Abstract
This study investigates how sentiment analysis can enhance resale price forecasting in the secondary luxury watch market, focusing on Rolex timepieces. Using signalling theory and social media influence theory, it examines whether discussions on luxury-watch-focused subreddits reflect market price movements. Over 4,000 Reddit posts collected between January and December 2023 were matched with weekly resale price data from Chrono24 for models like the Rolex Explorer, Oyster Perpetual, and Submariner. The analysis employed the FinBERT natural language processing (NLP) model, with forecasting performed using ARIMA and Long Short-Term Memory (LSTM) models. Results showed that 85.89% of comments were neutral, with most containing informational rather than emotional content. Although the more sophisticated LSTM outperformed the simpler ARIMA in capturing nonlinear market patterns, adding sentiment data yielded only marginal improvements in predictive accuracy. These findings suggest that short-term consumer sentiment has a limited impact on secondary market prices, which are primarily driven by stable brand signals such as heritage, craftsmanship, and scarcity. This research adds to the literature on luxury market analytics by emphasising the importance of domain-specific sentiment analysis. It also offers practical insights for brands and stakeholders, highlighting sentiment as a long-term indicator of brand perception rather than a short-term predictor of prices.