COVID-19 short-term forecasts Deaths 2020-10-17


General information

  • 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. 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.
  • A list of notes is below. The most recent note:
    [2020-10-11]Short-term forecasting of the coronavirus pandemic (with Jennie Castle and David Hendry) is now in press at the International Journal of Forecasting.

Peak increase in estimated trend of Deaths in Europe 2020-10-17

UKEUATBEBGBSCZDEDKESFIFRGRHRHUIEITNLPLPTROSENOCH
Peak date (mm-dd)04-1006-1909-0708-15 --10-06 --04-1508-1006-1904-2109-18 -- -- --07-2308-1509-26 --07-05 --09-0108-1406-16
Peak daily increment 964 805 5 9 2 226 1 914 30 60 4 127 19 7 4 3 7
Days since peak 190 120 40 63 11 185 68 120 179 29 86 63 21 104 46 64 123
Last total 43579 156378 889 10392 968 314 1352 9785 679 33775 351 33054 500 355 1109 1849 36474 6737 3524 2162 5812 5918 278 2122
Last daily increment 150 438 7 33 10 1 69 12 2 0 0 0 10 10 24 8 47 29 84 13 63 0 0 0
Last week 754 4202 38 201 76 11 365 159 10 846 5 698 51 31 155 23 308 153 520 82 401 24 3 34
Previous peak date --04-0604-0604-15 --04-11 -- --04-0403-31 --04-08 -- --04-2304-2403-2904-0404-2404-1505-0204-1204-0704-06
Previous peak daily increment 3095 20 328 7 15 946 943 14 167 781 153 26 31 25 98 8 60
Low between peaks -892 0 2 -2 0 -1351 7 0 6 -6 3 -5 0 1

Peak increase in estimated trend of Deaths 2020-10-17

AustraliaBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-ARUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-IDUS-ILUS-INUS-KSUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-MTUS-NCUS-NDUS-NEUS-NHUS-NJUS-NMUS-NVUS-NYUS-OHUS-OKUS-ORUS-PAUS-SCUS-SDUS-TNUS-TXUS-UTUS-VAUS-WAUS-WIUS-WV
Peak date (mm-dd)09-0307-2110-0207-1708-2309-1509-19 --10-0507-2309-14 --07-2209-2408-0409-1609-1507-2108-0609-0507-1309-0906-2308-0708-2208-2709-0905-1208-15 -- --08-0908-2004-2909-0905-2810-0908-26 --09-17 --05-2005-2906-2505-1308-1308-3108-29 --07-2805-0507-28 -- --07-2710-0209-1508-06 --09-16
Peak daily increment 35 1065 87 785 318 1167 118 1628 2947 161 323 68 1078 101 121 78 138 7 9 1 46 175 66 10 19 112 12 36 106 56 38 23 40 31 28 40 25 1566 9 18 26 92 5 137 45 386 6 68 13 5
Days since peak 44 88 15 92 55 32 28 12 86 33 87 23 74 31 32 88 72 42 96 38 116 71 56 51 38 158 63 69 58 171 38 142 8 52 30 150 141 114 157 65 47 49 81 165 81 82 15 32 72 31
Last total 904 153675 9782 13588 28803 112998 12431 30123 86059 33648 6603 23857 18408 9224 219286 2788 1684 5823 16956 2176 4542 641 664 15917 7423 1528 529 9192 3685 859 1312 5527 9716 4028 7258 2217 2419 3171 239 3929 399 546 466 16204 929 1707 33325 5067 1162 619 8442 3637 315 2866 17440 438 3419 2235 1574 401
Last daily increment 0 461 1 59 187 0 84 253 355 71 72 277 38 71 711 2 19 17 46 4 0 0 2 87 47 4 6 27 31 2 12 0 21 7 23 5 101 11 4 19 11 0 1 2 1 0 6 13 8 2 8 22 8 29 65 2 14 1 0 3
Last week 6 3187 116 270 969 3848 587 1579 2278 343 282 1386 628 387 4518 124 115 65 377 63 12 4 10 553 181 66 22 208 123 85 60 65 119 36 121 76 163 70 29 159 62 28 10 30 18 46 49 68 64 20 112 85 29 134 469 16 64 48 109 16
Previous peak date -- --05-06 -- --06-16 --03-20 -- --07-12 -- --04-1504-13 -- -- -- --04-2404-2004-29 -- --04-0705-21 -- --05-02 --04-1404-1404-25 --04-14 --04-30 -- -- -- -- -- --04-22 --04-1205-1804-2904-16 -- --04-29 -- -- -- --04-2204-0604-1004-23
Previous peak daily increment 168 1662 150 92 125 2223 72 105 11 44 14 99 10 64 459 137 17 317 10 3573 76 7 24 104 25 12 16
Low between peaks 2 388 20 16 532 2 3 1 12 3 5 12 -1 6 5 52 3 7 15 6 9 -17 0

