AI-Driven Financial Foresight

The overall media sentiment about bitcoin is in the high end of its range. According to academic research on social media and daily asset prices, "Sentiment predicts positive next-day returns, but attention predicts negative next-day returns." https://lnkd.in/g8GgXdHX. We see a similar effect at monthly and longer horizons for many assets (stocks, currencies, etc...). News media attention to bitcoin lately has been very high. The Bitcoin Halving is likely to occur in late April. From a psychological perspective, the increasing scarcity of bitcoin increases its attractiveness. Yet the last halving was May 11, 2020, and previously one occurred in July 2016. Before and after the past halvings, there was no clear price impact of the events. I've noticed in recent conversations at finance events that the launch of ETFs (specifically the SEC approval) has legitimized Bitcoin in the eyes of conservative financial institutions and family offices. Many are discussing new bitcoin products and investments. Surprisingly to me, the tradfi adoption of bitcoin as a store of value seems to be in early stages. So is Bitcoin peaking? What does the halving mean? It's hard to know in the short term: sentiment is higher (and vulnerable to a fall). The high overall attention implies downward pressure in future months. Longer term (years) the momentum is still positive.

Historically social media sentiment about individual cryptocurrencies tends to drive their prices higher. The highest sentiment large coins in March 2024 so far are #DOT #ARBITRUM #MATIC #AVAX #BNB.

Here's a ranked view on some of the smaller #cryptocurrencies' sentiment from #socialmedia covering the period March 1 through March 12, 2024.

We built a 30-day U.S. stock predictive model based on Media Sentiment using a simple Machine Learning algorithm - updated returns in the below image. Humans, even in hyper-competitive environments like the US stock market, respond predictably to news and social media content.The model was put in production in Aug 2019 and launched commercially on Jan 2020 in partnership with the StarMine team at LSEG Data & Analytics.The model ranks stocks (1 to 100) based on their likelihood of outperformance using a ML algo applied to only our LSEG MarketPsych Analytics news and social media-derived data. The blue line in the chart below depicts the spread between the top-ranked decile (10%) of stocks and the bottom-ranked decile - the absolute return portfolio. The gray-shaded regions are the 2-year training periods.The model performs similarly regardless of market cap. It is orthogonal to other StarMine models. It has about 30% monthly turnover, with no transaction costs assumed. Longer term models also work well with media data (we've tested out to 3 year forecasts).

For our UK readers, the evolving scandal at the UK's Post Office Ltd & Fujitsu is rapidly impacting Fujitsu's ESG scores. Background: The digital accounting system that the postmasters (regular people) entered their daily sales & cost calculations into were being fraudulently altered via remote access from the UK's Post Office Ltd / Fujitsu. In our ESG data Post Office Ltd has the lowest (1st percentile) Community & Social scores and a 4th percentile ESGControversies score (very bad). The blame has gradually shifted from Post Office Ltd to Fujitsu over the past month in social media, and Fujitsu's ESG scores have dropped substantially - see graphic below.

In our earlier post we identified #Robotics as the industry with the most positive media sentiment in 2023 (and 2022).Attached is a list of the companies most associated with Robotic initiatives or products in the past two months. Based on our #quant research, the valuations of such companies tend to (on average) outperform when associated with higher media sentiment (as they have been).  

Every year we look at the most positively viewed industries in global #businessnews and #socialmedia.For 2023 we see #Robotics on top (as it was in 2022) followed by #RenewableEnergy, #Tourism, #Fintech, and #Infrastructure.Further down the list the industries with a country code after them indicate the industry is primarily favored in one region (US = United States, CN = China).The stock prices of the highest #sentiment companies and industries on average outperform over the following year.In 2022 AI was #2 but fell to #33 based on legal and social challenges.We wish you all a happy, healthy, and prosperous 2024! #happynewyear  

#Bitcoin sentiment has been at highs and the bull keeps running - a phenomenon we call sentiment momentum. In this paper https://lnkd.in/gasgp9BQ we show that across thousands of global stocks, sentiment precedes price momentum (using Granger Causality Analysis), and the chart below illlustrates that effect in #BTC.In the chart, the black price bars are BTC/USD and the smooth lines are two moving averages of #sentiment (30 day and 90 day superimposed). The red shading indicates negative sentiment pressure (the 30-day sentiment is lower than 90-day), and blue shading indicates more positive sentiment.Social sentiment on BTC is the highest since 2020, and news sentiment is the highest since 2018. When sentiment falls substantially, we're likely to see a price correction, but that isn't happening yet.