sentiment that the Board's decision is, at least in part, motivated by the desire to downplay a statistic that appears to be contradictory to the achievement of price stability.  I'm surprised because such sentiment seems to imply that the FOMC places significant weight on the behavior of monetary statistics in the first place!

Let's go back in time, to the July 1993 Monetary Policy Report to the Congress.   Here is what Chairman Greenspan had to say in his testimony:

... at least for the time being, M2 has been downgraded as a reliable indicator of financial conditions in the economy, and no single variable has yet been identified to take its place.

By July 2000, this footnote appeared in the written report:

At its June meeting, the FOMC did not establish ranges for growth of money and debt in 2000 and 2001. The legal requirement to establish and to announce such ranges had expired, and owing to uncertainties about the behavior of the velocities of debt and money, these ranges for many years have not provided useful benchmarks for the conduct of monetary policy. Nevertheless, the FOMC believes that the behavior of money and credit will continue to have value for gauging economic and financial conditions, and this report discusses recent developments in money and credit in some detail

"Velocity" represents the number of times the stock of money turns over in support of spending, and it is the key piece of information policymakers need to to connect monetary measures to objectives.  (You can read a bit more about the theory here.)  In short, if you can't predict velocity, you really cannot predict how changes in the money supply will impact the rate of inflation.

With that thought in mind, the following picture of M2 velocity helps to explain why monetary statistics have been "downgraded " in monetary policy deliberations:



The lines represent the trends in velocity -- and the changes in those trends -- identified by formal statistical tests.  If you are interested in the formal details of these calculations, you can look them up in a paper by John Calrson, Ben Craig, and Jeff Schwartz that describes the methodology.  But you can also just believe your own lyin' eyes, which will tell you, I think, that velocity has not been very predictable over the past decade-and-a-half.

I would have shown you a picture of M3 velocity instead of M2 velocity, but for one problem.  One way to think about velocity is that it represents the part of money demand that is sensitive to changes in interest rates.  Estimating trends requires some estimate of the return on the monetary measure (so it can be compared to the return on nonmonetary assets -- giving us an idea of the "opportunity cost" of holding money). To the best of my knowledge, no regular estimate of the return to M3 is even available.  That alone speaks volumes.

So here's my last word on the subject.  If you have a sense that M3 is providing any information at all related to the objectives of monetary policy, you know something I don't know.

For all of you in the States -- or whose hearts are in the States -- Happy Thanksgiving.

November 24, 2005 | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference More Ado About M3 :







The Federal Reserve will raise its benchmark interest rate
to 4.5 percent in the first quarter from 4.0 percent now,
Citigroup said yesterday in a note to clients. The U.S. central
bank will bring the figure down to 4.25 percent by year-end,
matching the 10-year yield, it said

Posted by: stan jonas | November 24, 2005 at 11:33 AM

I will try the other approach and find out why M3 is so important for the ECB. It is the only figure that gets honored by the ECB with a regular monthly press release. Also of note is that Trichet and his chief economist Issing have been saying for months that M3 growth is a major concern to them and gives reason to think about hiking rates. When European central bankers talk about ample liquidity they always refer to M3.

Posted by: The Prudent Investor | November 24, 2005 at 01:25 PM

The ECB routinely accommodates overshoots of its "reference" rate for monetary growth, suggesting that it is only paying lip service to monetary aggregates.

Posted by: Stephen Kirchner | November 24, 2005 at 05:38 PM

"I will try the other approach and find out why M3 is so important for the ECB."

Yep, well if you come up with any meaningful answers, please let me know.

God, and the ECB both "move in mysterious ways" as far as I am concerned.

I won't say anything about Otmar Issing, since I've promised to be polite all this week.

Posted by: Edward Hugh | November 25, 2005 at 06:33 AM

"In short, if you can't predict velocity, you really cannot predict how changes in the money supply will impact the rate of inflation."

Yep, that's right. But somehow I keep hearing the argument that inflation is, always and everywhere, a monetary phenomenon.

For those who somehow find they can't treat this as gospel, here's an interesting article by Belgian economist Paul de Grauwe, published in the Scaninavian Journal of Economics, and entitled appropriately enough:

"Is inflation always and everywhere a monetary phenomenon?"


This issue has more than passing importance in the light of recent economic evidence from Germany and Japan, which suggests that "deflation" and "absence of inflation" is *not* always and everywhere a monetary phenomenon.

