COVID-19 short-term forecasts Deaths 2020-04-29


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 The documentation that is provided is still in progress and not peer reviewed.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) 2020-04-30 to 2020-05-06

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-04-29 26097 103020 580 7501 227 6467 443 24275 206 24087 300 1190 27682 4711 624 973 693 2462 207 1716
2020-04-30 26800 105000 590 7650 230 6620 450 24600 210 24400 310 1230 28000 4790 660 1000 720 2540 210 1740
2020-05-01 27400 106000 600 7810 230 6750 460 24900 220 24800 320 1280 28300 4870 690 1020 750 2590 210 1770
2020-05-02 28000 108000 610 7970 230 6890 460 25200 230 25100 330 1330 28600 4950 720 1050 780 2650 210 1800
2020-05-03 28600 110000 620 8130 240 7040 470 25600 230 25400 340 1380 29000 5030 760 1070 820 2710 210 1820
2020-05-04 29300 111000 630 8290 240 7190 480 25900 240 25800 340 1440 29300 5120 790 1100 850 2770 220 1850
2020-05-05 30000 113000 640 8470 240 7340 490 26200 250 26100 350 1500 29600 5200 830 1130 890 2840 220 1880
2020-05-06 30700 115000 640 8650 240 7500 490 26500 260 26500 360 1560 30000 5290 880 1150 930 2900 220 1900

Deaths count average forecast (bold black line in graphs) 2020-04-30 to 2020-05-06

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-04-29 5513 3155 1079 784 5957 1732 558 3081 60967 1961 764 2168 205 1216 1086 2221 1802 3405 3670 6770 18015 805
2020-04-30 5960 3310 1150 800 6030 1870 580 3180 62900 2040 790 2250 220 1260 1150 2310 1840 3610 3780 7020 18400 820
2020-05-01 6400 3470 1220 820 6100 1990 590 3280 64700 2100 820 2320 230 1290 1190 2390 1880 3830 3880 7210 18700 830
2020-05-02 6860 3630 1300 840 6180 2130 600 3380 66400 2170 850 2400 240 1330 1240 2480 1920 4060 3980 7400 19100 850
2020-05-03 7360 3800 1370 860 6250 2270 620 3490 68200 2240 880 2490 250 1360 1290 2560 1960 4310 4090 7600 19400 860
2020-05-04 7900 3980 1460 880 6320 2420 630 3600 70100 2310 910 2570 270 1400 1340 2660 2000 4570 4200 7810 19800 880
2020-05-05 8480 4180 1550 910 6400 2580 650 3710 72100 2390 940 2670 280 1430 1400 2750 2040 4850 4310 8030 20200 890
2020-05-06 9100 4380 1640 930 6480 2760 660 3830 74100 2470 970 2760 300 1470 1460 2850 2080 5140 4430 8250 20500 910

Deaths count forecast Europe (bold red line in graphs) 2020-04-30 to 2020-05-06

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-04-29 26097 103020 580 7501 227 6467 443 24275 206 24087 300 1190 27682 4711 624 973 693 2462 207 1716
2020-04-30 26600 104000 590 7650 230 6590 450 24500 210 24400 310 1240 28000 4790 650 1000 730 2530 210 1740
2020-05-01 27300 106000 600 7800 230 6720 460 24900 220 24700 310 1280 28300 4870 690 1020 770 2590 210 1770
2020-05-02 27900 108000 610 7950 230 6850 460 25200 220 25000 320 1320 28600 4950 720 1040 820 2660 210 1790
2020-05-03 28500 109000 620 8090 230 6980 470 25500 230 25400 330 1360 28900 5040 750 1070 860 2730 210 1820
2020-05-04 29200 111000 620 8250 230 7120 480 25800 230 25700 330 1400 29200 5120 790 1090 900 2810 210 1840
2020-05-05 29800 112000 630 8400 240 7260 490 26100 240 26000 340 1450 29500 5210 820 1120 950 2880 210 1870
2020-05-06 30500 114000 640 8560 240 7400 490 26500 250 26400 350 1500 29800 5290 860 1140 1000 2960 210 1890

Deaths count forecast (bold red line in graphs) 2020-04-30 to 2020-05-06

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-04-29 5513 3155 1079 784 5957 1732 558 3081 60967 1961 764 2168 205 1216 1086 2221 1802 3405 3670 6770 18015 805
2020-04-30 5910 3300 1150 800 6030 1900 570 3170 62600 2020 790 2230 220 1250 1170 2300 1840 3660 3760 6950 18300 820
2020-05-01 6330 3440 1220 820 6100 2080 590 3260 64400 2090 810 2290 230 1280 1250 2380 1880 3930 3850 7160 18700 830
2020-05-02 6770 3590 1290 830 6170 2260 610 3350 66300 2160 830 2350 240 1320 1330 2470 1910 4230 3950 7390 19000 840
2020-05-03 7240 3740 1370 850 6240 2460 620 3440 68200 2230 860 2410 250 1350 1420 2560 1950 4560 4050 7610 19300 860
2020-05-04 7730 3900 1460 870 6320 2670 640 3540 70100 2310 880 2470 260 1390 1510 2650 1990 4910 4150 7850 19600 870
2020-05-05 8270 4060 1550 880 6390 2900 660 3640 72200 2380 910 2540 280 1430 1610 2750 2030 5290 4260 8100 20000 880
2020-05-06 8840 4240 1650 900 6460 3160 680 3740 74300 2460 940 2600 290 1470 1710 2850 2070 5690 4370 8350 20300 900

