COVID-19 short-term forecasts Deaths 2020-05-08


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 The documentation that is provided is still in progress and not peer reviewed.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.
[2020-05-06] The New York Times is in the process of redefining its US state data. Unfortunately, at the moment only the last observation has changed (e.g New York deaths jumped from 19645 on 2020-05-05 to 25956 a day later). This means the data is currently useless; however it does bring it close to the Johns Hopkins/CSSE count (25626 on 2020-05-06). The aggregate US count is based on JH/CSSE so unaffected. We now use Johns Hopkins/CSSE US state data, including all states with sufficient counts. So the new forecasts cannot be compared to those previously.
A minor change is that we show the graph without scenario forecast if no peak has been detected yet.

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) 2020-05-09 to 2020-05-15

DateUKEUATBEBSCZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-05-08 31241 114251 614 8521 205 273 7510 522 26299 260 26192 392 1429 30201 5359 776 1114 923 3175 218 1823
2020-05-09 31500 115000 620 8610 210 280 7570 530 26500 260 26400 400 1440 30400 5400 780 1130 950 3200 220 1830
2020-05-10 32000 116000 620 8700 210 280 7650 530 26700 270 26600 410 1460 30700 5460 790 1140 980 3240 220 1840
2020-05-11 32500 117000 620 8800 210 290 7740 540 26900 280 26900 420 1470 30900 5510 810 1150 1010 3310 220 1850
2020-05-12 33000 118000 630 8910 210 290 7820 550 27100 280 27100 430 1490 31200 5570 820 1170 1050 3380 220 1860
2020-05-13 33500 120000 630 9030 210 300 7910 560 27300 290 27300 440 1510 31400 5630 840 1180 1080 3450 220 1870
2020-05-14 34000 121000 630 9150 220 300 7990 560 27500 290 27500 450 1530 31700 5690 860 1200 1120 3530 230 1890
2020-05-15 34600 122000 630 9270 220 310 8080 570 27700 300 27800 460 1550 31900 5760 870 1210 1150 3610 230 1900

Deaths count average forecast (bold black line in graphs) 2020-05-09 to 2020-05-15

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesRussiaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NVUS-NYUS-OHUS-OKUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
2020-05-08 10017 4697 1985 943 6541 3160 696 1723 3689 77180 383 516 2613 955 2871 304 213 1669 1378 243 3239 1328 298 2153 4698 1461 4346 534 467 409 530 8960 289 26243 1306 266 3715 320 236 1025 812 905 384
2020-05-09 10500 4850 2110 950 6590 3320 700 1800 3740 78300 400 530 2670 970 2920 310 220 1720 1410 250 3370 1360 310 2180 4760 1490 4390 560 470 430 540 9200 290 26500 1360 270 3940 330 240 1030 830 910 390
2020-05-10 11200 5020 2250 970 6650 3540 720 1890 3800 80000 410 550 2740 1000 2990 320 220 1790 1440 270 3530 1400 310 2210 4850 1540 4450 590 490 460 550 9500 300 26700 1420 270 4190 340 240 1050 850 920 390
2020-05-11 11900 5190 2400 980 6710 3780 730 1990 3850 81700 430 590 2810 1020 3060 330 230 1870 1480 280 3690 1440 320 2240 4960 1590 4510 620 510 480 570 9900 310 27000 1480 280 4450 340 250 1060 880 930 400
2020-05-12 12700 5380 2560 1000 6770 4030 750 2100 3910 83400 440 630 2890 1040 3140 350 240 1940 1520 290 3860 1480 330 2280 5080 1640 4570 650 530 510 590 10300 310 27300 1550 280 4730 350 260 1090 900 950 410
2020-05-13 13500 5560 2730 1020 6830 4310 760 2210 3970 85200 460 670 2960 1060 3220 360 250 2020 1560 310 4040 1520 340 2310 5200 1700 4630 680 550 540 600 10700 320 27600 1620 290 5020 360 260 1110 930 960 410
2020-05-14 14400 5760 2910 1040 6890 4600 780 2330 4030 87100 480 720 3040 1090 3300 370 260 2100 1610 320 4230 1560 350 2350 5330 1750 4690 720 570 570 620 11100 330 27900 1700 290 5340 370 270 1130 960 980 420
2020-05-15 15400 5970 3100 1060 6960 4900 800 2450 4090 89000 500 770 3120 1110 3380 390 270 2190 1650 340 4430 1600 360 2380 5460 1810 4760 760 590 610 640 11600 330 28200 1770 300 5670 370 270 1150 990 990 430

