COVID-19 short-term forecasts Deaths 2020-05-02


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 The documentation that is provided is still in progress and not peer reviewed.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) 2020-05-03 to 2020-05-09

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-05-02 28131 107023 596 7765 245 6812 475 25100 220 24729 335 1286 28710 4987 664 1023 771 2669 211 1762
2020-05-03 28600 108000 600 7850 250 6920 480 25300 220 25000 340 1310 28900 5040 680 1040 790 2730 210 1780
2020-05-04 29200 110000 610 7960 250 7050 480 25600 230 25300 340 1340 29200 5110 700 1060 810 2800 210 1800
2020-05-05 29900 111000 620 8070 250 7170 490 25900 230 25500 350 1370 29500 5190 720 1070 840 2870 210 1820
2020-05-06 30500 112000 630 8180 260 7300 500 26100 240 25800 360 1410 29800 5270 740 1090 860 2950 220 1840
2020-05-07 31200 114000 630 8290 260 7430 500 26400 240 26100 370 1440 30100 5350 770 1110 890 3020 220 1860
2020-05-08 31900 115000 640 8410 260 7560 510 26700 250 26400 380 1470 30400 5430 790 1130 920 3100 220 1890
2020-05-09 32600 117000 650 8520 260 7700 520 27000 250 26800 390 1510 30700 5520 820 1150 940 3180 220 1910

Deaths count average forecast (bold black line in graphs) 2020-05-03 to 2020-05-09

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-05-02 6761 3684 1323 831 6156 2061 603 3336 66369 2213 830 2436 240 1363 1161 2576 1950 3846 4020 7742 18909 833
2020-05-03 7200 3790 1390 840 6220 2180 610 3420 68100 2270 850 2470 250 1390 1190 2670 1980 4000 4070 7950 19200 840
2020-05-04 7800 3950 1480 850 6290 2310 630 3510 70000 2340 870 2530 260 1430 1220 2800 2020 4170 4170 8200 19500 850
2020-05-05 8300 4120 1570 860 6360 2450 650 3600 71900 2420 890 2600 270 1470 1260 2950 2070 4340 4270 8500 19800 870
2020-05-06 9000 4290 1670 870 6420 2590 660 3690 73800 2500 910 2670 290 1520 1300 3100 2110 4530 4380 8800 20100 880
2020-05-07 9600 4470 1770 890 6490 2740 680 3780 75800 2580 930 2740 300 1560 1340 3260 2160 4720 4490 9110 20400 890
2020-05-08 10300 4660 1880 900 6560 2910 700 3880 77900 2660 960 2810 320 1610 1380 3430 2200 4920 4600 9440 20700 900
2020-05-09 11100 4860 2000 910 6640 3090 720 3980 80100 2750 980 2890 330 1660 1420 3600 2250 5130 4720 9790 21100 920

Deaths count forecast Europe (bold red line in graphs) 2020-05-03 to 2020-05-09

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-05-02 28131 107023 596 7765 245 6812 475 25100 220 24729 335 1286 28710 4987 664 1023 771 2669 211 1762
2020-05-03 28700 108000 600 7850 250 6920 480 25400 220 25000 340 1310 29000 5070 680 1040 790 2730 210 1780
2020-05-04 29300 110000 610 7930 250 7030 490 25700 230 25200 350 1330 29300 5150 690 1050 820 2800 210 1800
2020-05-05 30000 111000 620 8010 260 7130 500 26100 230 25400 360 1350 29700 5230 710 1070 840 2860 220 1810
2020-05-06 30600 112000 630 8080 260 7240 510 26400 230 25700 370 1370 30000 5310 720 1080 870 2930 220 1830
2020-05-07 31300 114000 630 8160 260 7350 510 26700 240 25900 380 1390 30300 5390 730 1090 890 3000 220 1840
2020-05-08 31900 115000 640 8230 270 7460 520 27100 240 26100 390 1410 30700 5470 750 1110 920 3070 220 1860
2020-05-09 32600 116000 650 8300 270 7570 530 27400 250 26400 400 1430 31000 5560 760 1120 950 3140 220 1870

Deaths count forecast (bold red line in graphs) 2020-05-03 to 2020-05-09

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-05-02 6761 3684 1323 831 6156 2061 603 3336 66369 2213 830 2436 240 1363 1161 2576 1950 3846 4020 7742 18909 833
2020-05-03 7200 3850 1410 840 6220 2170 620 3410 68100 2290 850 2510 250 1410 1180 2690 1990 3990 4110 8030 19200 840
2020-05-04 7600 4010 1500 860 6280 2290 630 3490 69800 2360 870 2590 260 1460 1190 2820 2030 4130 4230 8330 19500 850
2020-05-05 8100 4180 1600 880 6340 2410 650 3570 71600 2440 890 2660 270 1500 1210 2950 2080 4280 4340 8640 19700 860
2020-05-06 8600 4350 1700 890 6410 2540 670 3650 73400 2520 900 2740 280 1550 1230 3080 2120 4430 4460 8950 20000 870
2020-05-07 9100 4530 1810 910 6470 2680 690 3730 75200 2600 920 2820 290 1600 1250 3220 2160 4590 4580 9270 20300 880
2020-05-08 9700 4720 1930 920 6530 2820 700 3810 77100 2690 940 2900 300 1650 1270 3370 2210 4750 4700 9610 20600 880
2020-05-09 10300 4910 2060 940 6600 2970 720 3890 79000 2770 960 2980 310 1710 1290 3520 2250 4920 4830 9960 20900 890

