This procedure monitors the performance of the portoflio and it calculates the MARKET RISK.
The stocks of 5 companies traded on the Frankfurt Stock Exchange are selected:
Hugo Boss (BOSS_X), Trading System: Xetra, ISIN: DE000A1PHFF7
E.on Se (EON_X), Trading System: Xetra, ISIN: DE000ENAG999
Heidleberg Druckmaschinen (HDD_X), Trading System: Xetra, ISIN: DE0007314007
Continental (CON_X), Trading System: Xetra, ISIN: DE0005439004
Siemens (SIE_X), Trading System: Xetra ISIN: DE0007236101
The portfolio is composed by:
for a total of 19 shares.
The data are downloaded from Quandl.com, which is a platform for financial, economic and alternative data that serves investment professionals.
The data cover a time period from 02-01-2014 till the last trading day available.
The stocks and the portfolio values are daily updated and the calculations of risk measures are on time.
The portoflio value today is:
| 2018-12-21 | 2014-01-02 | Delta | % change |
|---|---|---|---|
| 1263.652 | 1569.14 | -305.488 | -21.65 |
Plot the development of PF value for the last 100 trading days: from 2018-07-31 to 2018-12-21.
| 2018-12-21 | 2018-12-20 | 2018-12-19 | 2018-12-18 | 2018-12-17 | 2018-12-14 | 2018-12-13 | 2018-12-12 | 2018-12-11 | 2018-12-10 | 2018-12-06 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Value | 1263.652 | 1263.314 | 1275.982 | 1283.25 | 1277.784 | 1287.678 | 1304.432 | 1301.924 | 1288.85 | 1269.252 | 1288.146 |
| % change | 0.030 | -1.000 | -0.570 | 0.43 | -0.770 | -1.290 | 0.190 | 1.010 | 1.53 | -1.480 | -3.330 |
| 2018-12-21 | 2018-12-20 | 2018-12-19 | 2018-12-18 | 2018-12-17 | 2018-12-14 | 2018-12-13 | 2018-12-12 | 2018-12-11 | 2018-12-10 | 2018-12-06 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| BOSS | price | 53.320 | 52.960 | 53.860 | 53.640 | 54.880 | 56.880 | 58.000 | 57.300 | 59.040 | 58.760 | 59.420 |
| % change | 0.680 | -1.690 | 0.410 | -2.290 | -3.580 | -1.950 | 1.210 | -2.990 | 0.480 | -1.120 | -1.900 | |
| EON | price | 8.864 | 8.895 | 8.965 | 8.899 | 8.970 | 9.231 | 9.220 | 9.092 | 8.933 | 8.694 | 8.821 |
| % change | -0.350 | -0.780 | 0.740 | -0.790 | -2.870 | 0.120 | 1.400 | 1.760 | 2.710 | -1.450 | -2.310 | |
| HDD | price | 1.536 | 1.526 | 1.603 | 1.603 | 1.631 | 1.694 | 1.738 | 1.745 | 1.716 | 1.656 | 1.756 |
| % change | 0.650 | -4.920 | 0.000 | -1.730 | -3.790 | -2.560 | -0.400 | 1.680 | 3.560 | -5.860 | -0.230 | |
| CON | price | 122.700 | 122.550 | 123.250 | 124.050 | 122.800 | 122.650 | 125.600 | 126.950 | 124.900 | 121.800 | 124.800 |
| % change | 0.120 | -0.570 | -0.650 | 1.010 | 0.120 | -2.380 | -1.070 | 1.630 | 2.510 | -2.430 | -3.890 | |
| SIE | price | 98.170 | 98.390 | 99.510 | 100.320 | 99.580 | 100.200 | 100.440 | 99.510 | 97.900 | 96.960 | 97.670 |
| % change | -0.220 | -1.130 | -0.810 | 0.740 | -0.620 | -0.240 | 0.930 | 1.630 | 0.960 | -0.730 | -3.350 |
| Mean | Median | St. Dev. | Kurtosis | Skewness | |
|---|---|---|---|---|---|
| HugoBoss | -0.00052 | 0.00057 | 0.01836 | 21.04156 | -2.16242 |
| EON | -0.00031 | 0.00000 | 0.01898 | 6.08201 | -0.64582 |
| Heidelberg | -0.00044 | 0.00000 | 0.02389 | 6.68010 | -0.69125 |
| Continental | -0.00020 | 0.00000 | 0.01698 | 5.22867 | -0.56570 |
| Siemens | -0.00001 | 0.00031 | 0.01362 | 2.79942 | -0.06887 |
The market risk of the portfolio is computed applying 5 methodologies:
Variance Covariance method
Historical Simulation
Monte Carlo simulation with a multivariate Normal distribution
Monte Carlo simulation with a multivariate Student-t distribution
POT (Points Over Threshold) with a Generalized Pareto Distribution
In the following table, it is possible to see the maximum loss expected (in Euro), that the portfolio can experience on 2018-12-24 with a probability at 99%.
| VaR | ES | |
|---|---|---|
| Var-cov. | 38.07 | 43.58 |
| Hist. sim. | 45.05 | 57.15 |
| MC (normal) | 37.44 | 42.76 |
| MC (Student t) | 44.92 | 61.03 |
| POT | 45.09 | 57.32 |
This section checks if the portfolio survives real market conditions.
The VaR is calculated with a Monte Carlo with multivariate t-Student distribution.
The Back Test is evaluated from 2017-07-03 to 2018-12-21.
Let’s compare the VaR and Profit/Loss realized by the portfolio in the last 10 trading days:
| 2018-12-21 | 2018-12-20 | 2018-12-19 | 2018-12-18 | 2018-12-17 | 2018-12-14 | 2018-12-13 | 2018-12-12 | 2018-12-11 | 2018-12-10 | 2018-12-06 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Value of PF in Euro | 1263.65 | 1263.31 | 1275.98 | 1283.25 | 1277.78 | 1287.68 | 1304.43 | 1301.92 | 1288.85 | 1269.25 | 1288.15 |
| Return of PF in Euro | 0.34 | -12.67 | -7.27 | 5.47 | -9.89 | -16.75 | 2.51 | 13.07 | 19.60 | -18.89 | -43.63 |
| VaR in Euro | -44.78 | -45.38 | -45.65 | -45.45 | -45.54 | -46.46 | -46.57 | -45.66 | -45.47 | -45.75 | -47.18 |
Percentage of exceptions occured, i.e. how many times the Loss experienced was higher than the VaR predicted :
## [1] "0.55 %"
Now, the Conditional Coverage Test (Christoffen 1998) is implemented, which tests that the probability of observing an exception on a given day is independent if whether or not an exception was recorded on the previous day.
The Conditional Coverage Test states that:
\(H_{0}\) : the exceptions are serially independent
## [1] "The null hypothesis of Conditional Coverage Test cannot be rejected"
If the \(H_{0}\) cannot be rejected, the model can react promptly to new information.