Page 87 - Annual Report 2022

25 Banka Kombëtare Tregtare Annual Report 2022

Banka Kombëtare Tregtare Sh.a.
Notes to the Consolidated Financial Statements for the year ended
31 December 2022 (Amounts in USD, unless otherwise stated)

4. Use of estimates and judgements (continued)
Input in data model of application of IFRS 16 requirements – (continued)

a. Base rate yield curve (continued)
The Bank uses the Nelson-Siegel-Svensson model for extrapolation purposes for USD yield curve construction, which fits an
exponential approximation of the discount rate function directly to market prices. The Bank introduced the application of the
augmented NSS (Nelson-Siegel-Svensson) model as a version that has the ability to combine different forms of graphs, allowing
in essence negative rates as well as atypical interest rate distributions, which are not captured accurately by the classic Nelson-
Siegel model.
The Bank uses the Cubic spline interpolation for EUR yield curve construction. Cubic spline interpolation is a special case of
spline type interpolation that is used very often to avoid the problem of Runge’s phenomenon. This method gives an interpolating
polynomial that is smoother and has smaller error than some other interpolating polynomials such as Lagrange polynomial and
Newton polynomial. Cubic smoothing splines fitted to univariate time series data can be used to obtain local linear forecasts. The
approach is based on a stochastic state space model which allows the use of a likelihood approach for estimating the smoothing
parameter, and which enables easy construction of prediction intervals. In essence the same mathematical mechanic is followed
by the NSS (Nelson-Siegel-Svensson) model. Whereas an interpolation typically begins with specifying a functional form either
to approximate discount function or forward rates, and then estimates the unknown parameters. The cubic spline approach,
brings more flexibility on the shape of a yield curve and is thus good for financial practitioners who are looking for small pricing
anomalies.

To construct local currency, Albanian Lek (ALL), yield curve (YC) the Bank is using the Cubic spline interpolation, as described
above. Yields of government bonds (ON-1Y) are auction results published by Ministry of Finance and Bank of Albania at the end
of each respective auction. For auctions that are not so frequent, the rate is calculated by extrapolating between rate values of
the last 2Y bond and the rate derived from the last auction of the bond in question. The issue encountered by the bank’s forecasts
on Treasury Yields is of the Runge’s phenomenon type, which is a problem of oscillation at the edges of an interval that occurs
when using polynomial interpolation with polynomials of high degree over a set of equispaced interpolation points.
b. Credit spread

For the credit spread calculations, the Bank has approached the following logic:

1) Identify the international long-term Issuer Default Rating of the financial institution (“Bank”). International long-term IDR is
given by the External Credit Rating Agency such as Moody’s, Fitch or Standard & Poor. The Bank will use only the official,
world-wide accepted, external credit rating agencies such as Fitch, Moody’s and S&P because only these 3 agencies do the
analyses world-wide, make and publish the studies on PDs, LGD’s (where credit spread will be determined as PD*LGD) etc. on
the global level. These three agencies are also the only ones allowed to be used for the purpose of relying on the expert-data
parameters for e.g. in EU (as per CRD/CRR regulation etc.).

2) If the financial institution (Bank) does not have such a rating and it is part of a Group, the lower rating of the country ceiling
for the country where Bank is located and the external agency’s international long-term Issuer Default Rating of the ultimate
parent is used. The underlying reason for this approach is that when a bank is part of a group, support is more likely.

3) If neither of these steps results in a rating, country ceiling for the country in which Bank is located is identified and at least one
notch is subtracted. The country ceiling is the best rating that an entity based in that country can receive, so this is used as
a benchmark as we tend to work with the biggest and most robust institutions. Additionally, the downward risk adjustment is
made for the sake of prudence.

That particular rating of the Bank is assigned proper probability of default rate (PD rate), which is externally calculated – expert
data given by the external credit rating agency. However, PD is just a probability. In order to approximate full credit risk, LGD is
needed. By multiplying the PD rate and LGD rate, credit loss rate is obtained, and this is the approximation of credit risk.