COVID-19 short-term forecasts

Peak increase in estimated trend of Deaths in Europe 2020-07-22

	UK	EU	AT	BE	BG	BS	CZ	DE	DK	ES	FI	FR	GR	HU	IE	IT	NL	PL	PT	RO	SE	NO	CH
Peak date	04-09	04-04	04-06	04-15	--	04-11	04-08	04-15	04-04	03-31	04-21	04-08	04-03	04-19	04-24	03-29	04-04	04-25	04-15	05-03	04-21	04-07	04-06
Peak daily increment	984	3121	20	322		7	10	226	15	945	31	944	5	13	167	781	153	23	31	25	94	8	60
Days from 100 to peak	22	31	7	22		0	0	22	3	18	1	23	-11	6	21	25	15	19	17	31	24	-2	14
Days from peak/2 to peak	16	21	16	21		22	13	22	15	14	20	18	18	26	22	20	16	31	23	39	27	20	19
Last total	45501	135204	711	9808	321	260	364	9102	611	28426	328	30063	200	596	1754	35082	6139	1642	1702	2101	5667	255	1972
Last daily increment	79	106	1	3	8	0	4	3	0	2	0	7	3	0	1	9	3	6	5	27	21	0	0
Last week	382	458	0	13	28	1	9	15	1	10	0	28	7	1	5	65	2	37	23	130	74	1	3
Days since peak	104	109	107	98		102	105	98	109	113	92	105	110	94	89	115	109	88	98	80	92	106	107

Peak increase in estimated trend of Deaths 2020-07-22

	Brazil	Canada	Chile	Colombia	India	Indonesia	Iran	Mexico	Peru	Philippines	Russia	South Africa	Turkey	US	US-AL	US-AR	US-AZ	US-CA	US-CO	US-CT	US-DC	US-DE	US-FL	US-GA	US-IA	US-IL	US-IN	US-KS	US-KY	US-LA	US-MA	US-MD	US-MI	US-MN	US-MO	US-MS	US-NC	US-NE	US-NH	US-NJ	US-NM	US-NV	US-NY	US-OH	US-OK	US-OR	US-PA	US-SC	US-TN	US-TX	US-UT	US-VA	US-WA	US-WI
Peak date	07-17	05-06	07-17	--	06-16	--	--	06-22	06-14	07-12	05-29	--	04-15	04-14	--	--	--	--	04-24	04-20	04-29	06-23	--	--	05-21	05-12	05-01	04-14	04-14	04-14	04-25	04-29	04-14	05-29	04-30	--	06-22	05-20	05-29	06-25	05-11	--	04-09	04-29	--	--	05-05	--	--	--	04-17	04-22	04-06	04-11
Peak daily increment	1067	168	844		1662			695	279	104	171		120	2161					71	106	11	46			14	112	42	6	10	64	453	56	137	23	17		24	41	28	1543	9		999	76			137				6	105	25	12
Days from 100 to peak	111	36	93		72			77	68	102	48		18	29					21	18	10	60			27	41	27	-7	0	17	23	22	17	42	21		70	0	16	85	14		20	25			32				-45	7	15	2
Days from peak/2 to peak	102	37	61		52			70	65	107	49		20	20					27	20	30	74			50	44	35	19	19	21	20	27	18	57	35		82	30	31	81	39		15	30			33				18	10	30	18
Last total	82771	8913	8677	7373	29861	4459	14853	41190	13767	1843	12726	5940	5545	143190	1325	380	2974	8047	1771	4406	580	518	5345	3256	814	7347	2666	320	677	3558	8461	3393	6315	1552	1117	1423	1732	309	402	15707	591	704	32558	3235	474	271	7077	1285	876	4439	209	2051	1467	865
Last daily increment	1284	5	0	207	1129	139	219	790	188	6	165	572	19	1195	57	6	56	159	8	0	0	2	139	77	8	23	14	5	3	60	18	7	6	4	10	34	27	1	2	-30	3	28	38	16	13	2	26	64	16	240	7	3	3	6
Last week	6083	38	1387	1344	4259	586	1245	3616	1152	200	806	1271	105	4832	125	39	483	558	26	17	6	6	668	213	33	96	64	16	27	183	89	59	40	26	33	115	125	12	7	42	29	78	112	132	36	22	93	215	90	782	16	44	40	34
Days since peak	5	77	5		36			30	38	10	54		98	99					89	93	84	29			62	71	82	99	99	99	88	84	99	54	83		30	63	54	27	72		104	84			78				96	91	107	102

