COVID-19 short-term forecasts Deaths 2020-04-27


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 The documentation that is provided is still in progress and not peer reviewed.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.

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) 2020-04-28 to 2020-05-04

DateUKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCH
2020-04-27 21092 99386 549 7207 223 6126 427 23521 23293 280 1102 26977 4518 562 928 641 2274 205 1665
2020-04-28 21600 101000 560 7370 230 6260 430 23800 23600 290 1150 27300 4590 590 950 670 2330 210 1690
2020-04-29 22200 103000 570 7540 230 6430 440 24200 24000 300 1240 27700 4690 610 980 700 2420 210 1720
2020-04-30 22800 105000 570 7710 230 6600 450 24500 24300 310 1340 28000 4790 640 1010 730 2510 220 1750
2020-05-01 23500 107000 580 7890 240 6770 460 24900 24700 330 1440 28400 4890 670 1030 760 2600 220 1780
2020-05-02 24100 108000 590 8070 240 6950 470 25200 25000 340 1560 28800 4990 710 1060 790 2700 220 1800
2020-05-03 24800 110000 600 8260 240 7130 480 25600 25400 350 1690 29200 5100 740 1090 830 2800 230 1830
2020-05-04 25500 112000 610 8460 250 7330 490 26000 25700 370 1840 29600 5210 780 1120 870 2910 230 1860

Deaths count average forecast (bold black line in graphs) 2020-04-28 to 2020-05-04

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-04-27 4603 2841 939 765 5806 1434 511 2900 56259 1800 705 2012 1087 981 1992 1697 3003 3406 6044 17303 771 275002
2020-04-28 4950 3000 1000 790 5880 1520 520 3000 57800 1860 730 2100 1110 1010 2060 1730 3170 3510 6190 17700 790 282000
2020-04-29 5340 3170 1060 820 5960 1640 530 3120 59700 1940 770 2200 1150 1030 2150 1780 3360 3620 6410 18000 800 291000
2020-04-30 5760 3350 1130 840 6040 1780 550 3240 61600 2020 820 2310 1190 1060 2240 1820 3570 3740 6640 18400 810 299000
2020-05-01 6210 3550 1200 870 6110 1920 560 3360 63600 2100 870 2420 1230 1080 2340 1870 3800 3860 6870 18800 830 308000
2020-05-02 6700 3750 1270 890 6190 2070 570 3490 65600 2190 930 2540 1280 1110 2440 1920 4030 3990 7110 19200 840 316000
2020-05-03 7220 3980 1350 920 6280 2250 580 3620 67800 2280 990 2670 1320 1130 2550 1970 4290 4130 7370 19600 860 325000
2020-05-04 7780 4210 1430 950 6360 2440 600 3760 70000 2380 1060 2800 1370 1160 2670 2020 4560 4270 7630 20000 880 333000

Deaths count forecast Europe (bold red line in graphs) 2020-04-28 to 2020-05-04

DateUKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCH
2020-04-27 21092 99386 549 7207 223 6126 427 23521 23293 280 1102 26977 4518 562 928 641 2274 205 1665
2020-04-28 21500 101000 560 7370 230 6250 430 23800 23600 290 1130 27300 4580 580 950 670 2340 210 1690
2020-04-29 22000 102000 560 7540 230 6390 440 24100 24000 300 1160 27600 4650 600 980 700 2400 210 1720
2020-04-30 22400 104000 570 7710 230 6530 450 24400 24300 310 1190 28000 4720 630 1000 730 2450 210 1750
2020-05-01 22800 106000 570 7880 230 6670 450 24700 24700 310 1220 28300 4780 650 1030 760 2500 220 1780
2020-05-02 23300 107000 580 8050 240 6810 460 25000 25000 320 1250 28600 4850 670 1050 790 2560 220 1810
2020-05-03 23800 109000 590 8230 240 6950 460 25400 25400 330 1290 29000 4920 700 1080 820 2620 220 1840
2020-05-04 24200 110000 590 8420 240 7100 470 25700 25800 340 1320 29300 4990 720 1110 850 2670 220 1870

Deaths count forecast (bold red line in graphs) 2020-04-28 to 2020-05-04

DateBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-04-27 4603 2841 939 765 5806 1434 511 2900 56259 1800 705 2012 1087 981 1992 1697 3003 3406 6044 17303 771 275002
2020-04-28 4960 2970 1000 790 5880 1510 520 3000 57700 1860 730 2080 1110 1000 2040 1730 3160 3500 6190 17700 780 284000
2020-04-29 5350 3110 1070 810 5960 1600 530 3100 59200 1930 760 2150 1130 1030 2100 1770 3320 3590 6330 18000 800 292000
2020-04-30 5740 3250 1140 840 6040 1690 540 3200 60700 1990 790 2230 1150 1070 2150 1800 3490 3690 6470 18400 810 299000
2020-05-01 6170 3390 1210 860 6110 1780 550 3310 62200 2060 820 2310 1170 1100 2210 1840 3660 3780 6620 18700 820 307000
2020-05-02 6620 3550 1290 890 6190 1890 570 3410 63800 2130 850 2400 1190 1140 2270 1880 3840 3880 6760 19100 840 315000
2020-05-03 7110 3710 1370 910 6270 1990 580 3520 65400 2210 880 2480 1220 1170 2330 1920 4030 3990 6900 19400 850 322000
2020-05-04 7640 3880 1460 940 6350 2110 590 3640 67000 2280 920 2580 1240 1210 2390 1960 4230 4090 7050 19800 870 330000

