The S&P 500 has to be one of the most quoted items in all of finance. A search on Google yields 132 million results. Everyone speaks of the S&P as a single asset; however, it isn’t. It’s obviously a collection of 500 names. The current components of the S&P are readily available on-line. But finding changes in the components of the index over time is not. Although being added or deleted from the index is a material event, there is not a common database of changes without being a subscriber to Standard & Poors.
Considering our enormous focus on an accurate and complete historical data, I have charged myself with the task of replicating the S&P 500 index components over the years. I think I can crafted a plan to get changes back to 1995. Our historical database dates back to 1987. If anyone has a good source of publicly available changes over the years, please feel free to email me at Bill AT BloodhoundSystem DOT com.
The S&P isn’t a Static Vehichle
Companies move in and out of the S&P 500 for a number of reasons. The most obvious are traditional corporate actions such as mergers, spin-offs or in the rare case bankruptcy while a component. Other companies just no longer qualify from a market capitalization perspective. Additionally, throughout its history, Standard & Poors has changed the rules of qualification or composition makeup. For instance, in a rare set of circumstances, GM and AIG returned to the S&P in June of this year after their 2009 expulsion. In July, Nielsen Holdings (NLSN) replaced Sprint Nextel (S) as SoftBank Corp. is acquiring a substantial stake in Sprint and the public float is expected to fall below the 50% criteria necessary for continued inclusion. Between 1995 and 2000, Standard & Poors rapidly increased the number of NASDAQ listed stocks, and in the 2000s, they eliminated International stocks.
For large changes, such as 5% or more, S&P updates the number of shares immediately. For small changes, S&P waits until the end of the calendar quarter to update.
Today’s S&P vs. 1995
In our compilation of the yearly changes in the Index, I took a look at how the S&P has changed from 1995 to today. Out of the 500 companies that represent the S&P today, only 211 of them can be directly traced back to companies in the index in 1995 (including GM and AIG who left and came back). Now that doesn’t included mergers where two components were both in the index, and now only take one spot. Said another way, there are 289 companies in today’s S&P 500 that can not be traced back to companies from 1995.
Take a company like Cintas (CTAS). In 1995, Cintas reported $615 million in revenue and was a $2.1bn market cap at the end of the year. By our calculation, it only put Cintas just outside the 500 largest capitalizations; however, they were not added to the Index until 2001. On the other hand, companies like Beam (BEAM) and Allegheny Technologies (ATI) were not in the Index in 1995 but have roots in companies that were. Beam is the remaining liquor business that was spun off from American Brands. Allegheny Technologies was formed by the combination of Allegheny Ludlum and Teledyne in 1996, the later of which was a member of the S&P 500.
With this analysis, if two S&P companies merged, it opened a spot for a new component. It understates turnover to some degree. For instance, in 1995, Exxon and Mobil both represented spots in the Index. The former J.D. Rockefeller company Standard Oil Company of New York, Mobil has roots dating back prior to the 1880s. However, it merged with Exxon in 1999 to form ExxonMobil. Both 1995 companies are still represented in today’s index, but as one entity. To quantify that effect, we then took a different perspective. We looked at the components of the S&P as it stood in 1995.
1995’s S&P Represented in Today’s Index
Of the 500 names that comprised the index at the end 2005, 322 components of the S&P are still represented in some form or another in today’s index. For instance, many of the Regional Bell Operating Companies were rolled up into AT&T or Verizon. International Paper (IP) acquired five index components alone (Champion, Federal Paper, Temple-Inland, James River and Union Camp) and a sixth, Willamette Industries, was acquired by Weyerhaesuer, but now owned by IP.
Amongst those gone, I can account for 18 companies that participated in leverage buyouts. The mid-2000s saw aggressive financing lead to the acquisition of large, once unthinkable acquisitions. Bauch & Lomb, Hilton Hotels, First Data, Columbia/HCA all went private. A few such as Shell and Unilever are gone due to the S&P’s change in focus towards U.S.-only companies. 11 more were acquired by international companies. A few others are no longer in the index like Alco and Limited which split up and created smaller companies.
11% Cumulative Bankruptcy Rate
The surprising thing to me was not the number of companies that are still represented in some form or another, but rather the number of companies from the 1995 S&P 500 that went bankrupt. 40 companies of that group filed for bankruptcy between the end of that year and today. That figure does not include Fannie Mae or Freddie Mac which were insolvent and subject to government takeover, nor does it include Great Western Financial, Providian and HF Ahmanson & Co which were acquired by Washington Mutual which ultimately defaulted. Another handful like Navistar, Albertsons, Crown Cork, check-maker Deluxe Corp, Brunswick, Louisiana Pacific, Rite Aid, Sears, Radioshack, and Yellow Corp which have all come close. One out of every nine of America’s “best” companies from that year defaulted or were close to it in the coming years.
Survivorship-Bias Matters
It makes for an interesting example for the importance of including such companies for historical analysis. The landscape changes considerably year to year. Companies come and go, for good reasons and bad ones. Without a database that is survivor-bias free, you are blind to what happened in the real world.