Deaths count scenario forecast (bold purple line in graphs) 2020-04-30 to 2020-05-08

DateUKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCHBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-04-29 26097 103020 580 7501 227 6467 443 24275 206 24087 300 1190 27682 4711 624 973 693 2462 207 1716 5513 3155 1079 784 5957 1732 558 3081 60967 1961 764 2168 205 1216 1086 2221 1802 3405 3670 6770 18015 805
2020-04-30 26600 104000 580 7600 230 6600 450 24500 210 24300 310 1230 28000 4760 650 990 710 2500 210 1740 6000 3250 1140 800 6000 1830 560 3170 61800 2020 790 2230 210 1230 1120 2280 1840 3560 3760 6850 18200 820
2020-05-01 27200 105000 590 7690 230 6740 450 24700 220 24600 310 1260 28200 4820 680 1010 740 2560 210 1760 6500 3380 1210 820 6050 1910 570 3250 63400 2090 820 2290 220 1260 1160 2350 1870 3700 3860 7040 18500 830
2020-05-02 27800 107000 600 7780 230 6890 460 24900 220 24900 320 1290 28500 4880 700 1030 760 2630 210 1780 7000 3510 1280 830 6100 1990 580 3330 64900 2150 850 2350 220 1300 1190 2430 1900 3850 3950 7260 18700 840
2020-05-03 28400 108000 600 7860 230 7070 460 25100 230 25100 330 1340 28700 4950 720 1050 770 2680 220 1800 7600 3620 1340 840 6140 2080 590 3400 66300 2210 870 2410 230 1330 1230 2500 1940 3980 4020 7480 19000 850
2020-05-04 29000 109000 610 7940 240 7220 470 25300 230 25300 340 1380 29000 5020 750 1070 790 2730 220 1810 8100 3740 1400 850 6170 2160 600 3480 67800 2270 900 2470 240 1360 1260 2560 1960 4110 4100 7710 19200 870
2020-05-05 29500 110000 610 8020 240 7310 470 25500 240 25400 340 1390 29200 5080 770 1080 810 2780 220 1830 8700 3850 1460 860 6200 2240 600 3550 69100 2330 930 2520 240 1390 1280 2610 1990 4250 4200 7920 19400 880
2020-05-06 30000 111000 620 8090 240 7460 470 25600 240 25500 350 1420 29400 5140 780 1090 830 2840 220 1840 9500 3970 1520 870 6240 2300 610 3620 70200 2380 950 2580 250 1410 1300 2670 2000 4370 4260 8110 19500 890
2020-05-07 30500 112000 630 8170 240 7610 480 25800 250 25600 360 1450 29600 5200 810 1100 850 2880 230 1860 10200 4080 1580 880 6280 2370 610 3680 71500 2440 970 2640 250 1430 1330 2730 2020 4490 4330 8290 19700 890
2020-05-08 30800 112000 630 8240 240 7750 480 25900 250 25700 370 1480 29800 5250 830 1120 860 2920 230 1870 11000 4190 1630 880 6310 2390 610 3740 72400 2490 1000 2690 260 1450 1350 2780 2040 4610 4410 8450 19800 900

Peak increase in estimated trend of Deaths in Europe 2020-04-29

UKEUATBECZDEDKESFIFRHUIEITNLPLPTROSENOCH
Peak date04-0904-0704-0804-1004-0904-1504-0404-0204-2104-0704-1804-2403-2804-07 --04-11 --04-2204-0804-04
Peak daily increment 1014 3240 23 364 11 283 18 902 17 1014 14 113 826 163 33 113 9 61
Days from 100 to peak 22 34 9 17 0 23 3 20 1 23 5 22 24 18 13 25 -1 12
Days from peak/2 to peak 16 24 17 15 14 22 14 16 24 17 25 24 19 18 19 27 21 17
Days since peak 20 22 21 19 20 14 25 27 8 22 11 5 32 22 18 7 21 25

Peak increase in estimated trend of Deaths 2020-04-29

BrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-DCUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
Peak date --04-22 --04-1403-19 --04-1204-1904-1504-2204-2404-2004-2504-14 --04-2304-14 --04-2104-1604-0904-06
Peak daily increment 165 39 150 28 125 2278 87 59 115 13 49 99 76 144 304 777 30
Days from 100 to peak 21 16 14 11 22 29 25 21 18 5 13 23 18 24 20 18 15
Days from peak/2 to peak 22 26 16 25 24 21 29 28 20 26 20 26 21 25 18 14 30
Days since peak 7 15 41 17 10 14 7 5 9 4 15 6 15 8 13 20 23

Initial visual evaluation of forecasts of Deaths