Doombot: a machine learning algorithm for predicting downturns in OECD countries:
This paper describes an algorithm, "DoomBot", which selects parsimonious models to predict downturns over different quarterly horizons covering the ensuing two years for 20 OECD countries. The models are country- and horizon-specific and are automatically updated as the estimation sample p...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Electronic eBook |
Language: | English |
Published: |
Paris
OECD Publishing
2023
|
Series: | OECD Economics Department Working Papers
no.1780 |
Subjects: | |
Online Access: | DE-862 DE-863 |
Summary: | This paper describes an algorithm, "DoomBot", which selects parsimonious models to predict downturns over different quarterly horizons covering the ensuing two years for 20 OECD countries. The models are country- and horizon-specific and are automatically updated as the estimation sample period is extended, so facilitating out-of-sample evaluation of the algorithm. A limited combination of explanatory variables is chosen from a much larger pool of potential variables that include those that have been most useful in predicting downturns in previous OECD work. The most frequently selected variables are financial variables, especially those relating to credit and house prices, but also include equity prices and various measures of interest rates (such as the slope of the yield curve). Business cycle variables -- survey measure of capacity utilisation, industrial production, GDP and unemployment -- are also selected, but more frequently at very short horizons. The variables selected do not just relate to the domestic economy of the country being considered, but also international aggregates, consistent with findings from previous OECD work. The in-sample fit of the models is very good on standard performance metrics, although the out-of-sample performance is less impressive. The models do, however, provide a clear out-of-sample early warning of the Global Financial Crisis (GFC), especially when considered collectively, although they do generate 'false alarms' just ahead of the crisis. The models are less good at predicting the euro area crisis out-of-sample, but it is clear from the evolution of the choice of variables that the algorithm learns from this episode, for example through the more frequent selection of a variable measuring euro area sovereign bond spreads. The latest out-of-sample predictions made in mid-2023, suggest the probability of a downturn is at its greatest and most widespread since the GFC, with the largest contributions to such risks coming from house prices, interest rate developments (as measured by the slope of the yield curve and the rapidity of the change in short rates) and oil prices. On the other hand, warning signals from business cycle variables and equity prices, which are often good downturn predictors at short horizons, are conspicuously absent. |
Physical Description: | 1 Online-Ressource (63 Seiten) 21 x 28cm. |
Staff View
MARC
LEADER | 00000cam a22000002 4500 | ||
---|---|---|---|
001 | ZDB-13-SOC-101204922 | ||
003 | DE-627-1 | ||
005 | 20241028114255.0 | ||
007 | cr uuu---uuuuu | ||
008 | 240311s2023 xx |||||o 00| ||eng c | ||
024 | 7 | |a 10.1787/4ed7acc3-en |2 doi | |
035 | |a (DE-627-1)101204922 | ||
035 | |a (DE-599)KEP101204922 | ||
035 | |a (FR-PaOEC)4ed7acc3-en | ||
035 | |a (DE-627-1)101204922 | ||
040 | |a DE-627 |b ger |c DE-627 |e rda | ||
041 | |a eng | ||
084 | |a E01 |2 jelc | ||
084 | |a E58 |2 jelc | ||
084 | |a E65 |2 jelc | ||
084 | |a E66 |2 jelc | ||
084 | |a E17 |2 jelc | ||
100 | 1 | |a Chalaux, Thomas |e VerfasserIn |4 aut | |
245 | 1 | 0 | |a Doombot: a machine learning algorithm for predicting downturns in OECD countries |c Thomas, Chalaux and David, Turner |
264 | 1 | |a Paris |b OECD Publishing |c 2023 | |
300 | |a 1 Online-Ressource (63 Seiten) |c 21 x 28cm. | ||
336 | |a Text |b txt |2 rdacontent | ||
337 | |a Computermedien |b c |2 rdamedia | ||
338 | |a Online-Ressource |b cr |2 rdacarrier | ||
490 | 0 | |a OECD Economics Department Working Papers |v no.1780 | |