Deaths count forecast Europe (bold red line in graphs) 2020-07-23 to 2020-07-29

Date	UK	EU	BG	FR	IT	PL	PT	RO	SE
2020-07-22	45501	135204	321	30063	35082	1642	1702	2101	5667
2020-07-23	45610	135400	326	30070	35090	1649	1707	2121	5691
2020-07-24	45690	135500	332	30080	35110	1655	1712	2140	5712
2020-07-25	45740	135500	338	30080	35120	1662	1717	2160	5712
2020-07-26	45760	135500	344	30090	35130	1668	1721	2179	5712
2020-07-27	45770	135600	351	30100	35140	1675	1726	2199	5726
2020-07-28	45870	135700	357	30100	35150	1681	1731	2218	5730
2020-07-29	45950	135800	364	30110	35160	1688	1736	2238	5752

Deaths count forecast (bold red line in graphs) 2020-07-23 to 2020-07-29

Date	Brazil	Canada	Chile	Colombia	India	Indonesia	Iran	Mexico	Peru	Philippines	Russia	South Africa	Turkey	US	US-AL	US-AR	US-AZ	US-CA	US-CO	US-FL	US-GA	US-IA	US-IL	US-IN	US-KY	US-LA	US-MA	US-MD	US-MI	US-MN	US-MO	US-MS	US-NC	US-NJ	US-NM	US-NV	US-NY	US-OH	US-OK	US-OR	US-PA	US-SC	US-TN	US-TX	US-VA	US-WA	US-WI
2020-07-22	82771	8913	8677	7373	29861	4459	14853	41190	13767	1843	12726	5940	5545	143190	1325	380	2974	8047	1771	5345	3256	814	7347	2666	677	3558	8461	3393	6315	1552	1117	1423	1732	15707	591	704	32558	3235	474	271	7077	1285	876	4439	2051	1467	865
2020-07-23	83890	8922	8737	7611	30800	4541	15170	41800	13940	1905	12930	6096	5563	143600	1362	388	3065	8268	1771	5564	3319	820	7374	2671	686	3614	8480	3403	6324	1560	1122	1445	1732	15710	595	720	32580	3238	484	278	7108	1326	889	4753	2086	1474	865
2020-07-24	85030	8931	8963	7887	31660	4666	15440	42470	14110	1933	13100	6282	5581	144400	1399	395	3158	8442	1780	5760	3387	827	7391	2680	693	3668	8500	3412	6333	1566	1127	1466	1748	15710	599	736	32600	3253	496	283	7129	1365	901	5043	2095	1481	865
2020-07-25	85980	8940	9057	8111	32510	4774	15700	43000	14280	1994	13240	6492	5598	145100	1435	404	3254	8581	1781	5902	3451	832	7406	2693	698	3700	8521	3421	6342	1569	1133	1488	1769	15710	604	752	32620	3267	506	287	7137	1416	913	5281	2102	1489	873
2020-07-26	86620	8949	9057	8379	33340	4924	16000	43220	14450	2012	13340	6715	5615	145500	1472	412	3354	8690	1782	6016	3491	838	7411	2695	700	3746	8543	3430	6350	1572	1138	1511	1774	15710	608	769	32640	3296	511	291	7143	1439	925	5524	2105	1496	873
2020-07-27	87260	8958	9122	8562	34170	5061	16290	43550	14620	2044	13430	6951	5632	145800	1510	421	3456	8813	1786	6163	3535	845	7416	2697	703	3781	8565	3439	6359	1575	1143	1534	1791	15830	611	785	32660	3308	517	294	7147	1459	938	5757	2109	1503	874
2020-07-28	88550	8967	9156	8782	35000	5189	16580	44440	14790	2052	13580	7199	5650	146700	1549	430	3563	9037	1794	6363	3608	854	7437	2716	707	3830	8587	3448	6367	1578	1148	1558	1821	15880	615	802	32680	3335	526	302	7170	1522	950	6020	2122	1510	885
2020-07-29	89760	8977	9201	9057	35840	5345	16850	45170	14960	2057	13740	7456	5667	147800	1589	439	3673	9296	1802	6539	3655	860	7457	2728	711	3865	8610	3457	6375	1583	1153	1582	1842	15880	619	819	32700	3348	534	306	7192	1585	963	6300	2132	1517	889

Deaths count average forecast Europe (bold black line in graphs) 2020-07-23 to 2020-07-29