Deaths count forecast Europe (bold red line in graphs) 2020-10-18 to 2020-10-24

DateUKEUATBEBGCZDEESFRGRHRHUIEITNLPLPTROSECH
2020-10-17 43579 156378 889 10392 968 1352 9785 33775 33054 500 355 1109 1849 36474 6737 3524 2162 5812 5918 2122
2020-10-18 43700 156400 894 10430 972 1406 9813 33890 33050 510 359 1131 1851 36520 6737 3602 2173 5882 5920 2123
2020-10-19 43800 156600 899 10460 980 1492 9838 34000 33160 519 363 1153 1853 36560 6750 3679 2183 5951 5922 2124
2020-10-20 43920 157000 904 10490 988 1573 9862 34110 33350 529 366 1176 1855 36610 6781 3758 2193 6020 5924 2125
2020-10-21 44050 157300 909 10520 996 1665 9885 34220 33420 538 372 1198 1857 36650 6808 3835 2203 6089 5925 2126
2020-10-22 44160 157400 913 10550 1004 1762 9902 34320 33480 547 379 1221 1859 36690 6828 3914 2212 6158 5927 2128
2020-10-23 44280 157800 918 10590 1013 1847 9923 34420 33630 557 382 1245 1861 36730 6844 3995 2222 6229 5928 2129
2020-10-24 44390 158200 923 10620 1021 1925 9938 34520 33640 567 390 1269 1863 36770 6868 4078 2231 6300 5930 2130

Deaths count forecast (bold red line in graphs) 2020-10-18 to 2020-10-24

DateBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-ARUS-AZUS-CAUS-COUS-FLUS-GAUS-IAUS-IDUS-ILUS-INUS-KSUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-MTUS-NCUS-NDUS-NEUS-NJUS-NVUS-NYUS-OHUS-OKUS-ORUS-PAUS-SCUS-SDUS-TNUS-TXUS-VAUS-WAUS-WI
2020-10-17 153675 9782 13588 28803 112998 12431 30123 86059 33648 6603 23857 18408 9224 219286 2788 1684 5823 16956 2176 15917 7423 1528 529 9192 3685 859 1312 5527 9716 4028 7258 2217 2419 3171 239 3929 399 546 16204 1707 33325 5067 1162 619 8442 3637 315 2866 17440 3419 2235 1574
2020-10-18 154400 9801 13660 28960 113300 12530 30310 86060 33730 6665 24010 18500 9284 219800 2788 1703 5838 17030 2182 16040 7440 1532 529 9199 3685 859 1312 5549 9732 4033 7271 2217 2425 3185 241 3943 407 546 16210 1709 33330 5071 1162 622 8451 3650 316 2895 17520 3431 2235 1574
2020-10-19 154600 9820 13710 29120 113700 12620 30500 86200 33810 6726 24170 18590 9342 220100 2791 1721 5838 17040 2187 16070 7445 1549 533 9213 3694 867 1316 5561 9747 4038 7284 2217 2429 3185 244 3944 417 549 16210 1710 33340 5071 1167 623 8465 3655 318 2904 17560 3442 2238 1584
2020-10-20 155100 9838 13740 29270 114100 12710 30670 86570 33880 6787 24440 18680 9400 220900 2801 1739 5846 17100 2193 16170 7464 1560 536 9243 3721 874 1327 5568 9763 4044 7298 2218 2437 3201 246 3983 426 554 16220 1719 33350 5081 1182 629 8478 3671 319 2927 17640 3454 2257 1611
2020-10-21 155800 9857 13760 29430 114500 12800 30830 86880 33960 6847 24690 18770 9456 221800 2834 1757 5855 17190 2199 16260 7481 1571 541 9285 3735 884 1333 5576 9778 4049 7311 2229 2515 3224 249 4016 436 556 16230 1732 33360 5101 1193 632 8506 3690 323 2954 17750 3465 2265 1634
2020-10-22 156500 9875 13800 29580 114900 12890 30970 87100 34030 6908 24970 18860 9513 222600 2873 1774 5865 17280 2204 16390 7505 1581 542 9325 3753 896 1349 5585 9794 4054 7324 2231 2515 3235 252 4039 446 561 16240 1739 33370 5107 1202 636 8524 3703 336 2995 17850 3476 2273 1646
2020-10-23 157100 9894 13880 29740 115300 12990 31090 87360 34110 6969 25220 18950 9569 223400 2897 1792 5877 17340 2210 16490 7553 1591 547 9363 3773 914 1355 5605 9809 4059 7337 2239 2515 3242 255 4070 456 569 16250 1747 33380 5120 1211 641 8543 3710 339 3010 17930 3488 2281 1665
2020-10-24 157600 9912 13930 29890 115700 13080 31190 87780 34180 7030 25460 19040 9626 224100 2899 1810 5890 17380 2215 16550 7597 1595 552 9384 3796 918 1363 5605 9824 4064 7350 2247 2584 3255 258 4087 466 570 16250 1749 33380 5136 1217 643 8556 3735 348 3039 17990 3499 2281 1670