520 | |a This paper describes an algorithm, "DoomBot", which selects parsimonious models to predict downturns over different quarterly horizons covering the ensuing two years for 20 OECD countries. The models are country- and horizon-specific and are automatically updated as the estimation sample period is extended, so facilitating out-of-sample evaluation of the algorithm. A limited combination of explanatory variables is chosen from a much larger pool of potential variables that include those that have been most useful in predicting downturns in previous OECD work. The most frequently selected variables are financial variables, especially those relating to credit and house prices, but also include equity prices and various measures of interest rates (such as the slope of the yield curve). Business cycle variables -- survey measure of capacity utilisation, industrial production, GDP and unemployment -- are also selected, but more frequently at very short horizons. The variables selected do not just relate to the domestic economy of the country being considered, but also international aggregates, consistent with findings from previous OECD work. The in-sample fit of the models is very good on standard performance metrics, although the out-of-sample performance is less impressive. The models do, however, provide a clear out-of-sample early warning of the Global Financial Crisis (GFC), especially when considered collectively, although they do generate 'false alarms' just ahead of the crisis. The models are less good at predicting the euro area crisis out-of-sample, but it is clear from the evolution of the choice of variables that the algorithm learns from this episode, for example through the more frequent selection of a variable measuring euro area sovereign bond spreads. The latest out-of-sample predictions made in mid-2023, suggest the probability of a downturn is at its greatest and most widespread since the GFC, with the largest contributions to such risks coming from house prices, interest rate developments (as measured by the slope of the yield curve and the rapidity of the change in short rates) and oil prices. On the other hand, warning signals from business cycle variables and equity prices, which are often good downturn predictors at short horizons, are conspicuously absent. | ||
650 | 4 | |a Science and Technology | |
650 | 4 | |a Economics | |
700 | 1 | |a Turner, David |e MitwirkendeR |4 ctb | |
966 | 4 | 0 | |l DE-862 |p ZDB-13-SOC |q FWS_PDA_SOC |u https://doi.org/10.1787/4ed7acc3-en |m X:OECD |x Resolving-System |z lizenzpflichtig |3 Volltext |
966 | 4 | 0 | |l DE-863 |p ZDB-13-SOC |q FWS_PDA_SOC |u https://doi.org/10.1787/4ed7acc3-en |m X:OECD |x Resolving-System |z lizenzpflichtig |3 Volltext |
912 | |a ZDB-13-SOC-ebook | ||
912 | |a ZDB-13-SOC | ||
951 | |a BO | ||
912 | |a ZDB-13-SOC | ||
049 | |a DE-862 | ||
049 | |a DE-863 |
Record in the Search Index
DE-BY-FWS_katkey | ZDB-13-SOC-101204922 |
---|---|
_version_ | 1826942498674573313 |
adam_text | |
any_adam_object | |
author | Chalaux, Thomas |
author2 | Turner, David |
author2_role | ctb |
author2_variant | d t dt |
author_facet | Chalaux, Thomas Turner, David |
author_role | aut |
author_sort | Chalaux, Thomas |
author_variant | t c tc |
building | Verbundindex |
bvnumber | localFWS |
collection | ZDB-13-SOC-ebook ZDB-13-SOC |
ctrlnum | (DE-627-1)101204922 (DE-599)KEP101204922 (FR-PaOEC)4ed7acc3-en |
discipline | Wirtschaftswissenschaften |
format | Electronic eBook |
fullrecord | <?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>03661cam a22004212 4500</leader><controlfield tag="001">ZDB-13-SOC-101204922</controlfield><controlfield tag="003">DE-627-1</controlfield><controlfield tag="005">20241028114255.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">240311s2023 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1787/4ed7acc3-en</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627-1)101204922</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)KEP101204922</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(FR-PaOEC)4ed7acc3-en</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627-1)101204922</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rda</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">E01</subfield><subfield code="2">jelc</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">E58</subfield><subfield code="2">jelc</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">E65</subfield><subfield code="2">jelc</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">E66</subfield><subfield code="2">jelc</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">E17</subfield><subfield code="2">jelc</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Chalaux, Thomas</subfield><subfield code="e">VerfasserIn</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Doombot: a machine learning algorithm for predicting downturns in OECD countries</subfield><subfield code="c">Thomas, Chalaux and David, Turner</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="a">Paris</subfield><subfield code="b">OECD Publishing</subfield><subfield code="c">2023</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">1 Online-Ressource (63 Seiten)</subfield><subfield code="c">21 x 28cm.