Top-down Investing Toolkit: Growth Accounting

29 February 2016


A couple weeks ago, I saw Jamaica Finance Minister Peter Philips on Bloomberg Markets, discussing the economy in the context of the stellar 2015 performance of the Jamaica Stock Exchange (Main Index). You may read about this outperformance and some boons for the Jamaican economy here. Insofar as economic growth is used for investment decision-making, this news made me curious about how one of the concepts applies to a small island economy.

Growth accounting is one way to examine the performance of an economy and investment prospects by extension. Based on the textbook production function, the growth accounting equation is derived: the growth rate in output is equal to the growth rates of factor inputs given how much each input matters to growth. In practice, GDP growth is regressed on weighted growth for the variables labour and capital. In simple terms, GDP growth is ‚Äėexplained‚Äô by the weighted growth in labour, capital and every other possible variable that was not specified. That residual is called the growth in total factor productivity or TFP. Results from these models can allow for comparisons of whether labour or capital has contributed (or perhaps will contribute) more to a given year‚Äôs growth; however, the ‚Äúeverything else‚ÄĚ variable may often be the largest contributor, somewhat unhelpfully. The academic literature then is replete with papers that specify other variables, with different methods for measuring those variables.

The Conference Board Productivity Brief 2015 applies the concept to over 120 countries. Their explanatory variables here are the quantity of labour; the quality of labour; information, computer and telecommunications (ICT) capital services; and non-ICT capital services (non-residential construction, machinery and transport equipment), each weighted by their share of national income. Jamaica‚Äôs 0.5% growth in 2014, for example, was attributable to labour quantity more than it was by ICT Capital‚ÄĒalthough growth in the latter was far higher that year. This is influenced not only by the proportions of the inputs that create output in a given year but also by the elasticities for each variable that are estimated using the data.

160229 Blog Photo

The Conference Board‚Äôs methodology and data sources get a little involved. As the database sources are listed only generally, I was unable to verify the data for small economies or why ICT data is missing for Trinidad & Tobago, for example. I was also curious about comparability given that ‚Äúpersons employed‚ÄĚ is used where the number of hours worked is unavailable (as is for many of the small-island economies in their database). As I scoured the literature, I found myself doubling back to the websites for the Statistical Institute of Jamaica and the Central Statistical Office of Trinidad & Tobago to see what data was available to me among the variables that were unlikely to be collected by extra-regional databases. (Detractors of the CSOTT over its dataset lags and gaps have the National Statistical Institute of Trinidad & Tobago to look forward to.)

Data concerns aside, TCB’s methodology does appear sound to me, although I am interested in further specification such as the variables resource quality and public infrastructure that I read about elsewhere.  There is also the plethora of statistical concerns for time series data before any serious forecasting can be done for growth. As of October 2015, the IMF forecasts 2.1% and 1.4% real GDP growth for 2016 for Jamaica and T&T, respectively. Could these be reliable signals of how the average listed company performs next year?

As far as forecasts and leading macroeconomic indicators of growth go, I have wondered whether a macroeconomic model such as this or otherwise would have predicted the Jamaica Stock Exchange constituents’ 2015 earnings or at least produced enough insight into the impact of certain variables such that news could be acted upon. How do you systematically overturn the right rocks?

Jonathan Wilson
Risk Analyst

The views expressed are those of the author. As such, this post should not be construed as investment advice, nor does the opinion expressed necessarily reflect the views of Firstline Securities.

Comments are closed.