Deaths count average forecast Europe (bold black line in graphs) 2020-10-18 to 2020-10-24

DateUKEUATBEBGCZDEESFRGRHRHUIEITNLPLPTROSECH
2020-10-17 43579 156378 889 10392 968 1352 9785 33775 33054 500 355 1109 1849 36474 6737 3524 2162 5812 5918 2122
2020-10-18 43700 156900 894 10420 976 1412 9800 33880 33120 511 362 1135 1853 36520 6756 3616 2175 5874 5920 2125
2020-10-19 43790 157400 899 10450 989 1479 9822 33990 33200 519 365 1157 1855 36570 6767 3693 2186 5932 5921 2128
2020-10-20 43890 158000 904 10470 998 1546 9849 34100 33280 526 369 1184 1857 36610 6796 3771 2197 5996 5923 2132
2020-10-21 43990 158500 910 10500 1007 1619 9873 34210 33350 532 373 1212 1859 36650 6824 3851 2207 6059 5925 2135
2020-10-22 44080 159000 915 10520 1016 1692 9890 34320 33420 539 380 1241 1861 36690 6842 3932 2218 6120 5926 2138
2020-10-23 44190 159500 920 10540 1026 1767 9908 34430 33500 543 383 1270 1863 36730 6855 4016 2229 6179 5928 2141
2020-10-24 44290 160000 926 10570 1033 1839 9931 34540 33560 550 388 1298 1865 36770 6875 4101 2240 6242 5929 2144

Deaths count average forecast (bold black line in graphs) 2020-10-18 to 2020-10-24

DateBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-ARUS-AZUS-CAUS-COUS-FLUS-GAUS-IAUS-IDUS-ILUS-INUS-KSUS-KYUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-MTUS-NCUS-NDUS-NEUS-NJUS-NVUS-NYUS-OHUS-OKUS-ORUS-PAUS-SCUS-SDUS-TNUS-TXUS-VAUS-WAUS-WI
2020-10-17 153675 9782 13588 28803 112998 12431 30123 86059 33648 6603 23857 18408 9224 219286 2788 1684 5823 16956 2176 15917 7423 1528 529 9192 3685 859 1312 5527 9716 4028 7258 2217 2419 3171 239 3929 399 546 16204 1707 33325 5067 1162 619 8442 3637 315 2866 17440 3419 2235 1574
2020-10-18 154000 9795 13640 28960 113400 12510 30340 86210 33710 6662 24080 18490 9288 219600 2800 1702 5832 16960 2181 16030 7448 1534 532 9209 3696 869 1319 5543 9736 4033 7277 2227 2437 3174 242 3935 407 549 16210 1710 33340 5070 1167 620 8452 3642 318 2882 17460 3431 2236 1580
2020-10-19 154300 9812 13690 29100 114100 12600 30500 86500 33790 6720 24270 18590 9343 219800 2810 1721 5837 16970 2187 16080 7463 1548 535 9225 3704 882 1324 5555 9750 4039 7290 2233 2449 3175 245 3941 416 552 16210 1713 33350 5071 1173 622 8467 3651 322 2899 17500 3441 2237 1587
2020-10-20 154800 9830 13720 29250 114700 12700 30640 86970 33860 6780 24470 18700 9398 220400 2822 1740 5843 17030 2196 16170 7487 1560 538 9253 3731 895 1333 5564 9764 4044 7303 2240 2465 3189 249 3976 426 556 16210 1721 33350 5078 1186 627 8482 3666 325 2923 17570 3450 2251 1614
2020-10-21 155400 9848 13760 29410 115300 12800 30780 87380 33930 6844 24670 18800 9453 221000 2847 1759 5851 17110 2202 16260 7508 1572 542 9292 3746 910 1340 5574 9779 4049 7316 2259 2518 3209 255 4004 439 558 16220 1732 33360 5093 1196 630 8506 3682 330 2949 17680 3459 2257 1638
2020-10-22 156000 9864 13800 29530 115900 12900 30900 87960 34020 6908 24870 18900 9508 221500 2871 1778 5859 17200 2208 16390 7537 1582 546 9327 3762 924 1351 5584 9793 4054 7330 2270 2531 3223 259 4028 448 563 16220 1737 33370 5101 1205 635 8523 3697 339 2984 17780 3469 2263 1651
2020-10-23 156600 9887 13860 29720 116500 12990 31010 88570 34090 6975 25070 19000 9564 222100 2886 1798 5870 17270 2215 16490 7577 1594 549 9358 3780 941 1357 5601 9807 4059 7343 2281 2554 3232 264 4054 462 567 16230 1744 33380 5121 1213 639 8541 3709 344 3008 17880 3478 2268 1662
2020-10-24 157200 9906 13910 29880 117100 13070 31090 88980 34170 7043 25280 19110 9619 222500 2896 1817 5881 17340 2221 16540 7612 1601 553 9385 3798 953 1364 5607 9821 4064 7356 2293 2589 3247 267 4076 476 570 16230 1748 33390 5127 1218 642 8561 3730 351 3038 17950 3487 2268 1674