Now here's an extract from the paper:

"In this paper, we return to this issue using a sample of countries spanning the whole world over a period of 30 years. The key question we analyse concerns the link between inflation and the growth rate of money and how it depends on whether countries experience low or high rates of inflation.

Our results have some implications for the question regarding the use of the money stock as an intermediate target in monetary policy. As is well known, the European Central Bank continues to assign a prominent role to the growth rate of the money stock in its monetary policy strategy. The ECB bases this strategy on the view that ‘‘inflation is always and everywhere a monetary phenomenon’’. This may be true for high-inflation countries. Our results, however, indicate that there is no evidence for this statement in relatively low-inflation environments, which happen to be a characteristic of the EMU countries. In these environments, money growth is not a useful signal of inflationary conditions, because it is dominated by ‘‘noise’’ originating from velocity shocks. It also follows that the use of the money stock as a guide for steering policies towards price stability is not likely to be useful for countries with a history of low inflation."

Posted by: Edward Hugh | November 25, 2005 at 06:49 AM

In the same vein, here's some more grist to the mill from Paul de Grauwe in yesterdays Financial Times and talking about ECB President Trichet's volte face last week on interest rates:

"It is difficult to understand such a turnaround. One can only guess what underlies this “volte face” that put the financial markets on the wrong foot. Here is my explanation, which I cannot prove but which seems to make sense. There is a hard core of “hawks” in the governing council who cherish a monetarist agenda. This agenda has two items. One is to bring the rate of inflation back below 2 per cent, the only target the ECB has declared to be salutary and consistent with price stability. The second is to curtail the “excessive” growth of the money stock, M3, which has now reached an annual rate of 8 per cent (but this was also the case two weeks ago when the ECB declared that this was not sufficient to raise the interest rate). The hawks somehow managed to have the upper hand in the governing council and to make the president their spokesman when he came out declaring that a rate increase now had become necessary. This statement also makes a rate increase almost inevitable next month"

"The argument that a rate increase is necessary because the acceleration of M3 growth signals a higher future inflation is equally weak. In 2002-2003 the yearly growth rate of M3 was about 8 per cent during two consecutive years. Monetarists, reminiscent of Milton Friedman’s notorious predictions, forecast that this would inevitably lead to an upsurge of inflation two years later. It has not happened. Inflation moved in a flat band between 2 and 2.5 per cent. This will not surprise other central bankers who abandoned monetary targeting long ago after finding out that, in a low inflation environment, money numbers are very bad predictors of future inflation."

Posted by: Edward Hugh | November 25, 2005 at 06:56 AM

There seems to be a growing disconnect between what people are told the rate of inflation is, and what they experience in their own lives - and not just for energy. Healthcare, tuition, and many other service categories have been rising sharply, while at the same time the cost of imported goods remains stable or declines (how many DVD players do you really need?).

And then there's housing.

Congresswoman Carolyn B. Maloney put it best when asking the following question of Alan Greenspan before the Joint Economic Committee meeting on November 3rd:

"The question that my constituents ask me, I'm going to ask you, 'If the economy is so good and inflation is so well behaved, and there's price stability, then why does everything cost so much more when you go to buy something?"

Not just energy, "everything" (exclusive of DVD players, probably).

Over the last ten years inflation as measured by CPI-U has been in the 2-3 percent range, whereas money supply growth has been in the 5-10 percent range, with the fastest growth coming from the M3-M2 component, the reporting of which is being terminated.

The sense that I get is that the rise in prices felt by consumers is higher than what is being reported in government inflation statistics, and that past and future M3 growth is the uncomfortable confirmation of this.

P.S. (The name of my blog works equally well in the present and past tense).

