COVID-19 short-term forecasts Deaths 2020-05-04


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 The documentation that is provided is still in progress and not peer reviewed.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.
[2020-04-27] Our short-term COVID-19 forecasting paper is now available as Nuffield Economics Discussion Paper 2020-W06.
A small adjustment has been made to the scenario forecast methodology, and will be documented shortly.
[2020-04-29] See our blog entry at the International Institute of Forecasters.
US history of death counts revised in Johns Hopkins/CSSE data.
UK death counts have been revised to include the deaths in care homes. In the Johns Hopkins/CSSE data set, which we use, the entire history has been revised. So forecasts made up to 2020-04-29 cannot be compared to later outcomes. In the ECDC data set only the last observation has changed, causing a jump in the series.

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) 2020-05-05 to 2020-05-11

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-05-04 28734 108935 600 7924 252 6993 493 25428 240 25168 351 1319 29079 5082 698 1063 818 2769 214 1784
2020-05-05 29200 110000 600 8000 260 7080 500 25600 240 25400 360 1340 29300 5140 710 1070 840 2810 220 1790
2020-05-06 29700 111000 610 8080 260 7180 500 25800 240 25600 370 1360 29600 5210 720 1090 860 2860 220 1810
2020-05-07 30300 112000 620 8160 260 7280 510 26100 250 25800 380 1380 29900 5290 730 1110 890 2920 220 1820
2020-05-08 30800 114000 620 8250 270 7390 520 26300 250 26100 390 1410 30200 5360 750 1120 910 2980 220 1840
2020-05-09 31400 115000 630 8330 270 7500 530 26600 260 26300 400 1430 30500 5440 760 1140 940 3040 220 1850
2020-05-10 32000 116000 630 8410 270 7610 540 26800 260 26500 410 1450 30800 5510 770 1160 970 3110 220 1860
2020-05-11 32600 117000 640 8500 280 7720 550 27100 270 26800 420 1470 31100 5590 790 1170 990 3170 220 1880

Deaths count average forecast (bold black line in graphs) 2020-05-05 to 2020-05-11

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-05-04 7367 4003 1566 864 6277 2271 623 3461 68922 2297 849 2556 258 1398 1228 2673 1991 4090 4135 7910 19415 846 329210
2020-05-05 7700 4140 1600 870 6330 2380 630 3530 70300 2340 860 2600 270 1430 1240 2740 2020 4220 4220 8060 19700 850 336000
2020-05-06 8100 4310 1700 880 6390 2510 650 3600 71800 2410 880 2660 280 1470 1260 2830 2050 4370 4310 8300 19900 860 343000
2020-05-07 8600 4490 1820 890 6450 2640 660 3680 73400 2470 900 2730 290 1500 1290 2930 2090 4530 4410 8550 20200 870 351000
2020-05-08 9000 4670 1930 900 6520 2780 680 3760 75100 2540 920 2800 300 1550 1320 3030 2130 4690 4510 8810 20500 880 358000
2020-05-09 9500 4860 2060 910 6580 2920 690 3840 76800 2610 940 2870 310 1590 1350 3130 2160 4850 4610 9090 20800 890 366000
2020-05-10 10000 5060 2190 930 6640 3080 710 3920 78500 2680 960 2940 330 1630 1380 3240 2200 5030 4710 9370 21000 900 373000
2020-05-11 10600 5270 2340 940 6710 3250 720 4010 80300 2760 980 3020 340 1680 1420 3350 2240 5210 4820 9660 21300 910 380000

Deaths count forecast Europe (bold red line in graphs) 2020-05-05 to 2020-05-11

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-05-04 28734 108935 600 7924 252 6993 493 25428 240 25168 351 1319 29079 5082 698 1063 818 2769 214 1784
2020-05-05 29100 110000 600 8000 260 7090 500 25700 250 25400 360 1340 29300 5140 710 1080 840 2840 220 1800
2020-05-06 29500 111000 610 8070 260 7190 510 25900 250 25600 370 1360 29600 5190 730 1090 860 2890 220 1810
2020-05-07 29900 112000 610 8130 260 7280 520 26100 260 25800 380 1380 29800 5240 740 1110 890 2940 220 1820
2020-05-08 30300 113000 620 8200 270 7380 530 26300 270 26000 390 1400 30100 5300 760 1120 910 3000 220 1830
2020-05-09 30700 114000 620 8270 270 7470 540 26500 270 26200 400 1420 30300 5350 770 1140 940 3050 220 1840
2020-05-10 31100 116000 620 8330 280 7570 550 26800 280 26400 410 1430 30600 5410 780 1150 970 3110 220 1850
2020-05-11 31500 117000 630 8400 280 7670 560 27000 280 26700 420 1450 30800 5460 800 1170 990 3170 220 1860

