Skip to content
All library documents

Yahoo Finance Adjusted Price Reliability and Dividend Data Errors

Article Quant Q&A · Author: Ravi
Read the original

Summary

The document describes a reported discrepancy in AAPL returns between two historical calculations. A student found that the return over a specified period differed from class notes prepared earlier, prompting an investigation of Yahoo Finance adjusted prices. The author attributes the discrepancy to dividends that had been coded incorrectly in the data, and asks whether similar errors may affect other observations or downloaded datasets.

This is an anecdotal data-quality report, not a systematic audit. It offers no broader error count, validation method, or evidence about the reliability of the R download function mentioned. The episode highlights that adjusted historical prices depend on correctly recorded corporate actions and that data revisions can change calculated returns. Researchers using vendor-adjusted prices may want to check dividend records and compare critical observations with independent sources, while recognizing that the document alone cannot quantify how often such problems occur.

Key ideas

  • An AAPL return discrepancy was traced to incorrectly coded dividend data.
  • Adjusted historical prices can produce different returns when corporate-action records change or contain errors.
  • The report raises a question about the reliability of a data source and download function but does not assess them systematically.
  • A single incident does not establish the frequency or extent of errors in the broader dataset.

Tags

EquitiesUS marketsBacktesting
Full text
# Error in Yahoo! adjusted prices - The case of AAPL


# Error in Yahoo! adjusted prices - The case of AAPL












One of my students reported that return for AAPL between April 30, 2014 and June 30, 2014 was different than what was shown in my class notes which were prepared in early 2016. I investigated the error and found the problem in incorrectly coded dividends. The screenshot below shows the problem dividends:

This makes me wonder how many such errors are there in Yahoo! Finance data and how reliable are the results using data downloaded using the `getsymbols` function in `R`. Has anyone else experienced similar issues?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

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