COVID-19 short-term forecasts Deaths 2020-04-25


Disclaimer

  • Forecasts produced by Jennie Castle, Jurgen Doornik, and David Hendry, researchers at the University of Oxford. These are our forecasts, and should not be considered official forecasts from, or endorsed by, any of: University of Oxford, Oxford Martin School, Nuffield College, or Magdalen College.
  • These forecasts are short term time-series extrapolations of the data. They are not based on epidemiological modelling or simulations. The documentation that is provided is still in progress and not peer reviewed. All forecasts are uncertain: their success can only be determined afterwards. Many mitigation strategies are in place, which, if successful, invalidate our forecasts. An explanation of our methods is provided below.

Recent changes

[2020-03-24] Our forecasts are starting to overestimate in some cases. This was always expected to happen when the increase starts to slow down. Scenario forecasts that are based on what happened in China earlier this year, but only for Italy and Spain sofar.
[2020-03-26] Scenario forecasts that are based on what happened in China earlier this year, only for Italy.
[2020-03-31] Scenario forecasts, based on what happened in China earlier this year, are presented for several countries (line marked with x). Created more plausible 90% confidence bands (dotted line in same colour).
[2020-04-02] Now including more US States, based on New York Times data. And the world.
[2020-04-06] Added a post hoc estimate of the peak number of cases. This needs at least three confirmed observations (four for deaths) after the event. It is based on the averaged smooth trend, and can change later or be a local peak. It is marked with a vertical line with the date label, or a date with left arrow in the bottom left corner of the graph. This is backported to 2020-04-04.
[2020-04-08] Minor correction to peak estimates. Added table with scenario forecasts.
[2020-04-09] Added table with estimated peak dates (if happened) and dates to and since the peak. Note that this can be a local peak, and subsequent re-acceleration (or data revisions) can result in a new peak later.
[2020-04-10] Updated documentation with better description of short-term estimates and peak determination.
[2020-04-16] Added scenario forecasts to all graphs now. This would now be the preferred forecast for most.
This is the first time with a peak in confirmed UK cases (also for deaths, but this is uncertain because it is at the same date).
[2020-04-17] Bird and Nielsen look into nowcasting death counts in England.
[2020-04-24] A summary of our work on short-term COVID-19 forecasting appeared as a voxeu.

Further information

  • We believe these forecasts fill a useful gap in the short run. They give an indication of what is likely to happen in the next few days, removing some aspect of surprise. Moreover, a noticeable drop in comparison to the extrapolations could be an indication that the implemented policies are having some impact. It is difficult to understand exponential growth. We hope that these forecasts may help to convince viewers to adhere to the policies implemented by their respective governments, and keep all arguments factual and measured.
  • We use the data repository for the 2019 Novel Coronavirus Visual Dashboard operated by the Johns Hopkins University Center for Systems Science and Engineering. This is updated daily, but we tend to update our forecasts only every other day.
    US state data as of 2020-03-28 is courtesy of the New York Times.
  • We can only provide forecasts of what is measured. If confirmed cases are an underestimate of actual cases, then our forecasts will also be underestimates. No other epidemiological data is used. Data definition and collection differs between countries and may change over time.
  • We will update the methodology as we learn what is happening in the next few days or weeks. Once the number of cases levels off, there is no need to provide these forecasts anymore.
  • Countries where the counts are very low or stable have been omitted.
  • The graphs have dates on the horizontal axis (yyyy-mm-dd) and cumulative counts on the vertical axis. They show
    1. bold dark grey line (with circles): observed counts (Johns Hopkins CSSE);
    2. many light grey lines (with open circles): forecasts using different model settings and starting up to four periods back;
    3. red line (with open circles): single forecasts path using default model settings;
    4. black line (with crosses): average of all forecasts, recentered on the last observation;
    5. thin green lines: some indication of uncertainty around the red forecasts, but we do not know how reliable that is.
    Both the red line forecasts and the black lines are also given in the tables above. These forecasts differ, we are currently inclined to use the average forecasts.
  • The forecasts are constructed as follows:
    1. An overall `trend' is extracted by taking a window of the data at a time. In each window we draw `straight lines' which are selected using an automatic econometric procedure (`machine learning'). All straight lines are collected and averaged, giving the trend.
    2. Forecasts are made using the estimated trend, but we note that this must be done carefully, because simply extrapolating the flexible insample trend would lead to wildly fluctuating forecast. We use the `Cardt' method, which has been found to work well in other settings.
    3. Residuals from the trend are also forecast, and combined with trend forecasts into an overall forecast.
  • Scenario forecasts are constructed very differently: smooth versions of the Chinese experience are matched at different lag lengths with the path of each country. This probably works best from the peak, or the slowdown just before (but we include it for the UK nonetheless).
  • The forecast evaluation shows past forecasts, together with the outcomes (in the grey line with circles).
  • EU-BS is Estonia, Latvia, and Lithuania together.
  • This paper describes the methodology and gives further references. Also available as Nuffield Economics Discussion Paper 2020-W06. Still preliminary is the documentation of the medium term forecasts.

