COVID-19 short-term forecasts Deaths 2020-05-27


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 The documentation that is provided is still in progress and not peer reviewed.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.
[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.
[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.
[2020-05-18] Minor fixes to the improved version of scenario forecasting, backported to 2020-05-13.
[2020-05-20] Problem with UK confirmed cases: negative daily count. This makes the forecasts temporarily unreliable.
Updated the second paper.

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) 2020-05-28 to 2020-06-03

DateUKEUBEDEFRHUIEITNLPLPTROSE
2020-05-27 37460 125037 9364 8428 28546 505 1631 33072 5871 1028 1356 1227 4220
2020-05-28 37600 125000 9390 8450 28600 510 1640 33200 5890 1030 1370 1230 4240
2020-05-29 37800 125000 9420 8480 28700 510 1640 33300 5900 1040 1380 1240 4260
2020-05-30 38100 125000 9460 8510 28700 510 1650 33400 5920 1050 1390 1250 4290
2020-05-31 38300 126000 9490 8550 28800 520 1650 33500 5940 1060 1410 1260 4330
2020-06-01 38500 126000 9520 8590 28900 520 1660 33500 5950 1070 1420 1280 4380
2020-06-02 38700 126000 9550 8620 28900 530 1670 33600 5970 1080 1430 1290 4420
2020-06-03 39000 126000 9580 8660 29000 530 1670 33700 5990 1090 1440 1300 4470

Deaths count average forecast (bold black line in graphs) 2020-05-28 to 2020-06-03

DateBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NMUS-NYUS-OHUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
2020-05-27 25598 6876 841 803 4534 1473 7564 8597 3983 904 3968 552 4431 100418 583 832 3895 1392 3802 445 344 2319 1904 496 5082 1871 2618 6541 2325 5266 932 670 670 844 11339 329 29484 2044 5265 466 349 1581 1281 1095 539
2020-05-28 26700 6920 880 830 4720 1480 7600 8800 4070 910 4130 570 4450 101000 590 840 3910 1400 3830 450 350 2330 1910 500 5120 1890 2630 6590 2340 5280 940 670 680 850 11400 330 29600 2060 5300 470 350 1590 1290 1100 540
2020-05-29 28000 7010 930 860 4920 1500 7650 9200 4220 910 4300 600 4480 102000 600 850 3960 1400 3860 450 350 2350 1930 510 5160 1900 2660 6650 2380 5310 950 680 690 850 11500 340 29700 2090 5350 470 360 1610 1310 1100 540
2020-05-30 29200 7100 980 890 5130 1520 7700 9600 4390 910 4490 640 4510 103000 600 860 4000 1410 3900 460 350 2370 1940 530 5230 1920 2680 6710 2410 5340 970 680 700 860 11600 350 29800 2120 5400 480 360 1630 1340 1110 550
2020-05-31 30600 7190 1030 920 5340 1550 7750 10100 4560 920 4680 680 4530 103000 610 870 4040 1420 3930 460 350 2390 1970 540 5290 1930 2700 6770 2440 5370 990 690 710 870 11700 350 29900 2150 5450 480 370 1650 1360 1120 550
2020-06-01 32000 7290 1080 950 5570 1570 7800 10500 4740 930 4880 720 4560 104000 620 880 4080 1430 3970 470 360 2410 1990 550 5360 1950 2720 6840 2480 5400 1010 690 730 880 11900 360 30000 2180 5510 490 370 1670 1390 1120 560
2020-06-02 33400 7380 1140 990 5800 1600 7850 11000 4930 930 5090 760 4590 105000 630 890 4120 1430 4000 470 360 2440 2010 560 5430 1970 2750 6900 2510 5430 1020 700 740 890 12000 370 30100 2220 5560 500 380 1690 1420 1130 570
2020-06-03 35000 7480 1200 1020 6050 1630 7900 11500 5130 940 5310 800 4620 106000 640 900 4170 1440 4040 480 370 2460 2040 580 5500 1980 2770 6960 2550 5470 1040 700 750 910 12100 370 30300 2250 5620 510 380 1710 1450 1140 570

Deaths count forecast Europe (bold red line in graphs) 2020-05-28 to 2020-06-03

DateUKEUBEDEFRHUIEITNLPLPTROSE
2020-05-27 37460 125037 9364 8428 28546 505 1631 33072 5871 1028 1356 1227 4220
2020-05-28 37700 125000 9400 8470 28600 510 1640 33200 5890 1040 1370 1240 4270
2020-05-29 38000 125000 9420 8520 28700 510 1650 33300 5910 1050 1380 1250 4330
2020-05-30 38200 125000 9450 8560 28800 520 1660 33400 5920 1060 1390 1260 4380
2020-05-31 38500 125000 9480 8600 28900 520 1670 33500 5940 1060 1410 1270 4440
2020-06-01 38700 125000 9510 8640 28900 530 1680 33600 5960 1070 1420 1280 4490
2020-06-02 39000 126000 9540 8690 29000 530 1680 33700 5980 1080 1430 1290 4550
2020-06-03 39300 126000 9570 8730 29100 540 1690 33800 5990 1090 1440 1300 4610