Posted by: Tim | November 25, 2005 at 11:09 AM

First we define down cpi by a full point (Greenspan hyped Boskin recommendations), then we stop tracking our broadest money stock because it, in Wm. Polley's words, "has very little to do with monetary policy. But, the Fed's H.6 release indicates non-m2 M3 growth is mostly in large denomination deposits & eurodollars, not in repos. Isn't that monetary? It makes sense that a Fed that believes r.e. asset appreciation is NOT inflationary wouldn't want to track the "new money" created by asset appreciation. But, why aren't Economists questioning the Fed's decision?
Dave Altig, who knows the Fed, says the "Board" made the decision & he doesn't know all the factors that went into it. My experience is when staff's excluded from decisions, there's a good chance politics are involved.
For thousands of years inflation was addressed as the changing value of real assets over time. Now, the Fed (& Fed talking heads interpret inflation as "an increased quantity of money" and tell us the increase in cost of goods is responsive to the changed money supply. The confusion comes because the Fed (post Bretton Woods) rightly refuses to restrict its options by rigidly defining what "money" is. So, the question remains, if our broadest money stock is now deemed "NOT" money, what is money? And, why did the Fed descard M3 without first providing a "new" measure? The legitimacy of this question is validated by the agressive attacks on questioners of this Fed action. My response: 1. Repos are only a very small part of non-m2 m3 growth, too small to use as justification for abandonment. 2. Large denomination deposits are not attributable to "financial market innovation" & historically don't they better track inflation than m2? 3. Correlating current m2 to pre-Boskin cpi (+1%) would make a stronger argument for its recent relevancy over m3. 4. Ad hominem dismissals of "conspiracy theorists" is hardly a reasoned response worthy of honest brokers, the role I'd expect of Academic Economists.
Where does the pro-Fed spin from "independents" stop & when will Academic Economists recognize their societal
responsibilities? What's the value of tenure if it doesn't encourage argument independent of political consideration?
Below is a link to a blog by David Chapman maybe someone would respond to when time allows. I'm not familiar with him (probably a businessman) but I think he presents some strong points.

Posted by: bailey | November 26, 2005 at 09:17 AM

11/25/05 Prudent Bear's Doug Noland adds insight into M3 debate:
"A Quickie on “Money”:
“Money” connotes quite different things to different people. Some would argue that gold – a store of value over the ages that is nobody’s liability – is money in its purest form. Others have a more traditional (narrow) focus on currency and banking system reserves (“money stock”), and would generally analyze “money” primarily in terms of its role in consummating transactions. Many still view the money supply as something under the Federal Reserve’s control, holding “Fed pumping” responsible when the monetary aggregates expand rapidly. It is common for pundits to focus on what they believe “money” should be rather than the distinguishing characteristics of the creation, intermediation, risk profile, function and various effects of today’s extraordinary inflation of myriad financial claims.

And while most will view it as unconventional, I believe my “money” analytical framework is consistent with the thinking of some of the leading monetary economists of the past. Consistent with Ludwig von Mises’ “fiduciary media” approach, monetary analysis must be quite broad in scope and focused on the “economic functionality” of new financial claims. Allyn Abbott Young was keen to appreciate the “preciousness” attribute of money throughout history. It is the perceived preciousness (“moneyness”) of specific types of contemporary financial claims that leave them highly susceptible to over-issuance. Traditionally, when it came to financial claims expansion it was government issued currency and central bank created reserves that generally enjoyed the type of persistent (“store of value”) demand conducive to protracted Credit inflations. These days, the defining feature of contemporary Wall Street finance is the amalgamation of financial sector intermediation, the proliferation of credit insurance, financial guarantees, derivatives, implied and explicit GSE and government guarantees, and myriad sophisticated risk-sharing structures that have created to this point unlimited capacity to issue perceived “precious” financial claims. It is the Inherent and Dubious Nature of The Moneyness of Credit that Supply Creates its Own Demand.

I will loosely define contemporary “money” as financial claims perceived to be a highly safe and liquid store of nominal value. Simple enough, one would think, although it is a definition quite problematic for most. The catch is “perceived.” You can’t model perceptions, so my definition would be unacceptable to most academics, econometricians and trained economists (including Fed economists that measure and monitor money growth). Nonetheless, it is my view that the type of financial claims that demonstrates the “economic functionality” of “money” can vary greatly depending on evolving marketplace perceptions with respect to safety, liquidity, and the capacity to maintain nominal value.

And, importantly, I strongly argue that over the life of an inflationary boom the marketplace will come to perceive characteristics of “moneyness” in an expanding array of financial claims, and that this expanding universe of readily accepted instruments plays a defining role in perpetuating the boom. This is particularly the case when it comes to asset Bubbles and the underlying claims backed by inflated collateral values (i.e. after a protracted real estate boom, ABS and MBS today enjoy perceived moneyness qualities). Almost by definition, the final precarious boom-time speculative blow-off is financed through the frenetic expansion of dubious, yet momentarily treasured, financial claims.

With the above as background, I will attempt to clarify my view that we are at no analytical loss with the upcoming relegation of M3 to the government data scrapheap. First of all, M3 is today definitely not reflective of marketplace perceptions with respect to “moneyness.” With each boom year, the spectrum of perceived safe and liquid instruments expands. This year will see record ABS and commercial paper issuance, with the combined growth of these two categories of financial claims likely in the range of total M3 growth. M3 captures little of this imposing monetary expansion.

