WEBVTT 00:00.950 --> 00:08.030 And this video we're creating another variant of the S&P 500 return triangle and I call this the return 00:08.030 --> 00:12.440 radar because it pretty much looks like a weather radar. 00:12.440 --> 00:18.800 The only difference to the return triangles that the return rate US based on weekly returns instead 00:18.800 --> 00:20.530 of annual returns. 00:20.600 --> 00:27.050 And that means that we have a higher resolution a definition here and that might give us some more insight 00:27.650 --> 00:31.230 and the code to pretty much looks the same. 00:31.250 --> 00:39.140 So I will only highlight the differences and first of all we import the necessary libraries here and 00:39.140 --> 00:40.870 we also import our data 00:44.670 --> 00:49.050 so here we are for the S&P 500 price data. 00:49.140 --> 00:51.450 And now comes the very first difference. 00:51.750 --> 00:55.440 So we read sample the S&P 500 price data. 00:55.740 --> 01:02.520 And before we read sample to emulate frequency and here we read sample to weekly frequency and we take 01:02.520 --> 01:11.690 the very last trading price of each week so that's the weekly data frame and then in the next step we 01:11.690 --> 01:13.500 calculate weekly or lock return. 01:13.510 --> 01:15.500 So that's what we are doing here. 01:15.500 --> 01:21.530 And we actually annualize the weekly returns by multiplying with fifty two. 01:22.250 --> 01:25.510 And by doing so we're creating a new column return. 01:25.850 --> 01:28.670 And we also drop the very first row. 01:29.000 --> 01:30.840 So let's have a look. 01:31.070 --> 01:36.440 So here we have the weekly closing prices and the weekly returns the annualized 01:39.820 --> 01:45.430 and in total our period includes one thousand five hundred sixty five weeks. 01:45.610 --> 01:52.930 And uh that means uh that in our next step we are creating one thousand five hundred sixty five new 01:52.930 --> 01:57.280 columns and actually each column stands for a rolling statistic. 01:57.640 --> 02:04.630 So we have for the rolling mean return for a period of one week two weeks three weeks until one thousand 02:04.630 --> 02:07.180 five hundred sixty five weeks. 02:07.330 --> 02:09.450 And therefore we created our list. 02:09.460 --> 02:23.920 Windows so here we have a list with integers from 1565 till 1 and then finally we creating 1565 new 02:23.920 --> 02:31.510 columns with the for loop here and the column labels should be 1 week to week and so on and again here 02:31.540 --> 02:36.620 we calculate the rolling mean return. 02:36.710 --> 02:37.770 Let's have a look here. 02:37.790 --> 02:39.530 So this might take a while. 02:40.040 --> 02:48.080 So here we have our weekly data frame with um here on the right hand side the column 1 week and on the 02:48.080 --> 02:54.290 left hand side of the column one thousand five hundred sixty five weeks and still we have for the columns 02:54.290 --> 03:01.610 close and return our data frame and we can simply drop them and we save the resulting data frame and 03:01.610 --> 03:02.990 the variable triangular 03:05.960 --> 03:12.530 and finally we can create our plot slow here our seaborne heat map and let's have a look here 03:19.190 --> 03:26.240 so that's our return triangle with a high resolution or high definition and that pretty much looks like 03:26.270 --> 03:27.540 a weather radar. 03:27.980 --> 03:37.730 So here we can see the huge Harry Kane in the year 2000 and 2001 and also the huge Harry Kane in 2 8 03:37.730 --> 03:43.780 2 9 and the financial crisis and yes actually the combination of both hurricanes. 03:43.790 --> 03:51.530 So if your investment experienced both hurricanes then still you are in the negative numbers here and 03:51.530 --> 03:53.430 depending on the investment period. 03:53.450 --> 04:00.140 So we can see here with the short periods there are many isolated thunderstorms. 04:00.140 --> 04:06.980 So this is a pretty random year if your investment period is uh let's say one year or one and a half 04:06.980 --> 04:14.390 years then that's pretty much gambling if you end up with a gain or loss and by increasing the investment 04:14.390 --> 04:20.300 period the returns are getting a less volatile and better predictable actually. 04:20.300 --> 04:26.270 So for example here you are pretty much in the double digit area and here you are still influenced by 04:26.270 --> 04:32.250 the hurricanes of the year 2000 and 2008 2009. 04:32.330 --> 04:38.870 So that's a pretty informative year and with this we are finished with the return triangles and radars 04:39.470 --> 04:42.110 and I hope to see you also in the next video by.