Date	UK	EU	BG	FR	IT	PL	PT	RO	SE
2020-07-22	45501	135204	321	30063	35082	1642	1702	2101	5667
2020-07-23	45550	135300	324	30070	35090	1648	1706	2110	5680
2020-07-24	45620	135400	330	30080	35100	1654	1710	2128	5699
2020-07-25	45680	135400	334	30090	35110	1660	1715	2148	5699
2020-07-26	45710	135500	339	30110	35120	1665	1720	2168	5700
2020-07-27	45760	135600	346	30120	35130	1671	1724	2188	5710
2020-07-28	45850	135600	353	30130	35150	1679	1729	2208	5716
2020-07-29	45930	135700	361	30140	35160	1686	1733	2229	5736

Deaths count average forecast (bold black line in graphs) 2020-07-23 to 2020-07-29

Date	Brazil	Canada	Chile	Colombia	India	Indonesia	Iran	Mexico	Peru	Philippines	Russia	South Africa	Turkey	US	US-AL	US-AR	US-AZ	US-CA	US-CO	US-FL	US-GA	US-IA	US-IL	US-IN	US-KY	US-LA	US-MA	US-MD	US-MI	US-MN	US-MO	US-MS	US-NC	US-NJ	US-NM	US-NV	US-NY	US-OH	US-OK	US-OR	US-PA	US-SC	US-TN	US-TX	US-VA	US-WA	US-WI
2020-07-22	82771	8913	8677	7373	29861	4459	14853	41190	13767	1843	12726	5940	5545	143190	1325	380	2974	8047	1771	5345	3256	814	7347	2666	677	3558	8461	3393	6315	1552	1117	1423	1732	15707	591	704	32558	3235	474	271	7077	1285	876	4439	2051	1467	865
2020-07-23	83750	8920	8743	7621	30120	4529	15110	41720	13940	1865	12860	6018	5562	143500	1344	385	3048	8213	1772	5522	3268	817	7363	2669	681	3568	8474	3399	6323	1559	1119	1445	1741	15720	593	710	32570	3240	476	275	7097	1307	886	4570	2062	1470	865
2020-07-24	84850	8928	8926	7933	30770	4640	15370	42400	14130	1904	13010	6120	5579	144200	1374	393	3143	8377	1777	5693	3283	820	7379	2676	688	3578	8490	3408	6332	1564	1124	1474	1752	15760	596	719	32590	3256	479	280	7114	1338	901	4729	2071	1475	866
2020-07-25	85800	8935	9050	8229	31430	4738	15640	42930	14330	1947	13140	6286	5597	144800	1404	400	3241	8518	1780	5831	3299	824	7395	2690	693	3587	8507	3417	6341	1568	1129	1499	1769	15780	600	733	32600	3273	482	285	7124	1383	917	4884	2079	1480	871
2020-07-26	86530	8941	9117	8532	32100	4880	15920	43110	14530	1986	13260	6528	5614	145200	1432	408	3342	8623	1782	5961	3311	828	7403	2692	697	3603	8525	3425	6350	1571	1133	1521	1776	15810	604	742	32620	3305	484	289	7132	1410	930	5018	2086	1484	872
2020-07-27	87290	8949	9223	8795	32790	5003	16210	43460	14730	2027	13390	6783	5632	145400	1462	416	3447	8751	1785	6088	3325	832	7413	2694	700	3623	8543	3434	6359	1573	1138	1533	1791	15840	609	752	32640	3318	486	294	7142	1428	942	5148	2092	1488	873
2020-07-28	88520	8957	9324	9073	33480	5115	16500	44360	14940	2066	13550	7052	5650	146000	1494	424	3555	8972	1791	6271	3351	838	7434	2710	705	3644	8562	3443	6367	1577	1144	1574	1816	15870	613	774	32660	3342	491	300	7164	1474	956	5293	2106	1493	882
2020-07-29	89680	8966	9436	9380	34190	5243	16780	45050	15150	2106	13700	7333	5667	146700	1530	432	3667	9213	1797	6442	3371	844	7450	2720	712	3664	8580	3452	6376	1582	1149	1608	1835	15890	617	790	32670	3354	495	305	7185	1512	973	5479	2118	1498	884

Deaths count scenario forecast (bold purple line in graphs) 2020-07-23 to 2020-07-31

