Testing Stock Strategies with Crowdsourced Sentiment Scores

Summary

This article describes a simple way to test whether website sentiment indicators can inform stock trades. The author collects buy, sell, and hold readings for a stock list by extracting values from a financial portal, then stores them in a data frame or spreadsheet for analysis. An initial experiment captures scores before the market opens and compares them with each stock’s price change by the end of the session.

The reported single-day sample is described as unencouraging, and the author cautions that this brief observation cannot establish whether the strategy has potential. Suggested next steps include selecting stocks with more extreme readings, comparing indicators across sites, and combining sentiment with another strategy. The document offers no extended backtest, detailed performance statistics, or evidence that the site’s indicator represents overall market opinion. Its main contribution is a testable workflow and a clear reminder that a small one-day sample is inadequate for evaluating a trading signal. The data source and scraping method may also limit repeatability if the website changes how it presents its readings.

Key ideas

• The strategy uses a financial website’s buy, sell, and hold indicators as direct stock sentiment inputs.
• Sentiment readings can be collected before the market opens and compared with same-day stock price changes.
• The author’s single-day sample was not encouraging and is too limited to judge the strategy’s potential.
• Possible extensions include filtering for extreme sentiment, comparing multiple sources, and combining sentiment with another strategy.
• The article does not establish that the indicator accurately measures broad market sentiment or produces reliable signals.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.