</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="490" ind1="0" ind2=" "><subfield code="a">OECD Economics Department Working Papers</subfield><subfield code="v">no.1780</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">This paper describes an algorithm, "DoomBot", which selects parsimonious models to predict downturns over different quarterly horizons covering the ensuing two years for 20 OECD countries. The models are country- and horizon-specific and are automatically updated as the estimation sample period is extended, so facilitating out-of-sample evaluation of the algorithm. A limited combination of explanatory variables is chosen from a much larger pool of potential variables that include those that have been most useful in predicting downturns in previous OECD work. The most frequently selected variables are financial variables, especially those relating to credit and house prices, but also include equity prices and various measures of interest rates (such as the slope of the yield curve). Business cycle variables -- survey measure of capacity utilisation, industrial production, GDP and unemployment -- are also selected, but more frequently at very short horizons. The variables selected do not just relate to the domestic economy of the country being considered, but also international aggregates, consistent with findings from previous OECD work. The in-sample fit of the models is very good on standard performance metrics, although the out-of-sample performance is less impressive. The models do, however, provide a clear out-of-sample early warning of the Global Financial Crisis (GFC), especially when considered collectively, although they do generate 'false alarms' just ahead of the crisis. The models are less good at predicting the euro area crisis out-of-sample, but it is clear from the evolution of the choice of variables that the algorithm learns from this episode, for example through the more frequent selection of a variable measuring euro area sovereign bond spreads. The latest out-of-sample predictions made in mid-2023, suggest the probability of a downturn is at its greatest and most widespread since the GFC, with the largest contributions to such risks coming from house prices, interest rate developments (as measured by the slope of the yield curve and the rapidity of the change in short rates) and oil prices. On the other hand, warning signals from business cycle variables and equity prices, which are often good downturn predictors at short horizons, are conspicuously absent.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Science and Technology</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Economics</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Turner, David</subfield><subfield code="e">MitwirkendeR</subfield><subfield code="4">ctb</subfield></datafield><datafield tag="966" ind1="4" ind2="0"><subfield code="l">DE-862</subfield><subfield code="p">ZDB-13-SOC</subfield><subfield code="q">FWS_PDA_SOC</subfield><subfield code="u">https://doi.org/10.1787/4ed7acc3-en</subfield><subfield code="m">X:OECD</subfield><subfield code="x">Resolving-System</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="966" ind1="4" ind2="0"><subfield code="l">DE-863</subfield><subfield code="p">ZDB-13-SOC</subfield><subfield code="q">FWS_PDA_SOC</subfield><subfield code="u">https://doi.org/10.1787/4ed7acc3-en</subfield><subfield code="m">X:OECD</subfield><subfield code="x">Resolving-System</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ZDB-13-SOC-ebook</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ZDB-13-SOC</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">BO</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ZDB-13-SOC</subfield></datafield><datafield tag="049" ind1=" " ind2=" "><subfield code="a">DE-862</subfield></datafield><datafield tag="049" ind1=" " ind2=" "><subfield code="a">DE-863</subfield></datafield></record></collection> |
id | ZDB-13-SOC-101204922 |
illustrated | Not Illustrated |
indexdate | 2025-03-18T14:28:58Z |
