Page 19 - Invited Paper Session (IPS) - Volume 2

IPS178 Sana Antoine S. J.
 USD/EUR exchange rate (FED)
The domestic variables are the following:
 M3: money supply (BDL)
 WAIR: weighted average interest rate (BDL)
A dummy has been included for the year 2012 to account for the “jump”
in CPI due to the change in its housing component. In addition to this
“artificial” change, the year 2012 witnessed an increase of salaries in the public
sector as well as a massive arrival of Syrian refugees which translated into
higher demand and prices in the housing sector. As we can notice from figure
3, the variations in CPI closely follow the variations of the Brent price and Food
Index.

Figure 3: CPI, Brent and Food index Sources: CAS, US EIA, WB Commodity

Price Data

OLS estimation
Based on unit root tests, we decide to consider the first log-differenced
series for the OLS estimation. Furthermore, due to the highly noisy nature of
the month-on-month log-differenced series, we perform our estimation using
year-on-year logdifferences. Finally, the agricultural index and the Brent price
are strongly correlated (0.75) which lead us to keep the Brent time series for
the final estimation.
The following OLS regression has been performed:

Δ(log (CPIheadline))
= β0 + β1 Δ(log (M3)) + β2Δ(log(WAIR)) + β3Δ(log(BRENT)) +
β4Δ(log(EXR)) + dummy + ϵ
In addition to the headline CPI, we performed the same regression on a
measure of the “core CPI” which excludes Food and Energy items. Such

6 | I S I W S C 2 0 1 9