Sunday, March 13, 2011

Is The ECB Actually Targeting the Monetary Base?

Look at the figure below.  It shows the monetary base for the Eurozone and the data comes directly from the ECB.  The monetary base follows a striking trend that begins in 2002 and continues to the present.  Even the financial and Eurozone crises create only temporary deviations from this trend.   This trend is so straight it creates the appearance that the ECB is actually targeting the monetary base rather than following its two pillar strategy. 

This upward trend not only creates the appearance of a monetary base target, but it also turns out to be very important to the trend growth of nominal GDP in the Eurozone.  To see this, note that the monetary base, B, times the money multiplier, m, equals the money supply, M (i.e.  Bm = M).  In turn, the money supply times velocity, V, equals nominal spending or nominal GDP, PY (i.e. MV=PY). Putting this all together, we get the following identity:

BmV = PY.
Now let's unpack the first two components of this identity, Bm, that make up the money supply.  Using the M3 money supply to solve for the money multiplier, (i.e. m = M/B), the following figure shows what has been the main determinant of growth in the money supply: 

Other than a brief run up in late 2001-early 2002, the money multiplier had been relatively flat prior to the financial crisis.  Thus, most of the M3 money supply growth has come from increases associated with the trend growth in the monetary base.  After the financial crisis, the monetary base seems to adjust to the shifts in the money multiplier as an offset.  

The figure below shows M3 along with velocity, the third component of the above identity.  Here we see a sustained rise in M3 being offset by a sustained fall in velocity up till the financial crisis.  This drop in velocity seems strange given that it occurs during the housing boom. 

This and the previous figure reveal that the remarkable post-2002 trend in the monetary base was needed to offset the effects of a flat money multiplier and falling velocity on nominal GDP in the Eurozone. The figure below reveals how important the monetary base growth was to nominal GDP  growth during this time.  The figure shows actual nominal GDP and a counterfactual version, where the monetary base is grown at its 1999-2001 average monthly growth rate from 2002 on rather than  at its actual growth rate. This alternative monetary base series is then multiplied times the actual money multiplier and velocity (i.e. BmV) to get the counterfactual nominal GDP.  Here are the results:

So it seems the ECB was increasing the monetary base fast enough to keep nominal GDP growing at a stable pace.  While that can explain the upward trend in the monetary base, it does not explain (1) why the trend was so straight, (3) why the money multiplier was relatively flat, and (3) why velocity was falling.  Could it be the ECB was secretly targeting the monetary base via some kind of Bennet McCallum-type nominal GDP rule?


  1. Sure the MB pattern is linear in an arithmentic sense but not logarithmically. It's probably just a spurious pattern that is the aritifact of something else.

    The real test is probably which is more closely correlated to an exponential growth pattern, M3 or NGDP.

  2. If you see a straight line on a map, you know it's not natural. Someone made it. If you see a straight line in economics, you know someone made it straight. That line for the money base is too straight to be natural. The ECB must have made it straight. But why?

  3. Nick,
    I repectfully disagree. Straight lines depend on the metric. Why would anyone (especially the ECB) be mapping MB on an arithmetic plain?

  4. Mark:

    Nick is right. Anytime you see a straight line--fixed exchange rate, targeted interest rate--it is because of a conscious policy choice. Logs are only used in policy making to make it easier to think about issues, not to actually implement decisions.

  5. David,
    Don't be ridiculous (again with all due respect). Some patterns are always spurious. (Nothing comes to mind at the moment but if you stop and think about this it is of course true.)

    If one is targeting a rate of change and not an arithmetic rate of change one would see a line in log terms. Anyone who is targeting exponential growth should be looking for flat lines in log terms.

    A flat line in arithmetic terms in economics is almost always the result of something else.

  6. Mark:

    There are no naturally occurring straight lines in macro time series. Name one. Almost every (non-interest rate) macro time series has an exponential trend in level form. This is true for monetary aggregates as well. The ECB monetary base has no exponential trend. It is straight line. It is unique in this regard and not natural. It would require more than chance to create such a systematically straight line.

  7. David,
    You wrote:
    "There are no naturally occurring straight lines in macro time series. Name one."

    Don't be silly. One can always find spurious straight line correlations provided one is willing to limit the timeframe. I could absolutely bury you in examples if I didn't have better things to do with my time.

    My point is that the ECB is almost certainly not targeting the MB in an arithmetic fashion. (I'll bet you dollars to doughnuts.) I would look for a more rational explanation.

  8. Mark:

    All I said is show me a macro time series that is an almost perfectly straight line. I was vague on time, but obviously I meant for as long as the ECB monetary base. So show me a macro time series naturally occurring that is straight for about a decade. You can't--they don't exist. And this has nothing to do with spurious correlations. This is simply a univariate descriptive statistics exercise.

    To be clear, all I said is that this straight line makes it look like a MB target. There may be another answer, but it will have to include some explicit policy choices. This straight line didn't just occur based on random interactions of supply and demand for the monetary base.

  9. How one would account for countries joining the eurozone after 2001, like:

    Slovenia (2007)
    Cyprus (2008)
    Malta (2008)
    Slovakia (2009)
    Estonia (2011)

    So the extra addition of these countries should automatically increase the monetary base, so if you plot a graph of monetary base per country, it wouldn't look so straight.

    Or is that graph already adjusted for it?

  10. Mark: an arithmetic straight line strikes me as even weirder than a logarithmic straight line, and even less likely to occur "naturally".

    And when you add in Anonymous' observation, it looks weirder still.

    We see the hand of God here! (OK, the hand of man or woman).

  11. With the current Japanese tragedy heightening risk aversion in the markets, there would be little chance that ECB will raise rates in the short run. Spain the least exposed to the debt problems in the PIGS category. At govt yield of 4.2%, they seem to be able to make it out of the crisis in whole.