4.2. Real data applications
For the real data application of the Heston model, we use 253 daily interest returns between 02.01.2012 and 31.12.2012
(see Fig. 2) and the initial stock price as 51,340.96 TRY from BIST-100 on 02.01.2012 (see Fig. 6). We obtain an approximate
initial volatility level (on 02.01.2012) as an average of the last 5 years’ volatilities by using the data in Table 2.
At this step we perform analyses for two different initial volatilities with the corresponding long run volatilities. The first
analysis is done for the annual volatility whose initial volatility level on 02.01.2012 is obtained from the average of the annual
volatilities between 2007 and 2011 in the 5th column of Table 2. The average of the daily volatilities between 2007 and 2011
in the 7th column of Table 2 is used to approximate the initial daily volatility level on 02.01.2012 for the second analysis.
We use approximate long run volatility levels as 0.58 for the annual volatility and 228 × 10−5 for the daily volatility as
in Table 2 for 2012. The corresponding approximate long run variance levels between 02.01.2012 and 31.12.2012 are evaluated
by taking square of these volatility values (i.e. long run annual variance is approximately (0.58)2 ≈ 0.33 & long run
daily variance is approximately (228 × 10−5)2 ≈ 52 × 10−7). Then, we perform simulations for these values by using the
parameters in Table 1 and obtain 3-dimensional stock price expectation matrix M (M ∈ C1000×101×253). Now, for example,
let us select the length of time sub-intervals as 0.01 year (i.e. approximately 2.53 trading days) and the number of time
sub-interval as 3. Afterwards, we use the IMN in the 2-norm for M to analyze and quantify price impression approximately
and obtain the graphs in Figs. 3–5. These graphs show almost parallel behaviors to the graph of BIST-100 index between
02.01.2012 and 31.12.2012 which is shown in Fig. 6.