Deaths count forecast (bold red line in graphs) 2020-05-05 to 2020-05-11

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-05-04 7367 4003 1566 864 6277 2271 623 3461 68922 2297 849 2556 258 1398 1228 2673 1991 4090 4135 7910 19415 846 329210
2020-05-05 7670 4150 1660 880 6330 2390 640 3520 70200 2350 860 2610 270 1430 1250 2730 2020 4220 4220 8070 19700 850 336000
2020-05-06 7990 4320 1790 890 6400 2510 650 3590 71600 2400 880 2680 280 1460 1280 2780 2050 4350 4310 8240 19900 860 343000
2020-05-07 8310 4490 1920 910 6450 2630 670 3650 72900 2450 890 2750 290 1490 1310 2830 2080 4470 4400 8400 20100 860 350000
2020-05-08 8640 4670 2060 920 6510 2760 680 3720 74200 2510 900 2830 300 1520 1340 2880 2110 4600 4490 8580 20400 870 357000
2020-05-09 8990 4850 2210 930 6570 2890 700 3790 75600 2570 910 2900 310 1550 1370 2940 2130 4730 4580 8750 20600 880 364000
2020-05-10 9350 5050 2370 950 6630 3030 710 3850 77000 2620 920 2980 320 1590 1410 2990 2160 4870 4670 8930 20900 880 371000
2020-05-11 9730 5250 2540 960 6700 3180 730 3920 78400 2680 930 3060 330 1620 1440 3050 2190 5010 4760 9120 21100 890 378000

Deaths count scenario forecast (bold purple line in graphs) 2020-05-05 to 2020-05-13

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCHBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-05-04 28734 108935 600 7924 252 6993 493 25428 240 25168 351 1319 29079 5082 698 1063 818 2769 214 1784 7367 4003 1566 864 6277 2271 623 3461 68922 2297 849 2556 258 1398 1228 2673 1991 4090 4135 7910 19415 846 329210
2020-05-05 29400 110000 610 7990 260 7070 500 25700 240 25300 360 1340 29400 5150 710 1070 840 2820 220 1800 7680 4120 1620 870 6320 2370 630 3530 70400 2360 870 2600 270 1430 1240 2770 2030 4230 4230 8180 19600 860 338000
2020-05-06 29800 111000 610 8060 260 7140 510 25900 250 25500 370 1360 29600 5200 720 1090 860 2870 220 1810 7980 4250 1710 880 6370 2470 640 3580 71900 2410 880 2660 280 1470 1260 2850 2060 4360 4310 8360 19800 860 344000
2020-05-07 30300 112000 620 8140 260 7220 520 26100 250 25700 370 1380 29800 5250 730 1100 880 2920 220 1820 8260 4390 1790 890 6410 2560 640 3640 73400 2470 900 2710 290 1500 1270 2930 2090 4480 4390 8560 20000 870 349000
2020-05-08 30700 113000 620 8210 270 7290 520 26300 260 25800 380 1400 30000 5290 740 1120 900 2940 220 1830 8530 4510 1860 900 6440 2640 650 3690 74800 2520 910 2770 290 1530 1290 3000 2120 4610 4460 8740 20100 880 353000
2020-05-09 31100 114000 630 8300 270 7370 530 26400 260 26100 380 1420 30200 5320 750 1130 910 2980 220 1850 8810 4640 1930 910 6480 2720 650 3740 76200 2580 920 2800 300 1560 1300 3080 2140 4730 4530 8920 20300 880 357000
2020-05-10 31500 114000 630 8380 270 7440 530 26500 270 26200 390 1440 30300 5360 760 1140 930 3010 230 1860 9110 4750 1990 920 6510 2810 660 3780 77200 2630 940 2840 310 1580 1330 3140 2170 4860 4600 9100 20400 890 362000
2020-05-11 31900 115000 640 8450 280 7520 540 26700 270 26400 390 1460 30500 5390 770 1160 950 3040 230 1870 9340 4860 2050 930 6540 2890 660 3830 78100 2680 950 2870 320 1600 1340 3200 2190 4970 4670 9280 20600 890 366000
2020-05-12 32300 116000 640 8510 280 7600 540 26800 270 26500 400 1480 30600 5420 780 1170 960 3070 230 1880 9590 4960 2110 930 6570 2960 670 3860 79000 2730 960 2900 330 1620 1360 3250 2220 5090 4740 9420 20700 900 370000
2020-05-13 32700 116000 650 8580 280 7660 540 27000 270 26600 400 1500 30700 5450 790 1180 980 3100 230 1900 9870 5060 2170 940 6600 3040 670 3890 79900 2780 980 2930 330 1640 1370 3290 2240 5190 4810 9550 20900 910 374000

Peak increase in estimated trend of Deaths in Europe 2020-05-04

UKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
Peak date04-0904-0704-0804-1004-0904-1504-0404-0204-2104-0704-1804-2403-2804-0704-2504-1104-2904-2204-0804-04
Peak daily increment 1014 3241 23 364 11 280 18 902 19 1014 14 103 826 163 29 33 25 119 9 61
Days from 100 to peak 22 34 9 17 0 23 3 20 1 23 5 22 24 18 19 13 27 25 -1 12
Days from peak/2 to peak 16 24 17 15 14 22 14 16 23 17 25 25 19 18 29 19 35 27 21 17
Days since peak 25 27 26 24 25 19 30 32 13 27 16 10 37 27 9 23 5 12 26 30

Peak increase in estimated trend of Deaths 2020-05-04

BrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
Peak date04-30 -- --04-1403-1904-2904-1204-1904-1504-2304-2404-2004-2504-1404-20 --04-1504-2604-2104-2904-0904-0604-17
Peak daily increment 437 39 150 119 28 126 2279 88 57 118 13 48 45 75 176 148 299 777 30 9863
Days from 100 to peak 33 16 14 23 11 22 29 26 21 18 5 13 21 19 25 24 33 18 15 82
Days from peak/2 to peak 29 26 16 26 25 24 21 30 28 20 26 20 29 22 26 25 31 14 30 54
Days since peak 4 20 46 5 22 15 19 11 10 14 9 20 14 19 8 13 5 25 28 17

Initial visual evaluation of forecasts of Deaths