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.

Recent changes and notes

[2020-10-11]Short-term forecasting of the coronavirus pandemic (with Jennie Castle and David Hendry) is now in press at the International Journal of Forecasting.
[2020-10-10]Temporarily removed forecasts from the Chinese scenarios, while investigating possibility to use own history from the first wave.
Added information on the previous peak (if present) to the peak tables.
Local forecasts for England: now dropping last four observations.
[2020-07-01] Modified the short-term model to allow for (slowly changing) seasonality. Many countries show clear seasonality after the initial period, likely caused by institutional factors regarding data collection. This seasonality was also getting in the way of peak detection. As a consequence estimates of the peak date may have changed for countries with strong seasonality.
Added forecasts of cumulative confirmed cases for lower tier local authorities of England. The data is available from 2020-07-02 including all tests (pillar one and two). Only authorities with more than 5 cases in the previous week are included.
[2020-06-29] Tables in April included the world, but not the world as we know it (double counting China and the US). So removed the world from those old tables.
Why short-term forecasts can be better than models for predicting how pandemics evolve just appeared at The Conversation.
Thursday 2 July webinar at the FGV EESP - São Paolo School of Economics. This starts at 16:00 UK time (UTC+01:00) and streamed here.
[2020-06-24] Research presentation on short-term COVID-19 forecasting on 26 June (14:00 UK time) at the Quarterly Forecasting Forum of the IIF UK Chapter.
[2020-06-06] Removed Brazil from yesterday's forecasts (only; last observation 2020-06-05).
[2020-06-04] Data issues with confirmed cases for France.
Added an appendix to the short term paper with further forecast comparisons for European and Latin American countries.
Both Sweden and Iran have lost their peak in confirmed cases. For Sweden the previous peak was on 24 April (daily peak of 656 cases), for Iran it was on 31 March (peak of 3116). For Iran this looks like a second wave, with increasing daily counts for the last four weeks. For Sweden this is a sudden jump in confirmed cases in the last two days, compared to a fairly steady weekly pattern over the previous six weeks.
[2020-05-20] Problem with UK confirmed cases: negative daily count. This makes the forecasts temporarily unreliable.
Updated the second paper.
[2020-05-18] Minor fixes to the improved version of scenario forecasting, backported to 2020-05-13.
[2020-05-13] We now omit countries with fewer than 200 confirmed cases in the last week (25 for deaths).
The short-term paper has some small updates, including further comparisons with other models.
Data for Ecuador are not reliable enough for forecasting.
Switched to an improved version of scenario forecasting.
[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.
[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.
[2020-04-27] 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.
[2020-04-24] A summary of our work on short-term COVID-19 forecasting appeared as a voxeu.
[2020-04-17] Bird and Nielsen look into nowcasting death counts in England.
[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).
[2020-04-10] Updated documentation with better description of short-term estimates and peak determination.
[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.
[2020-04-08] Minor correction to peak estimates. Added table with scenario forecasts.
[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.
[2020-04-02] Now including more US States, based on New York Times data.
[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).
[2020-03-26] Scenario forecasts that are based on what happened in China earlier this year, only for Italy.
[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.

Initial visual evaluation of forecasts of Deaths