Increased flow of unfavorable information
Financial Times Unveils Economic Sentiment Series: A 40-Year Analysis of Weekly Sentiment
The Financial Times (FT) has revealed a unique series, dubbed 'macro mood,' which provides a comprehensive overview of average weekly sentiment since 1982. This series, marked by notable troughs of negative sentiment, offers an insightful perspective into the evolving economic climate over the past four decades.
The correlation between the FT's daily sentiment and Wall Street's Vix index, often referred to as a 'fear gauge,' is strong and statistically significant. This relationship explains nearly 20% of the daily Vix close price, suggesting a significant influence of the FT's sentiment on market fear levels.
The FT's sentiment analysis is not just about catering to the modern world's demand for negativity, but it plays a crucial role in optimizing user engagement. Each article is assigned a general sentiment score, ranging from -1 to 1, and a second model categorizes how closely each article aligns with the 'Platonic ideal of economics articles,' with a score ranging from 0 to 1.
The FT's macro mood series is constructed using advanced methods for extracting data from text, such as embeddings and fine-tuned large language models. The Monetary Policy Radar was instrumental in this construction.
The FT's financial crisis of 2007-2008 had profound impacts, not only on society but also on journalists. The average sentiment since the crisis has yet to recover, remaining significantly below the pre-crisis period (1982-2006).
Interestingly, a simple model using FT sentiment can predict US CPI inflation over the next 12 months more accurately than the Fed's own staff forecast and a standard benchmark model. Moreover, a model that includes a breakdown of topics covered in the FT further improves the prediction of US CPI inflation.
The current macro mood is influenced by a variety of negative factors, including Covid-19, inflation, rising populism, and extreme politics. The 2017 sentiment recovery, which occurred during a bull run in the 'stock market today' and a lessening of hard Brexit and Trump presidency concerns, was a brief respite and not sustained. This recovery turned sour due to the start of Trump trade wars and the anointment of Boris Johnson as UK prime minister.
The FT employs a model based on natural language processing (NLP) techniques, specifically leveraging machine learning models such as BERT, GPT, or 'RoBERTa,' to determine the overall sentiment in their articles. The FT's macro mood is updated daily on the Monetary Policy Radar data page.
Journalists, including those at the FT, provide a vital civic service by shining a light on scandals and holding the powerful to account. Despite the focus on negative sentiment, it is essential to remember the crucial role that balanced, accurate, and insightful journalism plays in our society.
Read also:
- Irish expatriates find strong pull towards homeland; require additional incentives to resettle in Ireland
- Berlin intends to host the Nova music festival exhibition at Tempelhof Airport
- Bringing the Nova music festival exhibition to Tempelhof Airport, Berlin
- Transports Berlin's Nova music festival showcase to Tempelhof Airport for a grand event
