Full metadata
Title
Evaluating the Utility of Investor Sentiment Measures in Predicting Performance of the S&P 500
Description
This paper investigates whether measures of investor sentiment can be used to predict future total returns of the S&P 500 index. Rolling regressions and other statistical techniques are used to determine which indicators contain the most predictive information and which time horizons' returns are "easiest" to predict in a three year data set. The five "most predictive" indicators are used to predict 180 calendar day future returns of the market and simulated investment of hypothetical accounts is conducted in an independent six year data set based on the rolling regression future return predictions. Some indicators, most notably the VIX index, appear to contain predictive information which led to out-performance of the accounts that invested based on the rolling regression model's predictions.
Date Created
2013-12
Contributors
- Dundas, Matthew William (Author)
- Boggess, May (Thesis director)
- Budolfson, Arthur (Committee member)
- Hedegaard, Esben (Committee member)
- Barrett, The Honors College (Contributor)
- School of Mathematical and Statistical Sciences (Contributor)
- Department of Finance (Contributor)
Topical Subject
Resource Type
Extent
19 pages
Language
Copyright Statement
In Copyright
Primary Member of
Series
Academic Year 2013-2014
Handle
https://hdl.handle.net/2286/R.I.19222
Level of coding
minimal
Cataloging Standards
System Created
- 2017-10-30 02:50:57
System Modified
- 2021-08-11 04:09:57
- 2 years 8 months ago
Additional Formats