Editor's Note: This article was prepared jointly by Julius de Kempenaer and Arthur Hill with assistance from members of the StockCharts data and business teams.
You may have noticed that in September, a lot was written about the changes in sector classifications that were implemented after the close on the 21st by S&P Dow Jones Indices, the company that maintains and publishes the underlying indices for the popular range of SPDR sector ETFs (XLF, XLK, etc.).
Changes to indices are not uncommon and happen on a regular basis (quarterly). Depending on the guidelines and the index methodology "rule book", companies will be added to or deleted from an index to better reflect the characteristics of the index as described in the rule book. Market capitalization and company activities are the major factors that determine in which sector a stock is classified.
This time, however, the changes not only included adding stocks to or deleting them from various sector indexes, but also an overhaul of the classification structure to better reflect the "new world order, which emerged over the last twenty years. Many of these "new world order" stocks did not exist during the dot com bubble of 2000, but they have gone on to become dominant mega-caps. Think Google, Facebook, Netflix and Twitter.
The GICS classification system consists of four levels:
- 11 - SECTORS
- 24 - INDUSTRY GROUPS
- 69 - INDUSTRIES
- 158 - SUB-INDUSTRIES
The methodology for classification is bottom-up. This means each stock gets classified to a "sub-industry" based on its principal business activity. Revenues are the key factor determining the principal business activity.
From the sub-industry level, a stock will eventually end up in a sector. The 158 sub-industries are rolled up into 69 industries. These 69 industries are then rolled up into 24 industry groups and these are ultimately be rolled up into the 11 sectors at the top of the pyramid.
Hence, once a stock is assigned to a sub-industry you will know immediately to which industry, industry group and sector it belongs.
For every classification level, index-values are calculated and published, and historical data is available since the inception of the GICS system.
Now be careful, these indexes are NOT INVESTABLE or TRADABLE products.
For example, XLK, the "Technology Select Sector SPDR® Fund" is not the same as $IXT, the "Technology Select Sector Index". XLK, an ETF, is an investable product that closely tracks an underlying index, which in this case is $IXT. They are close, but they are not exactly the same.
Numerous asset managers have created investable products based on (GICS) indices. The higher up in the classification pyramid the more products are available. The 11 SPDR Sector ETFs have become so popular and widely traded that they are now often seen as "the sectors", which they are but then they are not. Based on the management fee alone an ETF will always, marginally, underperform its underlying index.
Creating Historical Data
For the SPDR family of sector ETFs, there is plenty of data available for analysis as the ETF and the underlying index were launched at the same time or nearly the same time. This is not always the case. When a provider launches a new product based on an index that exists for many years already, there will be very little data available for the tradable product.
The changes in sector indexes and ETFs for Real-Estate in 2015 and now Communication Services in 2018 required attention because there was very little historical data.
Lack of data does not have to be a show-stopper as we can use the underlying index data for analysis and trade the newly launched product. Ideally, you want to analyze like for like so either you analyze investable products like ETFs or you analyze indexes. Mixing them could cause unexpected problems. For example, one of the two could be a total return series and the other not.
But what if a completely new index is launched together with a tracking ETF? Neither will have data and this makes historical comparisons impossible. And historical comparisons are exactly what we as technical analysts live by.
S&P, some math, and pragmatic (creative?) thinking to the rescue!
Let's start with the "easy" one. In 2015, S&P and MSCI decided to split the GICS financial sector into two new sectors – Financials and Real Estate. At the same time, a sector SPDR for the new Real Estate sector (XLRE) was launched by State Street.
Because Real Estate was an industry group under Financials, this new sector was essentially carved out from the Financials Sector and promoted one level higher in the GICS hierarchy. The historical data for the "new" sector was therefore already available.
At the index level, the historical data for the Real Estate Industry Group is the same as the data for the Real Estate Sector, the split was made at the third GICS level. The Real Estate Industry Group consisted of three industries; (Equity) REITs, Real Estate Management and Development, and Mortgage REITs.
The new Real Estate Industry Group holds only Equity REITs and Real Estate Management and Development. Mortgage REITs stayed with Financials Sector under the industry group Diversified Financials.
In the StockCharts database, we have copied the level-2 index data for Real Estate over to level-1, the Real Estate Sector Index. This now makes it possible to use the Real Estate Sector data in historical comparisons at the index level.
In order to be able to use the investable XLRE ETF in historical comparisons, we have taken the performance data for the Real Estate Index and simulated the performance of XLRE as if it was trading before October 2015 and back-filled the history of XLRE prior to introduction, assuming a 100% correlation.
Chartists can now use XLRE in longer-term charts, compare historical sector performance and use XLRE in tools like PerfCharts.
Note: the back-filled data prior to the start of the real estate as a sector includes mortgage REITs. As real estate used to be part of financials, there is an overlap between financials and real estate prior to the launch as a real sector. The history for XLF nor the underlying index has been adjusted.
From a historical data perspective, the introduction of the new Communication Services sector is a bit more complex. It is a completely new sector and also a dissemination of the old telecom sector. Unlike the Real Estate Sector, there is no historical data whatsoever for the Communication Services Sector.
Here, you can read a good piece on the "why and how" behind the introduction of this new sector as well as the impact on the other sectors. Some big names are changing from one sector to another.
In order to facilitate the financial markets and enable analysts to use historical data for this new sector, S&P Dow Jones Indices calculated the historical performance for the Communication Services Sector based on the current index methodology for the sector. This means they went back in time with a rule-book and created the index-composition at each review date, and then calculated the historical values of the new index.
This newly created data-set is available here.
In order to facilitate the use of XLC in long-term charts and other tools on the website, we have taken a similar route as described for XLRE. We have calculated the performance of the new sector index $IXC, based on the data provided by S&P, and used this performance data to backfill the historical prices for XLC, assuming a 100% correlation with the underlying index.
Chartists can now use XLC and XLRE in longer-term charts, compare historical sector performance and use these tickers in tools like PerfCharts and Relative Rotation Graphs.
Ticker Symbols Affected
- XLRE (history already added)
- $IXRE (history will be added soon)
- XLC (history already added)
- $IXC (history will be added soon)
Thanks for reading our explanation of these recent changes and how the sector additions have been handled on StockCharts. Remember, the SPDR sectors are an important component of our site organization and market analysis tools. Be sure to check out the following pages to view the sector and industry classifications.
- The StockCharts Team