Money and Output: The Missing Links: A Time Series Analysis

Abstract

There is an immense body of literature on the relationship of money and output, with result-integrity depending on the aggregates used. This paper proposes an alternative two-step approach, based on the reality that: (1) money creation is just the outcome of new bank loans extended, and (2) the (net) new demand for bank loans, when satisfied by the banking sector, parallels the growth in nominal GDP in the long-term – because, mainly, economic units borrow to consume (C) or invest (I), the major contributors to nominal GDP. The new deposits created are M1 deposits, but they can swiftly shift to M2-M1 or M3-M2, reflecting portfolio – or “demand for money” – decisions. Therefore, mining for an association between M1 and real GDP growth is not fruitful. The direct link between the monetary and real sectors lies is the association of new net bank loans to the private sector (the outcome of which is money creation) with nominal GDP growth: the R2 of the South African data (no massaging; quarterly; 1965-2012) is 0.99. The relationship in the short-term is less well correlated. The long- and short-term associations must be a fully considered Step 1, because it contains the seeds of the answer to Step 2: the translation of nominal GDP growth into real GDP growth and inflation, that is, the trade-off between these related aggregates. This article offers a time series analysis of Step 1, based on South African data.