The monetary aggregates (“M’s”) were constructed for a bygone monetary era largely dictated by banking sector liabilities, intermediation and payment clearing. The expansion of deposits and other bank liabilities (“repos,” euro deposits, etc.) would sufficiently capture total system Credit growth, with the M’s generally correlating well with nominal economic output and indicative of general financial conditions. However, several fundamental developments have profoundly altered the monetary landscape and the capacity for the M’s to reflect relevant economic and market developments. The rise to prominence of non-bank lending mechanisms has profoundly changed the nature of financial sector liability creation and intermediation.

Market-based securities issuance is now a major aspect of monetary expansion, and the M’s are undoubtedly ill-equipped for such an environment. The unprecedented expansion of GSE obligations (debt and MBS) created several Trillion dollars of perceived safe financial sector liabilities. The enormous growth of Wall Street intermediation has spurred both a boom in securities issuance and incredible growth in (individual and institutional) account balances held throughout the (international) broker/dealer community. Technological advancement has also played a key role in expanding “moneyness.” For example, the Internet now allows households and institutions to directly purchase Treasury bills and myriad securities online, when much of these funds would have in the past been held in bank or money fund deposits (and included in the M’s). I also believe our massive Current Account Deficits and the corresponding ballooning of foreign central bank dollar holdings have impacted the relevancy of the M’s. Every day now, a couple billion dollars of Credit growth immediately flows to overseas institutions, where much of it is recycled back to financial claims (Treasuries, agencies, MBS, and ABS) not included in the monetary aggregates. If these funds were instead held domestically as savings, it is quite likely that a large percentage would be held in instruments included in the M’s.

It is also worth noting that the M’s can at times prove especially flawed indicators of Credit expansion. In periods marked by a significant augmentation of Marketplace Risk Embracement, disintermediation out of low-yielding bank and money fund deposits into riskier instruments may meaningfully distort the M’s (recall 2003’s 4th quarter). Not only would stagnant monetary aggregate growth fail to reflect system Credit expansion, it would give decidedly erroneous signals with respect to system liquidity. And I know this is unconventional thinking, but I have come to completely disassociate the M’s from system liquidity. I would argue that system liquidity is today determined by the capacity of the broader financial system (including Wall Street, hedge funds, the “repo market”, foreign bank and global central bank dollar holdings) to expand, almost irrespective of the M’s.

In summary, the M’s no longer reflect either system Credit growth or system liquidity, and are prone to give erroneous signals at critical junctures (when Marketplace Risk Embracement is modulating). I have watched repeatedly over the past few years as analysts have pounced on any slowdown in the M’s as an indicator of waning Credit growth and liquidity. This year, it was the stagnation of MZM that captured analysts’ attention, notwithstanding that this development was largely related to continued disintermediation from the money fund complex and the shift to higher-yielding term deposits; Credit growth remained on record pace, and the Bubble economy carried on. Moreover, M3 is clearly not capturing the historic expansion in the securities-financing repurchase agreement (“repo”) marketplace. While primary dealer “repo” positions have expanded $975 billion over the past two years, the M3 component bank net “repo” liability position has increased $27 billion. And while some “repo” positions are being captured in Money Funds holdings, there are enormous perceived “money” assets held in the ballooning securities financing arena outside the purview of the M’s.

If the Fed endeavored to shroud the extent of current monetary inflation, I suggest they stick with publishing M3. And it is inconceivable at this point to expect the Fed – or the economic community – to embrace a broad-based measure of monetary instruments that would include Wall Street marketable securities and “repos.” Anyway, contemporary “money” is a moving target that changes at the whim of marketplace perceptions. And while I question the premise that the Fed has much to gain by eliminating M3, this nonetheless misses the much more salient point: The Fed has lost control of our nation’s “money” and Credit creation processes. The Greenspan/Bernanke Fed can now only administer feeble attempts to remove accommodation, hoping that over time baby-steps makes some headway but without ever attempting to impede, interrupt or discipline Wall Street Monetary Processes." Link:

Posted by: bailey | November 26, 2005 at 11:45 AM

The prudent bear analysis is very good. Over time, interest in money supply waxes and wanes.
Right now M 3 is not of much value. But who knows what will happen in a few years. It may become important again, but how will we know if it is not published?

Posted by: spencer | November 26, 2005 at 04:54 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad