Can Public Sentiment Improve Portfolio Performance?
(2024) DABN01 20241Department of Economics
Department of Statistics
- Abstract
- This study investigates whether asset allocation using public sentiment from social media data can enhance portfolio performance compared to traditional strategies. By constructing a sentiment index based on tweets about major NASDAQ-listed companies and employing machine learning and natural language processing techniques, we forecast public sentiment to inform a dynamic asset allocation strategy.
Focusing on the NASDAQ-100 index, we look at whether sentiment-based strategies can achieve higher risk-adjusted returns than the conventional 60/40 portfolio allocation, by comparing the Sharpe ratios of both portfolios over an 18 month period.
The results show that portfolios informed by public sentiment outperform the benchmark 60/40... (More) - This study investigates whether asset allocation using public sentiment from social media data can enhance portfolio performance compared to traditional strategies. By constructing a sentiment index based on tweets about major NASDAQ-listed companies and employing machine learning and natural language processing techniques, we forecast public sentiment to inform a dynamic asset allocation strategy.
Focusing on the NASDAQ-100 index, we look at whether sentiment-based strategies can achieve higher risk-adjusted returns than the conventional 60/40 portfolio allocation, by comparing the Sharpe ratios of both portfolios over an 18 month period.
The results show that portfolios informed by public sentiment outperform the benchmark 60/40 portfolio, yielding excellent Sharpe ratios. We find that forecasting a 5-day ahead moving average of the sentiment index results in greater portfolio performance compared to one-day-ahead sentiment forecasts. The results suggest that public sentiment can be a valuable indicator for asset allocation and market regime classification, offering great potential for enhancing investment strategies.
Our findings contribute to the growing literature on integrating social media data into financial analysis. Specifically, we fill a research gap on exploring the feasibility of forecasting sentiment to enhance portfolio performance. This study offers practical insights for investors seeking to leverage public opinion in portfolio construction. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9163151
- author
- Matthíasdóttir, Inga María LU and Eichholz, Friederike LU
- supervisor
-
- Simon Reese LU
- organization
- course
- DABN01 20241
- year
- 2024
- type
- H1 - Master's Degree (One Year)
- subject
- keywords
- machine learning, sentiment analysis, portfolio asset allocation, sentiment forecasting, natural language processing
- language
- English
- id
- 9163151
- date added to LUP
- 2024-09-24 08:35:37
- date last changed
- 2024-09-24 08:35:37
@misc{9163151, abstract = {{This study investigates whether asset allocation using public sentiment from social media data can enhance portfolio performance compared to traditional strategies. By constructing a sentiment index based on tweets about major NASDAQ-listed companies and employing machine learning and natural language processing techniques, we forecast public sentiment to inform a dynamic asset allocation strategy. Focusing on the NASDAQ-100 index, we look at whether sentiment-based strategies can achieve higher risk-adjusted returns than the conventional 60/40 portfolio allocation, by comparing the Sharpe ratios of both portfolios over an 18 month period. The results show that portfolios informed by public sentiment outperform the benchmark 60/40 portfolio, yielding excellent Sharpe ratios. We find that forecasting a 5-day ahead moving average of the sentiment index results in greater portfolio performance compared to one-day-ahead sentiment forecasts. The results suggest that public sentiment can be a valuable indicator for asset allocation and market regime classification, offering great potential for enhancing investment strategies. Our findings contribute to the growing literature on integrating social media data into financial analysis. Specifically, we fill a research gap on exploring the feasibility of forecasting sentiment to enhance portfolio performance. This study offers practical insights for investors seeking to leverage public opinion in portfolio construction.}}, author = {{Matthíasdóttir, Inga María and Eichholz, Friederike}}, language = {{eng}}, note = {{Student Paper}}, title = {{Can Public Sentiment Improve Portfolio Performance?}}, year = {{2024}}, }