International Capital and Growth Nexus in Turkey

In the figure above, I report the cumulative distribution functions of GDP growth in Turkey and the Capital Inflows to Turkey. So similar, no?

In a recent article, I and Ahmet Yilmaz of Marmara University have shown that the GDP growth in Turkey is highly dependent on the international capital. "OK", you may say, "Highly"... But what does "Highly" mean in this context?

Well, we first gathered detailed data for the capital inflows to Turkey from 1991 (this is the first date we could reach detailed quarterly data. Besides Turkey completely removed the capital account controls in August 1989) to Q3 2012. In the capital inflows, we included net errors and omission figures as well since net errors and omission numbers can be very (I mean very) high for countries like Turkey where substantially large amount of unregistered cross border trading activity exists.

And we formally compared the capital inflows data with the quarterly GDP growth data (YoY). The result is capital inflows Granger cause growth in Turkey but reverse causation does not hold! Plus, 44% of the variation in Growth can be attributable to the shocks to the capital inflows to Turkey!!

In a second paper, I am currently writing with Hasan Selcuk of Marmara University again, we have found that there exist nonlinear volatility interactions between the capital inflows and growth as well. Volatility is spilling over from capital inflows to growth in Turkey according to the Dynamic Conditional Correlation GARCH model we adopted in the paper with Hasan Selcuk.

In sum, what we have at hand is as follows:
- GDP growth in Turkey is dependent on capital inflows from abroad. Capital inflows cause growth but not the other way round.
- And there are volatility interactions between the two. Volatility spills over from capital inflows to growth. But not the other way round.

Can Volatility Predict Stock Returns?

I asked the question in the title in a paper that I deposited at the SSRN elibrary today.

The reverse question, i.e. whether the current stock returns are good predictors of future volatility, has been asked and studied many times in the literature. 

However, the causality from current volatility to future stock returns did not attract the academic attention at the same rate (that is not to say this way of causation has never been studied. I am only saying this way of causality is much less questioned.)

This is interesting because, although the academic community does not pay too much attention to whether volatility predicts stock returns or not, investment community (aka market professionals) seem to believe that the VIX index (a volatility index) is a gauge for future movements of S&P500 (a stock exchange index). 

Well, long story short what I did in this paper is I constructed a near-VAR system which allowed me to model the dynamic interdependencies between the variables (variables are the changes in the VIX implied volatility index - that is the volatility - and S&P500 returns - that is the stock returns).

Then I estimated the model and found that volatility (at least the implied volatility measure I used) could not predict the relevant stock exchange returns. Yikes!

See this paper on SSRN: 
http://ssrn.com/abstract=2200024