Exchange to Change January 2017 | Page 15

RESEARCH AT IOB
15

Interview with Dennis

Essers on the Synthetic Control Method

E2C : Can you briefly introduce your current research project ? What are the main research questions you want to answer ?
My current research project is an empirical evaluation of the Flexible Credit Line ( FCL ) 1 , the IMF ’ s prime precautionary lending instrument which so far has been taken up by only three emerging
1
The Flexible Credit Line or FCL is a precautionary instrument of the IMF that provides large amounts of resources to countries with strong macroeconomic fundamentals and a solid policy track record . For more on this see https :// www . imf . org / external / np / exr / facts / pdf / fcl . pdf . market economies : Mexico , Colombia and Poland . Together with my co-author Stefaan Ide ( of the National Bank of Belgium ) I try to answer two questions . First , which factors explain the participation of these three countries in FCL arrangements ? And second , to what extent have such arrangements delivered on their promise of boosting market confidence in their users ?
E2C : In this research , you make use of a quite novel method in quantitative impact evaluation , namely the Synthetic Control Method ( SCM )...
Yes , I use the Synthetic Control Method ( SCM ) to answer the second question . More specifically , I employ it to evaluate the effects that the FCL arrangements have had on international ( EMBI 2 ) bond spreads and capital inflows in Mexico , Colombia and Poland .
2
JP Morgan ’ s Emerging Markets Bond Index or EMBI tracks dollar-denominated sovereign bonds issued by a selection of emerging market countries .
E2C : Can you tell us what the main intuition / idea is behind this SCM ?
The method measures the impact of an intervention ( here : the FCL ) as the difference between post-intervention outcomes ( here : bond spreads or capital flows ) in the treated unit ( here : Mexico , Colombia or Poland ) and those same outcomes in a ‘ synthetic ’ control group . This control group is constructed as a weighted average of untreated units ( here : other emerging market economies that are governed by structural processes similar to those in Mexico , Colombia and Poland ). The weights attached to the different control units are chosen so that the characteristics of the synthetic control ( here : bond spreads or capital flows and their determinants ) over the pre-intervention period
Exchange to change January 2017