Turkish CA deficit and growth
As I mention in this week’s Hurriyet Daily News column, IMF’s latest Turkey forecasts were criticized for having let the current account deficit stick around 10 percent even with sharply falling growth.
In the column, I noted that there isn’t simple relationship between Turkey’ s current account deficit and growth, giving the 2006-2008 episode as an example. That was kind of cheating, as oil prices were an important factor that period:
Of course, the fact that the non-energy deficit did not adjust by a lot validates my claim as well, but a more complete explanation is in order. I’ll give my own simple explanation first, and then refer you to a more complete explanation, which explains my one-sentencer fully:
Simply put, given the size of imports compared to exports, it will take time for the current account deficit to fall unless Turkey has chronic low growth (or a crisis). After this executive summary, I leave you to the excellent explanation by Murat Ucer of Istanbul Analytics – GlobalSource Partners – Turkey Data Monitor, who incidentally took on the same issue in his latest Weekly, which came out a couple of hours after I submitted my column:
First, the `arithmetic of adjustment` is not on our side. Once you are up at around a $75-$85 billion figure, to even stabilize it in nominal terms, exports (of goods, services, income and transfers) have to grow some 40%-50% faster than imports, which is not easy. Second, assuming that nominal exchange rate moves in line with the inflation differential between Turkey and its trading partners, which is what‘s typically assumed at the IMF, the IMF desk economist has no choice but to “force the adjustment”, i.e. stabilize CAD at these already elevated levels of around 10% of GDP, through lower growth.
The first point is simply my point on the executive summary. But Murat, being an ex-IMF economist himself, goes to the details behind the Fund’s current account projections. BTW, all these are also discussed in the Fund’s Financial Programming book on the Turkish economy- the data are obviously outdated, but the methodologies are not. Anyway, Murat goes on to expand on his second point:
Historical import elasticities of growth as well as of domestic demand in Turkey are provided below (see graph). These are crude `unconditional` elasticities that do not take into account the exchange rate effect. But in a scenario where the real exchange rate does not move much, an import elasticity of around 2-2.5, looks to be a plausible number to work with. This yields some 5%-6% or so import volume growth as a starting point; now add on top of this, another 5% or so change in import prices, which is typically exogenously given to the IMF economist, we then have import growth at over 10% in nominal U.S. dollar terms. This means that exports will have to grow in the 15%-20% range just to stabilize or slightly reduce the CAD. Also note, not much help will be coming from the denominator (i.e. nominal GDP in dollar terms), because it won‘t change much given the low growth (2.5%) and stable real exchange rate assumptions in the IMF scenario.
In the bigger scheme of things, the real challenge here is that fixing the CAD problem requires more than a few quarters of slow growth. Put differently, we have a level problem with the CAD in the sense that it has ratcheted up to around 10% of GDP, and it is likely to stay there even in a sub-par growth scenario. Basically, the latest global liquidity cycle that began early last decade has moved Turkey to a higher CAD level, as a percent of GDP, every year till 2007. The global crisis caused an aberration from this, but since then, we adjusted, overstretched and got back to the pre-crisis CAD trajectory of rising CADs. While the exact figures may vary amongst forecasters, this is the critical challenge of Turkish macroeconomics, which is yet to be comprehended fully.
I went ahead and reproduced Murat’s exercise of elasticities, which I call “knockoff elaticities”, for the simple reason that that they don’t take into account the exchange rate effect and “other stuff”, i.e. are not produced by a proper elasticity equation:
And finally, the see Murat’s point on the evolution of the current account deficit, have a look at the second graph in my column. So what needs to be done? Murat goes on to list his short and medium-term policy recommendations:
So unless more is done in terms of: 1) implementing very prudent policies for a considerable period of time, led by a tighter fiscal policy (like targeting surpluses!); and 2) stepping up structural reforms geared toward expanding Turkey‘s tradable sector, this structural problem will not disappear simply because growth slows for a few quarters.
I would add to the structural reform agenda policies geared towards increasing the Turkish savings rate, as the current account problem is in essence a savings-investment gap.
Comments are closed.