Deaths count forecast Europe (bold red line in graphs) 2020-05-09 to 2020-05-15

DateUKEUATBEBSCZDEDKESFIFRHUIEITNLPLPTROSENOCH
2020-05-08 31241 114251 614 8521 205 273 7510 522 26299 260 26192 392 1429 30201 5359 776 1114 923 3175 218 1823
2020-05-09 31700 115000 620 8640 210 280 7600 530 26500 260 26400 400 1450 30400 5410 790 1130 950 3270 220 1830
2020-05-10 32200 116000 620 8750 210 280 7710 540 26700 270 26600 410 1470 30700 5460 810 1140 970 3360 220 1840
2020-05-11 32700 118000 630 8870 210 290 7820 540 26900 270 26900 420 1490 30900 5520 830 1150 1000 3450 220 1850
2020-05-12 33200 119000 630 8990 210 290 7930 550 27100 280 27100 430 1520 31200 5570 850 1160 1030 3540 220 1860
2020-05-13 33700 120000 630 9110 210 300 8040 560 27300 280 27300 440 1540 31400 5630 860 1170 1070 3630 220 1870
2020-05-14 34200 121000 640 9230 220 300 8160 570 27500 290 27500 450 1560 31700 5690 880 1180 1100 3720 220 1880
2020-05-15 34700 122000 640 9360 220 310 8270 570 27700 290 27700 460 1580 31900 5740 900 1190 1130 3810 230 1890

Deaths count forecast (bold red line in graphs) 2020-05-09 to 2020-05-15

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesRussiaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NVUS-NYUS-OHUS-OKUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
2020-05-08 10017 4697 1985 943 6541 3160 696 1723 3689 77180 383 516 2613 955 2871 304 213 1669 1378 243 3239 1328 298 2153 4698 1461 4346 534 467 409 530 8960 289 26243 1306 266 3715 320 236 1025 812 905 384
2020-05-09 10700 4840 2110 960 6600 3390 710 1810 3740 78900 400 560 2690 970 2940 320 220 1730 1420 250 3420 1360 310 2180 4820 1500 4400 560 490 440 550 9200 300 26600 1370 270 4120 330 240 1050 830 920 390
2020-05-10 11300 5000 2240 980 6660 3640 720 1910 3790 80600 430 610 2760 990 3010 330 240 1800 1460 270 3600 1390 310 2210 4940 1540 4450 590 510 460 570 9500 300 26800 1440 280 4470 330 250 1070 860 930 400
2020-05-11 12000 5160 2370 1000 6720 3900 740 2010 3840 82300 450 670 2830 1010 3080 350 250 1870 1500 280 3790 1420 320 2240 5070 1590 4510 620 540 480 580 9800 310 27100 1520 280 4840 340 250 1100 890 950 410
2020-05-12 12800 5330 2500 1020 6780 4170 760 2110 3890 84100 480 730 2900 1030 3150 370 260 1940 1540 290 3980 1460 330 2270 5190 1640 4560 650 560 500 600 10100 310 27400 1590 290 5240 350 250 1120 930 960 420
2020-05-13 13600 5500 2650 1040 6840 4470 770 2220 3940 85800 510 790 2980 1050 3230 380 280 2020 1590 300 4190 1490 340 2300 5330 1690 4620 690 590 530 620 10500 320 27700 1670 290 5670 350 260 1150 960 980 420
2020-05-14 14400 5670 2800 1060 6910 4790 790 2340 3990 87600 540 860 3060 1080 3300 400 290 2100 1650 310 4400 1530 340 2330 5460 1740 4670 720 620 550 640 10900 320 28000 1750 300 6140 360 260 1170 990 1000 430
2020-05-15 15300 5850 2970 1080 6970 5120 800 2460 4040 89500 570 940 3140 1100 3380 420 310 2180 1700 330 4630 1560 350 2360 5600 1790 4730 760 650 580 660 11300 330 28300 1840 300 6640 370 270 1200 1030 1020 440

Deaths count scenario forecast (bold purple line in graphs) 2020-05-09 to 2020-05-17

