ChatGPT and Deepseek: Can They Predict the Stock Market and Macroeconomy?
Jian Chen (Xiamen University), et al.
February 2025
We study whether ChatGPT and DeepSeek can extract information from the Wall Street Journal to predict the stock market and the macroeconomy. We find that ChatGPT has predictive power. DeepSeek underperforms ChatGPT, which is trained more extensively in English. Other large language models also underperform. Consistent with financial theories, the predictability is driven by investors’ underreaction to positive news, especially during periods of economic downturn and high information uncertainty. Negative news correlates with returns but lacks predictive value. At present, ChatGPT appears to be the only model capable of capturing economic news that links to the market risk premium.

Regimes
Amara Mulliner (Man AHL), et al.
March 2025
We propose a new systematic method for detecting the current economic regime and show how to use this information for predicting returns. Rather than presupposing a set of possible regimes, we rely on economic state variables and determine for which historical dates the values of these variables were most similar. To establish our position in an asset today, we identify historically similar periods and measure subsequent performance of the asset. If the historical performance is positive, we initiate a long position; conversely, if it is negative, we initiate a short position. We illustrate the efficacy of our method on six common long-short equity factors over 1985-2024. Our results show that using this information our regime classification leads to significant outperformance. Interestingly, we also find important information in what we call anti-regimesperiods in the past that are the most dissimilar to today.

What Moves Prices? The Dynamics of Fundamentals and Returns
Fabio Girardi (Vienna U. of Econ. and Business) and C. Schlag (Goethe U. Frankfurt)
February 2025
We propose a dynamic model that explains a large share of both in-sample and out-of-sample variation in the annual return and growth of fundamentals on the aggregate S&P 500 index. To capture the time variation in investors’ beliefs, we rely on a penalized vector autoregressive model and predictors that summarize a substantial portion of the information available in the market. We combine model-implied conditional expectations and present value identities to investigate what drives the variations in the price-to-dividend and price-to-earnings ratios. We find that time-varying expected returns account for most of the movements in the price-to-dividend ratio over the period 1980-2021, but play a smaller role in the price-to-earnings ratio. Notably, over the period 2001-2021, the expected growth of fundamentals explains a significantly larger share of the variation in both valuation ratios.

Trends and Reversion in Financial Markets on Time Scales from Minutes to Decades
Christof Schmidhuber and Sara A. Safari (Zurich U. of Applied Sciences)
March 2025
We empirically analyze the reversion of financial market trends with time horizons ranging from minutes to decades. The analysis covers equities, interest rates, currencies and commodities and combines 14 years of futures tick data, 30 years of daily futures prices, 330 years of monthly asset prices, and yearly financial data since medieval times. Across asset classes, we find that markets are in a “trending regime” on time scales that range from a few hours to a few years, while they are in a “reversion regime” on shorter and longer time scales. In the “trending regime”, weak trends tend to persist, which can be explained by herding behavior of investors. However, in this regime trends tend to revert before they become strong enough to be statistically significant, which can be interpreted as a return of asset prices to their intrinsic value. In the “reversion regime”, we find the opposite pattern: weak trends tend to revert, while those trends that become statistically significant tend to persist. Our results provide a set of empirical tests of theoretical models of financial markets. We interpret them in the light of a recently proposed lattice gas model, where the lattice represents the social network of traders, the gas molecules represent the shares of financial assets, and efficient markets correspond to the critical point. If this model is accurate, the lattice gas must be near this critical point on time scales from 1 hour to a few days, with a correlation time of a few years.

Informative Price Pressure
Salman Arif (University of Minnesota), et al.
February 2025
Informed investors often hedge their stock bets right before FOMC meetings. The resulting price pressure, when aggregated across stocks, reveals their long-term view of the stock market (with a minus sign). Consistent with this, we find that the average stock market return on the day before recent FOMC meetings, while completely reverted the next day, strongly and negatively predicts stock market returns up to two years in the future. The market return predictability is robust to additional controls, various sample cuts and extends to other important macroeconomic announcements. The day before the FOMC meeting is associated with low informed trading intensity, which explains the decision of informed investors to hedge on that day. At the same time, the VIX index is higher on that day, resulting in detectable price pressure.

Media tone is a priced risk factor in currency markets
Kari Heimonen (University of Jyväskylä), et al.
March 2025
Media tone constructed from 7,000,000 articles from 2,000 global media and 800 social media sites is found to be a genuine risk factor that cross-sectionally prices currencies. It can predict excess US dollar returns for up to six months and surpasses the no-change benchmark in predicting returns out of sample. Its predicted value contains information beyond those predicted by currency factors and business cycles. Evidence collaborates with the theory that Media tone increases investment returns, has pronounced predictive power for the currencies associated with hard-to-value characteristics, and its predictive power increases with media sources. Trading of rational investors, including banks, is associated with Media tone.

Learn To Use R For Portfolio Analysis
Quantitative Investment Portfolio Analytics In R:
An Introduction To R For Modeling Portfolio Risk and Return
By James Picerno