# More on the McCallum-Christensen rule (and something on Selgin and the IMF)

I have just printed three papers to (re)read when the rest of the family will be sleeping tonight. You might want to have a look at the same papers.

The two first are connected. It is Lastrapes’s and Selgin’s 1995 paper “Gold Price Targeting by the Fed” and McCallum’s 2006 paper “Policy-Rule Retrospective on the Greenspan Era”. Both papers are basically about how the Greenspan conducted monetary policy.

The hypothesis in the first paper is that the Greenspan Fed used gold prices as an indicator for inflationary pressures, while the other is a restatement and an empirical test of the so-called McCallum rule. The McCallum rule basically saying that the Fed is targeting nominal GDP growth at 5% by controlling the money base.

As both papers confirm their hypothesis why not combine the results from the two papers? The Fed reserve controls the money base to ensure 5% NGDP growth and use the the gold price to see whether it is on track or not. Okay, lets be a little more open-minded and lets include other asset prices and lets look at more commodity prices than just gold. Then we have rule, which I have earlier called a McCallum-Christensen (yes, yes I have a ego problem…).

The McCallum-Christensen rule can be estimated in the following form:

dB=a+b*dV+cNGDPMI

d is %-quarterly growth, B is the money base (or rather I use MZM), V is the 4-year moving average of MZM-velocity and NGDPMI is a composite index of asset prices that all are leading indicators of NGDPMI.

In my constructed NGDPMI I use the following variables: S&P500, the yield curve (10y-2y UST), the CRB index (Commodity prices) and an index for the nominal effective dollar rate. I have de-trended the variables with a four-year moving average (thats simple), but one could also use a HP-filter. I have then standardized each of the variable so they get an average of 0 and a standard deviation of 1 – and then taken the average of the four sub-indicators.

And guess what? It works really well. I can be shown that during the 1990ties the Fed moved MZM up and down to track market expectations of NGDP captured by my NGDPMI indicator. This is the time where Manley Johnson and Bob Keleher played an important role in the conduct and formulation of monetary policy in the US. As I have earlier blogged about I fundamentally think that the Johnson-Keleher view of monetary policy is closely connected to the Market Monetarist view.

The McCallum-Christensen rule also fit relatively well in the following period from 2000 to 2007/8, but it is clear that monetary policy is becoming more erratic during this period – probably due to Y2K, 911, Enron etc. Hence, there is indications that the influence of Johnson and Keleher has been faithing in that period, but overall the McCallum-Christensen rule still fits pretty well.

Then the Great Recession hits and it is here it becomes interesting. Initially the Fed reacts in accordance with the McCallum-Christensen rule, but then in 2009-2010 it becomes clear that MZM growth far too slow compared to what the MC rule is telling you. Hence, this confirms the Sumnerian hypothesis that monetary policy turned far to negative in 2009-10.

So why is it that I am not writing a Working Paper about these results? Well, I might, but I just think the result are so extremely interesting that I need to share them with you. And I need more people to get involved with the econometrics. So this is an invitation. Who out there want to write this paper with me? And we still need some more number crunching!

But for now the results are extremely promising.

Okay, on to the third paper “Reserve Accumulation and International Monetary Stability”. Its an IMF working paper. I have a theory that the sharp rise in the accumulation of FX Reserves after the outbreak of the Great Recession has prolonged the crisis…but more on that another day…

PS I promised something on Selgin and this was not really enough…but hey the guy is great and you are all cheating yourselves of great inside into monetary theory if you don’t read everything George ever wrote.

1. #### Alex Salter

/  October 19, 2011

Unfortunately I don’t know anything about time series beyond Fourier analysis and the classical decomposition, and I know even less about VAR models, otherwise I’d volunteer in a heartbeat. Good luck!