Deaths count average forecast Europe (bold black line in graphs) 2020-04-26 to 2020-05-02

DateUKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCH
2020-04-25 20319 96522 536 6917 218 5877 418 22902 22614 262 1063 26384 4409 524 880 601 2192 201 1599
2020-04-26 21100 99000 550 7120 220 6070 430 23300 23000 270 1170 26800 4530 560 910 630 2290 210 1630
2020-04-27 21700 101000 560 7290 230 6260 440 23600 23400 290 1270 27200 4640 590 940 660 2400 210 1660
2020-04-28 22400 103000 570 7470 230 6470 450 24000 23800 300 1380 27600 4750 630 960 690 2530 220 1690
2020-04-29 23200 105000 590 7650 240 6670 460 24400 24200 320 1500 28100 4870 660 990 720 2660 220 1720
2020-04-30 23900 107000 600 7830 240 6890 470 24700 24600 330 1640 28500 4990 700 1020 750 2800 230 1760
2020-05-01 24700 109000 610 8020 250 7120 480 25100 25000 350 1790 28900 5120 740 1050 790 2940 230 1790
2020-05-02 25500 111000 630 8220 250 7360 490 25500 25400 370 1960 29400 5250 780 1080 820 3100 240 1830

Deaths count average forecast (bold black line in graphs) 2020-04-26 to 2020-05-02

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-04-25 4057 2547 825 720 5650 1305 494 2706 53755 1677 670 1862 1054 897 1884 1644 2730 3273 5863 16599 743 263877
2020-04-26 4410 2720 880 750 5730 1450 510 2820 55800 1760 720 1980 1100 920 1980 1700 2920 3420 6130 17000 760 272000
2020-04-27 4720 2900 930 770 5810 1590 520 2950 57700 1850 770 2100 1140 950 2080 1750 3120 3550 6410 17400 770 281000
2020-04-28 5040 3100 980 790 5900 1730 530 3090 59700 1940 830 2220 1190 980 2190 1810 3340 3690 6690 17800 790 290000
2020-04-29 5390 3310 1040 810 5980 1890 540 3230 61700 2040 890 2360 1250 1010 2300 1860 3570 3830 6990 18100 810 298000
2020-04-30 5760 3540 1100 830 6060 2070 550 3380 63800 2150 960 2500 1300 1050 2410 1920 3810 3980 7310 18500 820 307000
2020-05-01 6170 3780 1170 850 6150 2270 570 3540 66000 2260 1040 2650 1360 1080 2540 1980 4080 4130 7640 18900 840 316000
2020-05-02 6600 4050 1240 880 6240 2480 580 3710 68200 2380 1120 2820 1420 1120 2670 2040 4370 4300 7990 19300 860 324000

Deaths count forecast Europe (bold red line in graphs) 2020-04-26 to 2020-05-02

DateUKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCH
2020-04-25 20319 96522 536 6917 218 5877 418 22902 22614 262 1063 26384 4409 524 880 601 2192 201 1599
2020-04-26 21000 98000 550 7110 220 6050 430 23200 23000 270 1130 26800 4520 560 910 630 2280 200 1630
2020-04-27 21700 100000 560 7300 230 6220 440 23600 23300 280 1190 27200 4630 600 930 660 2370 210 1650
2020-04-28 22400 102000 570 7480 230 6400 450 23900 23700 290 1280 27600 4740 640 960 690 2480 210 1670
2020-04-29 23100 104000 580 7670 230 6580 460 24300 24000 290 1360 28000 4850 680 990 720 2580 220 1700
2020-04-30 23800 106000 590 7860 240 6770 470 24600 24300 300 1450 28400 4970 720 1020 750 2690 220 1720
2020-05-01 24600 108000 600 8050 240 6960 480 25000 24700 310 1550 28800 5080 770 1050 780 2810 220 1750
2020-05-02 25300 110000 610 8260 250 7150 490 25400 25000 320 1660 29200 5200 820 1080 820 2920 230 1770

Deaths count forecast (bold red line in graphs) 2020-04-26 to 2020-05-02

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-04-25 4057 2547 825 720 5650 1305 494 2706 53755 1677 670 1862 1054 897 1884 1644 2730 3273 5863 16599 743 263877
2020-04-26 4470 2700 870 740 5730 1430 500 2820 55700 1760 690 1970 1090 920 1970 1700 2910 3390 6130 17000 760 272000
2020-04-27 4950 2860 920 760 5810 1600 520 2930 57600 1840 710 2090 1130 930 2060 1750 3100 3530 6390 17400 770 280000
2020-04-28 5480 3040 970 790 5890 1770 530 3050 59600 1920 750 2210 1180 950 2160 1810 3300 3680 6660 17700 790 289000
2020-04-29 6040 3230 1020 810 5970 1960 540 3170 61600 2000 780 2330 1220 970 2260 1870 3510 3830 6940 18100 800 296000
2020-04-30 6670 3430 1080 840 6050 2160 560 3290 63600 2090 810 2460 1270 980 2370 1930 3730 3990 7230 18500 820 304000
2020-05-01 7360 3640 1140 870 6130 2390 570 3420 65800 2180 850 2600 1310 1000 2480 1990 3970 4160 7530 18900 830 312000
2020-05-02 8130 3870 1210 900 6220 2640 580 3550 68000 2280 880 2750 1360 1010 2600 2060 4220 4330 7850 19300 850 320000