Deaths count forecast (bold red line in graphs) 2020-05-28 to 2020-06-03

DateBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NMUS-NYUS-OHUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
2020-05-27 25598 6876 841 803 4534 1473 7564 8597 3983 904 3968 552 4431 100418 583 832 3895 1392 3802 445 344 2319 1904 496 5082 1871 2618 6541 2325 5266 932 670 670 844 11339 329 29484 2044 5265 466 349 1581 1281 1095 539
2020-05-28 26800 6960 880 830 4700 1500 7610 9000 4110 910 4120 590 4460 101000 590 840 3950 1410 3830 450 350 2340 1930 510 5160 1890 2640 6600 2350 5300 950 670 680 850 11500 330 29600 2070 5330 470 360 1600 1310 1100 540
2020-05-29 28000 7050 920 860 4890 1530 7660 9400 4280 920 4290 620 4490 102000 590 850 3990 1430 3870 450 350 2370 1950 520 5240 1900 2660 6660 2390 5340 970 680 700 870 11600 340 29800 2100 5380 480 360 1620 1340 1110 550
2020-05-30 29300 7150 960 890 5090 1560 7710 9800 4450 930 4470 660 4520 103000 600 860 4040 1450 3900 460 360 2400 1980 540 5310 1910 2680 6710 2420 5370 990 680 710 890 11700 340 29900 2120 5430 490 370 1640 1370 1120 560
2020-05-31 30700 7240 1010 920 5290 1590 7760 10300 4620 930 4650 690 4550 104000 610 870 4080 1470 3930 460 360 2430 2010 550 5380 1920 2690 6770 2460 5410 1010 680 720 910 11900 350 30000 2150 5470 500 380 1660 1400 1130 570
2020-06-01 32100 7340 1050 960 5510 1620 7810 10800 4800 940 4840 730 4570 105000 610 890 4130 1490 3960 470 360 2450 2040 570 5460 1940 2710 6830 2490 5450 1030 690 740 930 12100 350 30200 2180 5520 510 380 1680 1430 1140 580
2020-06-02 33500 7440 1100 990 5730 1650 7860 11300 4990 950 5040 770 4600 106000 620 900 4180 1510 3990 470 370 2480 2060 590 5540 1950 2730 6880 2520 5480 1050 690 750 950 12200 360 30300 2200 5560 520 390 1700 1470 1150 590
2020-06-03 35100 7540 1150 1030 5960 1680 7910 11800 5190 960 5250 810 4630 107000 630 910 4220 1530 4020 470 370 2510 2090 600 5620 1960 2750 6940 2560 5520 1070 690 770 970 12400 360 30400 2230 5610 530 400 1720 1500 1160 600