DateUKEUATBEBSCZDEDKESFIFRHUIEITNLPLPTROSENOCHBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesRussiaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NVUS-NYUS-OHUS-OKUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
2020-05-08 31241 114251 614 8521 205 273 7510 522 26299 260 26192 392 1429 30201 5359 776 1114 923 3175 218 1823 10017 4697 1985 943 6541 3160 696 1723 3689 77180 383 516 2613 955 2871 304 213 1669 1378 243 3239 1328 298 2153 4698 1461 4346 534 467 409 530 8960 289 26243 1306 266 3715 320 236 1025 812 905 384
2020-05-09 31500 115000 620 8610 210 280 7570 530 26500 270 26400 400 1440 30400 5380 790 1130 940 3180 220 1830 10400 4810 2100 960 6600 3370 710 1770 3740 78200 390 520 2660 970 2910 310 220 1690 1390 250 3320 1360 300 2180 4810 1500 4400 550 470 440 540 9100 300 26500 1350 270 3920 330 240 1030 820 910 390
2020-05-10 32000 116000 620 8690 210 280 7660 530 26600 270 26500 400 1460 30600 5420 800 1140 960 3240 220 1840 11000 4920 2190 970 6660 3590 720 1840 3780 79800 410 560 2710 990 2960 320 230 1720 1420 260 3400 1390 310 2210 4930 1540 4460 570 480 460 560 9300 300 26800 1390 270 4080 330 240 1050 840 920 390
2020-05-11 32400 117000 620 8750 210 280 7730 530 26800 280 26700 410 1480 30800 5460 810 1150 980 3300 220 1850 11600 5030 2290 990 6710 3800 730 1900 3830 80900 420 580 2750 1000 3000 330 230 1760 1440 270 3480 1420 310 2240 5050 1580 4510 580 490 480 570 9400 300 27000 1430 280 4190 340 250 1070 850 930 400
2020-05-12 32800 118000 630 8790 210 290 7810 540 27000 280 26900 410 1500 31000 5500 830 1160 990 3340 220 1860 12100 5120 2370 1000 6760 4030 740 1960 3880 82100 430 610 2800 1010 3030 340 240 1790 1460 280 3550 1450 320 2260 5150 1620 4560 590 500 500 580 9600 310 27200 1460 280 4320 350 250 1090 870 940 410
2020-05-13 33100 119000 630 8840 210 290 7870 540 27100 280 27000 420 1520 31100 5530 840 1170 1010 3370 220 1870 12600 5200 2450 1010 6800 4240 750 2020 3920 83400 440 620 2830 1020 3050 340 240 1810 1470 280 3640 1470 320 2280 5250 1660 4600 600 500 500 590 9700 310 27400 1500 280 4430 360 250 1110 880 950 420
2020-05-14 33400 119000 630 8870 220 290 7930 550 27300 290 27100 420 1530 31200 5560 850 1180 1020 3390 220 1880 13200 5300 2530 1020 6840 4460 750 2080 3950 84600 450 640 2890 1040 3080 350 250 1850 1490 290 3700 1510 320 2300 5350 1700 4640 620 510 510 600 9800 310 27500 1530 290 4480 370 250 1120 890 950 420
2020-05-15 33700 120000 630 8910 220 290 7970 550 27400 290 27200 420 1550 31400 5600 860 1190 1040 3420 220 1890 13700 5370 2610 1020 6870 4660 760 2140 3990 86100 460 640 2920 1050 3110 360 250 1870 1500 290 3720 1530 330 2320 5420 1730 4680 630 520 530 610 10000 310 27600 1560 290 4570 370 250 1130 900 950 430
2020-05-16 34000 121000 640 8940 220 290 8030 550 27600 290 27200 430 1570 31500 5640 860 1200 1060 3460 230 1900 14200 5460 2660 1030 6910 4850 770 2190 4030 86800 470 650 2950 1060 3140 360 260 1900 1520 300 3750 1560 330 2340 5490 1760 4730 640 520 540 620 10100 320 27600 1590 290 4650 380 260 1140 910 960 430
2020-05-17 34200 121000 640 8960 220 300 8080 560 27700 290 27400 430 1580 31700 5670 870 1210 1060 3490 230 1910 14800 5540 2720 1030 6950 5040 770 2240 4060 87400 480 650 2980 1060 3170 370 270 1920 1530 300 3800 1580 330 2360 5570 1790 4760 650 540 550 620 10200 320 27600 1620 290 4740 390 260 1150 920 960 430

Peak increase in estimated trend of Deaths in Europe 2020-05-08

UKEUATBEBSCZDEDKESFIFRHUIEITNLPLPTROSENOCH
Peak date04-0904-0704-0804-1004-1104-0904-1504-0404-0204-2104-0704-1804-2403-2804-0704-2404-11 --04-2204-0804-04
Peak daily increment 1014 3241 23 364 8 11 280 18 902 19 1014 14 110 826 163 30 33 117 9 61
Days from 100 to peak 22 34 9 17 1 0 23 3 20 1 23 5 22 24 18 18 13 25 -1 12
Days from peak/2 to peak 16 24 17 15 22 14 22 14 16 23 17 25 25 19 18 28 19 27 21 17
Days since peak 29 31 30 28 27 29 23 34 36 17 31 20 14 41 31 14 27 16 30 34

Peak increase in estimated trend of Deaths 2020-05-08

BrazilCanadaIndiaIndonesiaIranMexicoPhilippinesRussiaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NVUS-NYUS-OHUS-OKUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
Peak date -- -- --04-1403-19 --04-12 --04-1904-15 -- --04-2304-2404-20 -- -- --04-07 -- --04-2904-1604-1504-2504-2904-22 -- -- --04-2004-2904-1204-07 --04-21 --04-2904-0304-3004-2204-0604-05
Peak daily increment 39 150 28 125 2279 90 58 120 46 50 13 75 272 64 148 20 303 11 1066 10 20 10 37 58 32 13
Days from 100 to peak 16 14 11 22 29 26 20 18 7 25 1 19 24 22 25 7 28 0 18 8 14 -8 26 7 14 -3
Days from peak/2 to peak 26 16 25 24 21 30 27 20 11 32 20 22 23 27 26 20 28 17 13 26 33 5 35 13 29 11
Days since peak 24 50 26 19 23 15 14 18 31 9 22 23 13 9 16 18 9 26 31 17 9 35 8 16 32 33

Initial visual evaluation of forecasts of Deaths