The recent update to the MA compilation method revealed a sudden reduction in the growth rate. However this was driven by a mysterious “improvements in reporting at one institution”, which saw £85bn vanish in January 2014. I made a shadow M’ series which added this back in, but that’s not ideal.
I’ve just tried an alternative response, which is to strip MFI deposits from the measure. We can call this MAex, and here’s the series from April 1991:
If you want to see a more recent look, here it is from January 2001:
I’m continuing efforts to improve the measure.
Indeed. It is invalid to include interbank liabilities as part of the money supply so what is here called ‘MAEX’ is actually just ‘MA’.
To see why, suppose Barclays grants a GBP100 loan ex nihilo in true fractional bank fashion to someone who uses it to pay me for some tennis lessons (in practice, what we probably mean is that my trainee overdraws his account as per his standing arrangements with Barclays).
I deposit the sum in my demand account at Santander, thereby turning it into money. Santander, having no immediate outlet for it, jobs it on in the money market to Lloyds who similarly pass the parcel to RBS who finally lend it back to Barclays. Assets and liabilities are all now matched at the close of business.
The original version – MA-cum Banks – which counts ALL the liabilities regardless of their ownership, amounts to GBP400 and so grossly overstates the amount of money involved in this series.
It should be clear that there is no economic significance (as opposed to implications for financial network instability effects) of there being three interposed banks or three hundred between the originating bank asset (Barclay’s loan) and the key monetary liability (my deposit)
‘MAEX’ however only registers Santander’s original demand liability to me, a member of the non-bank world, and that is the only real money involved in this whole, otherwise incidental, daisy chain of transactions.
Ergo MAEX = MA
QED I hope!