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

DateUKEUBEDEFRHUIEITNLPLPTROSEBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NMUS-NYUS-OHUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
2020-05-27 37460 125037 9364 8428 28546 505 1631 33072 5871 1028 1356 1227 4220 25598 6876 841 803 4534 1473 7564 8597 3983 904 3968 552 4431 100418 583 832 3895 1392 3802 445 344 2319 1904 496 5082 1871 2618 6541 2325 5266 932 670 670 844 11339 329 29484 2044 5265 466 349 1581 1281 1095 539
2020-05-28 37600 125000 9390 8450 28600 510 1630 33200 5890 1040 1370 1240 4240 26500 6940 880 830 4710 1470 7610 8900 4080 910 4120 600 4460 101000 590 840 3910 1390 3830 450 350 2330 1910 510 5100 1880 2630 6590 2350 5280 940 680 680 850 11400 330 29500 2060 5310 470 350 1590 1290 1100 540
2020-05-29 37800 125000 9420 8490 28700 510 1640 33200 5900 1050 1370 1240 4290 27400 7030 920 850 4890 1490 7650 9200 4210 920 4280 640 4480 101000 600 850 3930 1400 3860 450 350 2350 1920 520 5150 1900 2640 6650 2380 5290 960 680 690 860 11500 340 29600 2090 5350 470 360 1610 1310 1100 540
2020-05-30 37900 125000 9440 8520 28800 510 1640 33300 5910 1060 1380 1250 4330 28200 7090 950 880 5070 1500 7690 9500 4350 920 4440 690 4500 102000 600 850 3970 1410 3890 460 350 2370 1940 540 5200 1910 2660 6690 2410 5300 970 680 700 880 11500 340 29700 2100 5390 480 360 1630 1330 1110 540
2020-05-31 38100 126000 9460 8550 28800 520 1650 33400 5930 1060 1390 1260 4360 29000 7160 980 900 5250 1500 7730 9800 4490 930 4610 730 4530 103000 610 860 3990 1430 3910 460 350 2380 1950 560 5250 1930 2670 6740 2430 5330 980 690 720 900 11600 350 29800 2120 5420 480 370 1640 1340 1110 550
2020-06-01 38200 126000 9490 8580 28900 520 1650 33500 5940 1070 1400 1270 4390 29800 7210 1020 920 5430 1510 7770 10000 4620 930 4760 770 4550 103000 610 870 4020 1440 3940 460 360 2390 1950 580 5290 1940 2680 6790 2450 5350 990 690 730 910 11600 350 29900 2140 5470 480 370 1660 1350 1120 550
2020-06-02 38400 126000 9510 8600 28900 520 1650 33600 5960 1080 1400 1280 4400 30500 7260 1060 940 5610 1510 7810 10300 4740 940 4920 820 4570 104000 610 880 4050 1450 3970 470 360 2410 1960 600 5330 1950 2700 6830 2470 5370 1000 700 740 930 11700 350 30000 2160 5520 490 380 1670 1360 1120 550
2020-06-03 38500 126000 9530 8630 29000 520 1660 33600 5970 1080 1410 1290 4420 31200 7300 1090 960 5790 1510 7840 10500 4870 950 5110 870 4590 105000 620 880 4080 1460 3990 470 360 2420 1970 620 5370 1970 2710 6870 2500 5390 1020 700 750 940 11800 350 30000 2170 5550 490 380 1680 1370 1130 550
2020-06-04 38600 127000 9550 8650 29100 530 1660 33700 5980 1090 1410 1290 4440 32000 7340 1120 980 5960 1520 7880 10800 4990 950 5250 910 4610 105000 620 890 4100 1460 4020 480 360 2440 1980 630 5400 1980 2720 6910 2520 5410 1030 710 760 960 11800 360 30100 2190 5590 490 390 1700 1380 1130 560
2020-06-05 38800 127000 9570 8670 29100 530 1660 33800 5990 1090 1420 1300 4440 32700 7380 1160 1000 6130 1530 7910 11100 5140 950 5390 970 4620 106000 630 900 4120 1470 4040 480 360 2450 1990 650 5440 1990 2730 6940 2540 5430 1040 710 770 970 11900 360 30200 2210 5630 500 390 1710 1390 1140 560

Peak increase in estimated trend of Deaths in Europe 2020-05-27

UKEUBEDEFRHUIEITNLPLPTROSE
Peak date04-0904-0704-1004-1504-0704-1804-2403-2804-0704-2404-1105-0104-22
Peak daily increment 1011 3233 354 270 1017 14 95 832 160 29 33 25 118
Days from 100 to peak 22 34 17 23 23 5 22 24 18 18 13 29 25
Days from peak/2 to peak 16 24 15 22 16 25 25 19 18 28 19 37 27
Days since peak 48 50 47 42 50 39 33 60 50 33 46 26 35

Peak increase in estimated trend of Deaths 2020-05-27

BrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-IAUS-ILUS-INUS-LAUS-MAUS-MDUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NMUS-NYUS-OHUS-PAUS-SCUS-TNUS-TXUS-VAUS-WAUS-WI
Peak date --05-05 -- -- --04-1403-19 -- --04-12 -- --04-1904-1505-0605-0804-2304-2404-2004-2505-1505-0704-20 --05-1304-2904-1404-2504-2904-2205-0605-0805-06 --04-3005-1404-0704-2905-0504-2904-0305-1404-2204-0604-06
Peak daily increment 178 38 148 29 125 2270 17 38 88 52 116 12 11 54 45 128 50 76 245 66 146 25 20 23 315 11 1072 56 170 20 10 40 53 31 12
Days from 100 to peak 34 16 14 11 22 30 23 26 26 21 18 5 21 36 20 43 25 18 24 22 25 19 27 23 29 17 18 25 32 14 -8 40 7 14 -2
Days from peak/2 to peak 35 26 16 25 24 21 37 36 30 28 20 26 46 42 25 44 32 21 23 26 26 33 41 38 29 42 13 31 33 33 5 48 13 30 13
Days since peak 22 43 69 45 38 42 21 19 34 33 37 32 12 20 37 14 28 43 32 28 35 21 19 21 27 13 50 28 22 28 54 13 35 51 51

Initial visual evaluation of forecasts of Deaths