Page 424 - Special Topic Session (STS) - Volume 3

STS552 Natalia Nehrebecka
Figure 3: Estimated coefficients for Recovery Rate using Quantile
Regression and OLS along with 95% confidence intervals for economic
sectors 2

Source: Authors’ calculations

4. Discussion and Conclusion
In order to estimate the risk parameter LGD in a suitable manner must
meet a number of requirements imposed by the regulator. The main aspects
are the right approach to the default definition - consistent within all credit
risk parameters, creating a reliable reference data set, based on which the LGD
is estimated, considering all historical defaults in modelling and selecting the
right modelling method. It is necessary to verify and validate the methods of
estimated losses due to defaults and to correct any discrepancies. Validation
should pay attention to compliance with regulatory requirements as well as
the correctness of the estimated parameters and the predictive power of
models. Correct estimation of the LGD parameter affects the maintenance of
adequate capital for expected credit losses, which is a key element of the
bank’s operation.

2 Sector=2: “B” - Mining and quarrying; Sector=3: “DE” - Energy, water and waste; Sector=4:
“F” – Construction, Sector =5: “G” - Trade, Sector=6: “H” - Transportation and storage,
Sector=7: “L” - Real estate activities, Sector=8: “Others”.

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