Real or nominal shock – which does more to destabilize developing economies? The case of money velocity in Kazakhstan

What causes instability in a national economy?  To find out, economists often take two approaches.  One is to specify a model of structured vector autoregression -- an ad hoc model constrained by a few conditions derived from theory.  Its hybrid nature leaves it open to question – why specify one condition from theory but not others?  The other approach is a Monte Carlo simulation, which is inherently unrealistic.  This paper proposes a third approach:  Begin with a tautology – the quantity equation of exchange – and decompose it into the sources of instability in spending.  This approach seems structured and realistic.

The paper develops a gauge of volatility in money velocity, which destabilizes spending and output, generating business cycles.  The determinants of this volatility arise from money supply, output, and the price level.  The paper’s algorithm allows covariances among these variables.  An application to a fast-growing transition economy, Kazakhstan, finds that at the margin, price shocks affect the volatility of spending more than do real shocks, by several orders of magnitude, despite real-business-cycle claims.  An oil exporter, Kazakhstan may be vulnerable to the gyrating price of crude oil.  The paper also finds that price shocks are more destabilizing than money-supply shocks, despite monetarist claims. 

The paper speculates that transition economies like Kazakhstan, which are based on resource exports, may not be well-explained by either the monetarist model, the real-business-cycle model, or the short-run Keynesian model (which holds the price level constant because it is irrelevant). As a secondary goal, the paper illustrates how the algorithm can help estimate the forecast error in simulations. 

Keywords:  real shocks, monetary shocks, monetary policy, simulations, forecasting in transitional economies