For our Deep Dive this week I look at the ability of machine learning techniques to predict financial crises based on a new BoE study. It confirms the importance of yield curves and credit growth in predicting crises but also captures the non-linearities which conventional models miss. We also include an update on our various indicators.
We now have three models that give odds of US recessions. There is some disagreement amongst them. We update our US election tracker now that the results from Super Tuesday are in. Finally, we show which countries are experiencing the fastest growth in cases of COVID-19 and which countries have yet to report any cases.
This article is only available to Macro Hive subscribers. Sign-up to receive world-class macro analysis with a daily curated newsletter, podcast, original content from award-winning researchers, cross market strategy, equity insights, trade ideas, crypto flow frameworks, academic paper summaries, explanation and analysis of market-moving events, community investor chat room, and more.
(total reading time: 2 mins)
For our Deep Dive this week I look at the ability of machine learning techniques to predict financial crises based on a new BoE study. It confirms the importance of yield curves and credit growth in predicting crises but also captures the non-linearities which conventional models miss.
We also include an update on our various indicators. We now have three models that give odds of US recessions. There is some disagreement amongst them. We update our US election tracker now that the results from Super Tuesday are in. Finally, we show which countries are experiencing the fastest growth in cases of COVID-19 and which countries have yet to report any cases.
Enjoy!
Bilal
Can Machine Learning Help Predict Crises (3 min read) Forecasting financial crises is difficult. The data comprises a limited set of observed instances and signals a warning only when it is already too late to intervene. Moreover, the early warning models are complex and difficult to distil into simple, transparent indicators for macroprudential authorities. But can we use machine learning to model the crises? A new staff working paper from the Bank of England tries to find out.
(Bilal Hafeez │ 4th March, 2020)
• The Fed’s actions seem to have done the trick – the probability of a US recession has fallen by 5 percentage points to 29% over the past week. That is, if we look at the odds implied by the US yield curve.
• The consensus of economists is still at the probability of 25%.
• Biden was the clear winner from Super Tuesday. His odds of getting the Democratic nomination have surged to around 70%. Sanders odds have collapsed.
• Yesterday, we showed which states Sanders and Biden were expected to win. The latest results show that polling and prediction markets got Massachusetts, Minnesota and Maine wrong in predicting a Sanders win. It looks like Biden has won all three.
• Europe continues to see the fastest growth in confirmed cases of COVID-19. Netherlands, Belgium and Switzerland have seen noticeable jumps over the past week.
• Recently, Argentina, Chile, Poland and Tunisia have also reported their first cases.
• In a new table, we show which major countries have yet to report cases. They are Colombia, Hungary, South Africa and Turkey.
(The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs.)