Deaths count scenario forecast (bold purple line in graphs) 2020-05-03 to 2020-05-11

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCHBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-05-02 28131 107023 596 7765 245 6812 475 25100 220 24729 335 1286 28710 4987 664 1023 771 2669 211 1762 6761 3684 1323 831 6156 2061 603 3336 66369 2213 830 2436 240 1363 1161 2576 1950 3846 4020 7742 18909 833
2020-05-03 28700 108000 600 7850 250 6920 480 25200 230 25000 340 1300 28900 5040 680 1040 790 2750 210 1780 7300 3790 1390 840 6200 2160 610 3410 68000 2260 840 2490 250 1390 1200 2680 1990 3990 4070 8010 19100 850
2020-05-04 29300 109000 610 7940 250 7020 490 25400 230 25200 350 1320 29200 5110 690 1050 810 2830 210 1800 7900 3940 1460 850 6250 2260 620 3480 69600 2330 860 2570 260 1440 1230 2770 2030 4130 4160 8290 19400 860
2020-05-05 29800 110000 620 8010 250 7110 490 25500 240 25400 360 1340 29500 5170 710 1070 830 2900 220 1810 8400 4080 1530 860 6290 2360 620 3550 71100 2390 880 2650 270 1480 1260 2860 2070 4270 4240 8520 19600 870
2020-05-06 30300 111000 620 8080 260 7200 500 25700 240 25600 370 1370 29700 5220 720 1080 850 2970 220 1820 8900 4200 1600 870 6330 2440 630 3610 72600 2450 900 2740 280 1510 1290 2940 2100 4400 4320 8750 19800 880
2020-05-07 30900 112000 630 8150 260 7280 500 25800 250 25800 370 1390 30000 5260 730 1090 870 3040 220 1840 9500 4320 1670 870 6370 2540 630 3660 73900 2510 910 2810 280 1550 1320 3020 2140 4530 4390 8950 19900 880
2020-05-08 31300 112000 630 8230 260 7360 510 26000 250 25900 380 1410 30200 5310 740 1110 890 3080 220 1850 10100 4430 1730 880 6410 2620 640 3710 75200 2560 930 2860 290 1570 1350 3090 2170 4650 4470 9140 20100 890
2020-05-09 31700 113000 640 8300 260 7450 510 26100 260 26100 390 1420 30400 5350 750 1120 910 3120 230 1870 10700 4540 1790 890 6440 2710 640 3760 76400 2620 940 2900 300 1600 1370 3170 2190 4770 4540 9310 20200 900
2020-05-10 32000 114000 640 8360 260 7530 510 26300 260 26200 390 1440 30500 5390 760 1130 930 3150 230 1880 11300 4630 1840 900 6480 2800 650 3800 77300 2670 960 2940 310 1630 1390 3240 2210 4900 4600 9440 20400 910
2020-05-11 32300 115000 650 8440 270 7580 520 26500 260 26300 400 1460 30700 5420 770 1140 940 3180 230 1890 11900 4780 1900 910 6510 2880 660 3840 78400 2720 970 2990 310 1650 1400 3320 2240 5010 4660 9610 20500 910

Peak increase in estimated trend of Deaths in Europe 2020-05-02

UKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
Peak date04-0904-0704-0804-1004-0904-1504-0404-0204-2104-0704-1804-2403-2804-0704-2504-1104-2504-2204-0804-04
Peak daily increment 1014 3241 23 364 11 280 18 902 18 1014 15 100 826 163 29 33 25 116 9 61
Days from 100 to peak 22 34 9 17 0 23 3 20 1 23 5 22 24 18 19 13 23 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 31 27 21 17
Days since peak 23 25 24 22 23 17 28 30 11 25 14 8 35 25 7 21 7 10 24 28

Peak increase in estimated trend of Deaths 2020-05-02

BrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
Peak date -- -- --04-1403-19 --04-1204-1904-1504-2304-2404-2004-25 -- -- --04-1504-2504-21 --04-0904-06
Peak daily increment 39 150 28 125 2279 87 55 117 13 75 178 145 777 30
Days from 100 to peak 16 14 11 22 29 26 21 18 5 19 24 24 18 15
Days from peak/2 to peak 26 16 25 24 21 30 28 20 26 22 24 25 14 30
Days since peak 18 44 20 13 17 9 8 12 7 17 7 11 23 26

Initial visual evaluation of forecasts of Deaths