Skip to content
All library documents

Historical Market Capitalization Data and Survivorship Bias

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

Summary

The document discusses obtaining market capitalization for S&P 500 stocks on a historical date using low-cost data sources. Suggested approaches include extracting data with R, downloading data from a general-purpose service, or combining historical prices with shares outstanding. These options are presented as possible routes rather than verified, reliable datasets; the responses explicitly question the accuracy and availability of free data.

A central warning is that reconstructing historical index constituents from today’s list introduces survivorship bias, because companies that later left the index or delisted would be missing. Historical shares outstanding can also differ after buybacks, acquisitions, share issuance, and other corporate actions. Further complications include multiple share classes and convertible securities, which mean that price multiplied by a simple share count may not match reported market capitalization. Any source’s methodology should be checked, and professional data may be needed when accuracy matters.

Key ideas

  • Using current index constituents for a historical study can create survivorship bias.
  • Historical market capitalization depends on the shares outstanding at the historical date.
  • Mergers, buybacks, issuance, and delistings can change the inputs to market capitalization.
  • Multiple share classes and convertible securities complicate price-times-shares calculations.
  • Free data sources may be unreliable, so their definitions and methods need scrutiny.

Tags

EquitiesUS marketsBacktestingStatistics
Full text
# Get market cap by ticker on 1.7.2013?


# Get market cap by ticker on 1.7.2013?












I have a list of all S&P500 tickers, e.g. AAPL, GOOG, JPM.

I would like to get their market cap on 1.7.2013 (I don't have Bloomberg, only free internet).

Is there an excel addin or other tool/page where I can obtain historic market cap?

The page https://www.quandl.com/ seems to provide financial information, but I did not find how to retrieve the market cap by ticker.

## Answer by Quantopik (score 2, accepted)

https://quant.stackexchange.com/a/17495

I faced the same issue some years ago and I solved by implementing the R script reported here; now, with new Yahoo disclaimer rules, it seems to be broken, but, anyway you should be able to replicate the data mining process using that script together with this.

If you're pretty confident with R, you should be able to do that.

Alternatively, you can visit wolframalpha.com and download the data.

You can also do the same by using python and QUANDL API, but it should more complicated. The only one problem is about the data reliability in both cases, but, as usual, if database is for free, it is normal to get low-quality data.

## Answer by Richard at NorgateData (score 3)

https://quant.stackexchange.com/a/17505

Some of the issues with this sort of request is:

a) Today's S&P 500 components are not the same from 1 Jul 2013. By using today's components you are introducing pre-inclusion/survivorship bias. Are you going to be able to find data on the delisted stocks? eg. Since 1 Jul 2013, Sprint Corporation, BMC Software, NYSE Euronext, Molex, Life Technologies, Beam Suntory, International Game Technology, Forest Laboratories, Safeway, Covidien, Allergan have all delisted (due to takeover/merger). There's also a lot more that have been removed as part of the quarterly rebalancing.

b) Today's shares outstanding are not the same as 1 Jul 2013. There may have been mergers/takeovers (where shares are issued as part of the takeover), demergers, buybacks, employee share issues, private placements etc. since that time.

c) Shares outstanding * share price does NOT equal market cap in all cases. Some companies (e.g. Berkshire Hathway, Google etc.) have multiple types of share classes, sometimes with different ownership levels, others with unlisted shares. Others have securities that convert to normal securities (Convertible Preference, Warrants etc.) - so these also need to be considered.

Due to these complexities, this sort of data is just not freely available. Given you only need a date a coupe of years back then the difference may not be crucial for your exercise, but if you need absolute accuracy then you need to look at more professional sources.

With any source you use, you must determine exactly what methodology is used to produce the metrics.

## Answer by Rime (score 1)

https://quant.stackexchange.com/a/15606

@emcor I would suggest downloading the Stock's daily Prices & then downloading Shares Outstanding or average shares outstanding. Then find the product of the two to arrive at market caps.

I don't know how reliable quandl's data is or if they have shares outstanding data, but if they do this can probably be done using R code since you will be downloading many ticker's data

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.