Deaths count scenario forecast (bold green line in graphs) 2020-04-26 to 2020-05-04

DateUKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCHBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-04-25 20319 96522 536 6917 218 5877 418 22902 22614 262 1063 26384 4409 524 880 601 2192 201 1599 4057 2547 825 720 5650 1305 494 2706 53755 1677 670 1862 1054 897 1884 1644 2730 3273 5863 16599 743 263877
2020-04-26 20700 98000 550 7100 220 6080 420 23200 22900 280 1110 26800 4510 550 900 610 2310 210 1630 4270 2680 850 730 5720 1310 500 2810 56000 1750 720 1950 1100 930 1960 1690 2880 3310 6080 17000 760 272000
2020-04-27 21300 100000 560 7300 230 6270 430 23500 23300 290 1160 27100 4610 560 930 620 2400 210 1670 4510 2800 890 750 5780 1400 510 2910 57900 1820 740 2040 1140 960 2040 1740 3020 3410 6310 17300 770 280000
2020-04-28 21900 102000 570 7460 230 6450 440 23800 23600 300 1210 27500 4710 570 950 630 2490 220 1690 4780 2930 920 770 5830 1430 520 3010 59800 1890 760 2130 1180 980 2120 1780 3150 3520 6530 17700 790 287000
2020-04-29 22400 104000 580 7610 230 6610 450 24000 23900 310 1250 27800 4810 590 970 640 2570 220 1720 5000 3060 950 790 5870 1470 530 3100 61700 1960 770 2210 1210 1010 2190 1830 3290 3620 6720 18000 800 295000
2020-04-30 23000 105000 590 7760 240 6780 450 24200 24200 320 1300 28200 4900 600 990 640 2670 220 1750 5230 3170 980 800 5900 1490 540 3190 63600 2030 770 2300 1250 1040 2270 1870 3430 3710 6930 18300 810 302000
2020-05-01 23400 107000 600 7880 240 6940 460 24400 24500 330 1340 28500 5000 610 1000 650 2760 230 1770 5430 3290 1010 810 5940 1500 540 3280 65400 2090 770 2380 1290 1060 2330 1910 3580 3800 7130 18500 820 310000
2020-05-02 23900 108000 610 7980 240 7100 470 24600 24800 340 1380 28800 5080 630 1020 660 2850 230 1790 5620 3410 1040 820 5970 1540 550 3360 67900 2150 780 2450 1330 1090 2400 1940 3700 3880 7330 18700 830 319000
2020-05-03 24400 109000 610 8080 240 7250 470 24700 25100 350 1420 29100 5170 640 1040 670 2940 230 1810 5790 3520 1070 830 6000 1590 550 3450 70200 2210 790 2530 1360 1110 2470 1980 3820 3960 7510 18900 840 327000
2020-05-04 24800 110000 620 8180 250 7400 470 24900 25400 360 1460 29400 5250 650 1060 680 3050 240 1830 5890 3630 1100 840 6030 1610 560 3530 72300 2270 790 2600 1390 1130 2530 2020 3940 4040 7700 19000 850 335000

Peak increase in estimated trend of Deaths in Europe 2020-04-25

UKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCH
Peak date04-1004-0704-0804-1004-0904-1504-0404-0204-07 -- --03-2804-07 --04-12 -- --04-0804-04
Peak daily increment 890 3240 23 368 11 275 18 902 1014 826 163 33 9 61
Days from 100 to peak 22 34 9 17 0 23 3 20 23 24 18 14 -1 12
Days from peak/2 to peak 16 24 17 15 14 22 14 16 17 19 18 20 21 17
Days since peak 15 18 17 15 16 10 21 23 18 28 18 13 17 21

Peak increase in estimated trend of Deaths 2020-04-25

BrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
Peak date -- -- --04-1203-19 --04-1204-1904-16 -- --04-20 --04-20 --04-14 -- --04-1604-0904-0604-17
Peak daily increment 38 150 30 125 3112 110 43 76 297 777 30 11223
Days from 100 to peak 14 14 11 22 30 18 21 18 20 18 15 82
Days from peak/2 to peak 24 16 25 24 20 20 29 21 18 14 30 49
Days since peak 13 37 13 6 9 5 5 11 9 16 19 8

Initial visual evaluation of forecasts of Deaths