Monday, 22 April 2013

The Reinhart-Rogoff controversy: some microeconomic evidence

The duel between pro-austerity and pro-stimulus advocates is now being fought on the relationship between economic growth and debt/GDP ratios, following the Reinhart and Rogoff finding that there was a "tipping point" around 90 per cent of debt-to-GDP ratio when the correlation between debt and economic growth would become negative.

Their finding was questioned by Herndon-Ash-Pollin who estimated that the strength of the negative relationship was actually much stronger at low ratios of debt-to-GDP. Recently Dube, using the same set of data, estimated that current period debt-to-GDP is a pretty poor predictor of future GDP growth at debt-to-GDP ratios of 30 or greater but it does a great job predicting past growth which he claims is a tell-tale sign of reverse causality. That is, recession leads to increased spending and greater government borrowing not the other way around. Krugman joined the debate on Dube´s side but cautioned about claiming any causal relationship, or, if it existed, it would be pretty slim.

The debate is not over and, most likely, macroeconomists will continue sabre-rattling with lags and correlation studies to discuss the direction of causality and the location of the “tipping point”. In my view, it is unlikely that they will arrive at any unequivocal conclusion at the macroeconomic level for three main reasons. First, if we assumed a closed economy with a single firm we would end up with an accounting identity between total assets and their financing that would prevent any causality conclusions. Second, since leverage amplifies both gains and losses, any relationship must be very sensitive to the business cycles. Finally, the microeconomics of debt financing is too complex to build a one-way macroeconomic theory.

Yet at the firm level it must be easier to find if there is such a tipping point. There are basically three ways in which we can use debt-financing – to finance consumption, failed investments (including gambling) and profitable investments. In the long run only the third use is sustainable, but in the short run the three types of spending have a positive multiplier effect on economic growth. Elsewhere, I have shown why at the corporate level debt financing has simultaneously contracting and expansionary effects on investment, with the positive generally offsetting the negative effect depending on lender’s mark-ups and borrowing limits.

The corporate level is the right place to find out if and where there is any tipping point in the spectrum of leverage, because if there is one it should be close to the maximum debt-capacity financiers impose on the basis of several debt coverage ratios. Moreover, we may extrapolate those results to the macro level under the following, not very extravagant, assumptions: a) the shares of labour and capital in total income are relatively stable; b) an ever increasing number of companies do not pay dividends so that the growth of equity is a good proxy for economic growth; and c) listed companies give a good representation of the entire business sector.

Since profitable companies should use leverage to increase the return to their shareholders, the correct way to verify if they benefit from increased leverage is to check if the elasticity of equity in relation to debt is greater than one or at least positive. Since at SADIF Investment Analytics we cover more than 20,000 stocks worldwide we quickly pulled the quarterly growth rates of equity and debt for the last four years which allow us to gauge the relationship between equity growth and debt-financing.

We used data from countries that in the popular imagination epitomise the three types of use for debt financing. Americans are often seen as reckless shopaholics pursuing consumption-led growth policies, Euromeds (Portuguese, Spaniards, Italians, Slovenes and Greeks) are generally perceived as castle-in-the-air investors in loss-making projects in transportation and alternative energies generously financed by the EU/EIB and Germans are traditionally depicted as successful thrifty mercantilists. The distribution of firms and the median equity elasticity in each quadrant of the equity and debt growth space is given below.

The median equity elasticities highlighted in the table seem to validate the popular view on the use of debt financing since the Germans have the highest value and the Euromeds the lowest. The only discordant note is that the percentage of firms with a positive elasticity is much higher in spendthrift America than in thrifty Germany which weakens any macroeconomic extrapolation.

To test the location of a possible tipping point I plotted the equity elasticity against the extra debt capacity measured as the spare level of debt capacity as a percentage of total outstanding debt, so that we can measure the leverage spectrum from left to right in the chart below for the USA.
The expectation is that elasticities rise as firms deleverage or leverage towards their maximum debt capacity (0% extra-capacity). The maximum of the quadratic functions fitted are indeed close to zero (i. e. -1.7% in the US and 7.3% in Germany). However the function is meaningless for the Euromeds and the coefficient of determination is too low for the other two countries.

So, in conclusion, this microeconomic evidence suggests that it is not possible to settle the debate on the correlation between growth and debt without allowing for the business cycle and the microeconomic complexities of debt financing.


  1. While I was looking for some stuff I visited this site and found great stuff in it. The design of the site is impressive and the content presented is very easy to understand. I am very much thankful for this site.

  2. I think the Reinhart-Rogoff replication scandal has not only consequences for how we think about debt. It might well change the way we think about replication and data sharing in the social sciences.