Date	UK	EU	BG	FR	IT	PL	PT	RO	SE	Brazil	Canada	Chile	Colombia	India	Indonesia	Iran	Mexico	Peru	Philippines	Russia	South Africa	Turkey	US	US-AL	US-AR	US-AZ	US-CA	US-CO	US-FL	US-GA	US-IA	US-IL	US-IN	US-KY	US-LA	US-MA	US-MD	US-MI	US-MN	US-MO	US-MS	US-NC	US-NJ	US-NM	US-NV	US-NY	US-OH	US-OK	US-OR	US-PA	US-SC	US-TN	US-TX	US-VA	US-WA	US-WI
2020-07-22	45501	135204	321	30063	35082	1642	1702	2101	5667	82771	8913	8677	7373	29861	4459	14853	41190	13767	1843	12726	5940	5545	143190	1325	380	2974	8047	1771	5345	3256	814	7347	2666	677	3558	8461	3393	6315	1552	1117	1423	1732	15707	591	704	32558	3235	474	271	7077	1285	876	4439	2051	1467	865
2020-07-23	45550	135300	323	30100	35100	1649	1707	2110	5676	83410	8925	8988	7581	30090	4523	15050	41550	13890	1893	12830	6035	5564	143700	1346	384	3067	8135	1772	5451	3266	814	7360	2673	684	3580	8475	3404	6326	1556	1120	1435	1736	15710	595	713	32570	3251	474	276	7088	1300	887	4532	2065	1471	867
2020-07-24	45580	135400	326	30120	35110	1653	1712	2131	5683	84340	8931	9255	7805	30670	4626	15240	41990	14050	1895	12930	6176	5581	144500	1369	389	3162	8227	1777	5560	3307	817	7372	2682	688	3613	8488	3413	6333	1560	1124	1454	1757	15710	599	730	32580	3265	479	280	7098	1337	902	4636	2073	1476	870
2020-07-25	45630	135400	330	30140	35110	1657	1716	2149	5690	85300	8937	9467	8002	31120	4714	15460	42450	14180	1913	13010	6365	5596	145200	1390	394	3250	8318	1780	5677	3360	820	7383	2691	692	3647	8499	3421	6340	1563	1128	1472	1778	15710	602	746	32600	3278	484	285	7108	1375	918	4800	2079	1481	873
2020-07-26	45660	135500	335	30140	35120	1660	1721	2163	5696	86200	8941	9467	8212	31450	4810	15700	42870	14300	1932	13070	6507	5612	146000	1418	399	3340	8420	1782	5797	3418	823	7394	2700	696	3682	8507	3429	6348	1566	1132	1493	1795	15710	605	764	32610	3290	487	291	7118	1413	935	4971	2085	1485	876
2020-07-27	45700	135600	341	30140	35130	1663	1723	2178	5702	87190	8948	9467	8418	31740	4910	15930	43190	14430	1950	13130	6665	5626	146700	1436	404	3432	8519	1787	5930	3469	825	7404	2707	699	3709	8515	3436	6354	1570	1135	1511	1817	15710	608	785	32620	3304	493	297	7127	1444	952	5103	2090	1487	879
2020-07-28	45720	135700	345	30150	35130	1666	1727	2194	5708	88150	8954	9467	8591	32120	5018	16170	43560	14540	1966	13190	6765	5639	147500	1458	413	3529	8616	1791	6047	3545	828	7413	2714	703	3734	8520	3442	6361	1572	1138	1530	1835	15710	611	805	32630	3316	496	303	7137	1477	969	5251	2093	1490	881
2020-07-29	45740	135700	349	30150	35140	1669	1730	2209	5712	89240	8960	9467	8776	32400	5128	16380	43860	14640	1984	13240	6917	5652	148200	1479	420	3628	8712	1794	6207	3610	830	7422	2721	706	3754	8524	3448	6368	1574	1142	1547	1851	15710	614	825	32640	3326	504	310	7146	1505	987	5384	2096	1493	884
2020-07-30	45770	135800	354	30160	35150	1671	1732	2224	5719	90140	8964	9467	8962	32620	5229	16620	44210	14730	1999	13290	7151	5663	148800	1525	428	3718	8799	1798	6386	3690	833	7430	2726	709	3774	8529	3455	6373	1576	1144	1564	1870	15710	616	847	32650	3339	510	315	7155	1559	1004	5522	2101	1496	887
2020-07-31	45800	135900	358	30170	35150	1673	1738	2240	5724	90990	8969	9467	9129	32920	5331	16840	44510	14810	2015	13350	7360	5674	149500	1558	432	3805	8889	1801	6502	3773	836	7439	2730	713	3791	8532	3461	6379	1578	1147	1581	1889	15710	619	863	32660	3352	515	321	7165	1593	1016	5652	2105	1497	891

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

[2020-07-01] 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.

COVID-19 short-term forecasts Deaths 2020-07-22

Gnereal information