Deaths count scenario forecast (bold purple line in graphs) 2020-04-28 to 2020-05-06

DateUKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCHBrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
2020-04-27 21092 99386 549 7207 223 6126 427 23521 23293 280 1102 26977 4518 562 928 641 2274 205 1665 4603 2841 939 765 5806 1434 511 2900 56259 1800 705 2012 1087 981 1992 1697 3003 3406 6044 17303 771 275002
2020-04-28 21600 101000 560 7370 230 6300 440 23900 23600 290 1170 27300 4620 590 950 670 2340 210 1680 4980 2970 990 780 5860 1520 520 2990 57900 1860 730 2090 1120 990 2070 1740 3180 3510 6280 17600 780 284000
2020-04-29 22200 102000 560 7510 230 6470 440 24200 24000 300 1240 27600 4710 610 980 690 2410 210 1700 5390 3100 1040 810 5920 1600 530 3090 59500 1920 760 2180 1140 1010 2140 1780 3330 3620 6490 17900 800 291000
2020-04-30 22700 104000 570 7630 230 6630 450 24500 24300 300 1300 27900 4800 630 1000 720 2460 210 1720 5830 3230 1100 820 5970 1680 540 3170 60900 1980 790 2260 1150 1020 2210 1810 3480 3720 6650 18100 810 298000
2020-05-01 23100 105000 570 7730 230 6770 460 24700 24600 310 1340 28200 4890 650 1020 740 2530 210 1740 6290 3350 1140 840 6020 1760 550 3260 62300 2040 810 2340 1170 1040 2270 1830 3620 3810 6860 18400 820 305000
2020-05-02 23500 106000 580 7820 240 6950 460 24900 24900 310 1400 28400 4970 670 1040 770 2590 220 1750 6750 3470 1180 850 6060 1840 560 3350 63700 2100 830 2420 1180 1060 2340 1860 3760 3890 7030 18600 830 312000
2020-05-03 23900 107000 580 7900 240 7120 470 25100 25100 320 1450 28700 5060 680 1060 790 2650 220 1770 7230 3580 1220 860 6090 1910 560 3430 65100 2160 850 2500 1210 1080 2400 1880 3890 3980 7220 18700 840 319000
2020-05-04 24300 108000 590 7990 240 7270 480 25200 25300 330 1490 28900 5140 700 1080 810 2710 220 1790 7700 3700 1270 870 6130 1990 570 3510 66500 2220 880 2580 1230 1100 2460 1900 4010 4050 7400 18900 850 325000
2020-05-05 24700 109000 590 8070 240 7420 480 25400 25600 330 1540 29100 5210 720 1100 830 2780 220 1800 8180 3810 1310 870 6160 2060 580 3580 67700 2280 900 2660 1250 1110 2510 1930 4130 4130 7540 19100 860 330000
2020-05-06 25100 110000 600 8150 250 7560 480 25500 25800 340 1590 29300 5260 740 1120 850 2870 230 1820 8610 3920 1350 880 6200 2120 580 3640 68900 2330 920 2730 1280 1130 2560 1950 4250 4210 7730 19300 860 335000

Peak increase in estimated trend of Deaths in Europe 2020-04-27

UKEUATBECZDEDKESFRHUIEITNLPLPTROSENOCH
Peak date04-1004-0704-0804-1004-0904-1504-0404-0204-0704-18 --03-2804-07 --04-12 --04-2204-0804-04
Peak daily increment 883 3240 23 364 11 280 18 902 1014 14 826 163 33 112 9 61
Days from 100 to peak 22 34 9 17 0 23 3 20 23 5 24 18 14 25 -1 12
Days from peak/2 to peak 16 24 17 15 14 22 14 16 17 25 19 18 20 27 21 17
Days since peak 17 20 19 17 18 12 23 25 20 9 30 20 15 5 19 23

Peak increase in estimated trend of Deaths 2020-04-27

BrazilCanadaIndiaIndonesiaIranMexicoPhilippinesTurkeyUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WAWorld
Peak date --04-22 --04-1203-19 --04-1204-1904-1504-22 --04-2004-1404-2004-2304-14 --04-1504-1604-0904-0604-17
Peak daily increment 165 38 150 29 125 2264 86 114 49 43 97 76 140 301 777 30 9791
Days from 100 to peak 21 14 14 11 22 29 25 18 13 21 23 18 18 20 18 15 82
Days from peak/2 to peak 22 24 16 25 24 21 29 20 20 29 26 21 19 18 14 30 54
Days since peak 5 15 39 15 8 12 5 7 13 7 4 13 12 11 18 21 10

Initial visual evaluation of forecasts of Deaths