institution | BVB |
language | English |
open_access_boolean | |
owner | DE-862 DE-BY-FWS DE-863 DE-BY-FWS |
owner_facet | DE-862 DE-BY-FWS DE-863 DE-BY-FWS |
physical | 1 Online-Ressource (63 Seiten) 21 x 28cm. |
psigel | ZDB-13-SOC FWS_PDA_SOC ZDB-13-SOC-ebook ZDB-13-SOC |
publishDate | 2023 |
publishDateSearch | 2023 |
publishDateSort | 2023 |
publisher | OECD Publishing |
record_format | marc |
series2 | OECD Economics Department Working Papers |
spelling | Chalaux, Thomas VerfasserIn aut Doombot: a machine learning algorithm for predicting downturns in OECD countries Thomas, Chalaux and David, Turner Paris OECD Publishing 2023 1 Online-Ressource (63 Seiten) 21 x 28cm. Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier OECD Economics Department Working Papers no.1780 This paper describes an algorithm, "DoomBot", which selects parsimonious models to predict downturns over different quarterly horizons covering the ensuing two years for 20 OECD countries. The models are country- and horizon-specific and are automatically updated as the estimation sample period is extended, so facilitating out-of-sample evaluation of the algorithm. A limited combination of explanatory variables is chosen from a much larger pool of potential variables that include those that have been most useful in predicting downturns in previous OECD work. The most frequently selected variables are financial variables, especially those relating to credit and house prices, but also include equity prices and various measures of interest rates (such as the slope of the yield curve). Business cycle variables -- survey measure of capacity utilisation, industrial production, GDP and unemployment -- are also selected, but more frequently at very short horizons. The variables selected do not just relate to the domestic economy of the country being considered, but also international aggregates, consistent with findings from previous OECD work. The in-sample fit of the models is very good on standard performance metrics, although the out-of-sample performance is less impressive. The models do, however, provide a clear out-of-sample early warning of the Global Financial Crisis (GFC), especially when considered collectively, although they do generate 'false alarms' just ahead of the crisis. The models are less good at predicting the euro area crisis out-of-sample, but it is clear from the evolution of the choice of variables that the algorithm learns from this episode, for example through the more frequent selection of a variable measuring euro area sovereign bond spreads. The latest out-of-sample predictions made in mid-2023, suggest the probability of a downturn is at its greatest and most widespread since the GFC, with the largest contributions to such risks coming from house prices, interest rate developments (as measured by the slope of the yield curve and the rapidity of the change in short rates) and oil prices. On the other hand, warning signals from business cycle variables and equity prices, which are often good downturn predictors at short horizons, are conspicuously absent. Science and Technology Economics Turner, David MitwirkendeR ctb |
spellingShingle | Chalaux, Thomas Doombot: a machine learning algorithm for predicting downturns in OECD countries Science and Technology Economics |
title | Doombot: a machine learning algorithm for predicting downturns in OECD countries |
title_auth | Doombot: a machine learning algorithm for predicting downturns in OECD countries |
title_exact_search | Doombot: a machine learning algorithm for predicting downturns in OECD countries |
title_full | Doombot: a machine learning algorithm for predicting downturns in OECD countries Thomas, Chalaux and David, Turner |
title_fullStr | Doombot: a machine learning algorithm for predicting downturns in OECD countries Thomas, Chalaux and David, Turner |
title_full_unstemmed | Doombot: a machine learning algorithm for predicting downturns in OECD countries Thomas, Chalaux and David, Turner |
title_short | Doombot: a machine learning algorithm for predicting downturns in OECD countries |
title_sort | doombot a machine learning algorithm for predicting downturns in oecd countries |
topic | Science and Technology Economics |
topic_facet | Science and Technology Economics |
work_keys_str_mv | AT chalauxthomas doombotamachinelearningalgorithmforpredictingdownturnsinoecdcountries AT turnerdavid doombotamachinelearningalgorithmforpredictingdownturnsinoecdcountries |