Study guide
12.2 Indexes
CMT Level I Chartered Market Technician
Track your progress as you read
Finished sections get a check mark.
1 Introduction and Exam Overview
1.1 Current CMT Level I Exam Facts 1.2 Charter Path and Study Plan
2 Theory and History Foundations
2.1 A Brief History of Technical Analysis 2.2 The Dow Theory 2.3 Markets, Instruments, Data, and the Technician
3 Market Hypotheses and Fibonacci
3.1 The Efficient Markets Hypothesis 3.2 The Fibonacci Sequence and the Golden Ratio
4 Chart Types and Construction
4.1 An Overview of Charting 4.2 The X Axis 4.3 The Y Axis 4.4 Charting Volume and Open Interest
5 Trend Analysis
5.1 What Is a Trend 5.2 A Trend's Four Phases 5.3 Trend Identification and Following
6 Volume, Open Interest, and Internals
6.1 Introduction to Volume Analysis 6.2 Volume, Open Interest, and Price 6.3 Market Internals
7 Chart Pattern Analysis
7.1 Classical Chart Patterns 7.2 Introduction to Candlesticks 7.3 Introduction to Candlestick Patterns 7.4 Basics of Point-and-Figure Charting
8 Internal and External Technical Indicators
8.1 Moving Averages 8.2 Technical Indicator Construction 8.3 Introduction to Bollinger Bands
9 Statistics for Technicians
9.1 Introduction to Statistics, Part 1 9.2 Introduction to Statistics, Part 2 9.3 Introduction to Probability
10 Behavioral Finance and Sentiment
10.1 Behavioral Finance 10.2 Market Sentiment and Technical Analysis 10.3 Sentiment Measured from Market Data 10.4 Sentiment Measured from External Data
11 Cycle Analysis
11.1 Foundations of Cycle Theory 11.2 Common Cycles
12 Equities, Indexes, and Fixed Income
12.1 Equities 12.2 Indexes 12.3 Fixed Income and Bonds
13 Futures, ETPs, and Currencies
13.1 Futures 13.2 Exchange-Traded Products 13.3 Foreign Exchange
14 Digital Assets and Options
14.1 Digital Assets 14.2 Options
15 Relative Strength
15.1 Introduction to Relative Strength 15.2 Relative Strength and Its Uses
16 Volatility Analysis
16.1 The Meaning of Volatility to a Technician 16.2 Measuring Historical Volatility 16.3 Options-Derived Volatility
17 Systematic Methods and Ethics
17.1 Introduction to Quantitative Methods 17.2 Code of Ethics 17.3 Standards of Professional Conduct
12.2 Indexes
Key Takeaways Explain This Topic AI
An index compresses a market, sector, or style into one series so a technician can benchmark relative strength, confirm averages, and read breadth against a known tape.
The Dow Jones Industrial Average is a price-weighted index of 30 U.S. stocks; a higher-priced constituent moves the average more than a lower-priced one regardless of company size.
The S&P 500 is a float-adjusted market-capitalization-weighted index of leading U.S. large-cap companies, so larger investable names dominate index return.
An equal-weighted index assigns each constituent the same weight at rebalance, giving smaller members more influence than they have in a cap-weighted benchmark and requiring periodic rebalancing.
Survivorship bias is the error of studying only names that remain in an index or remain listed today, which omits failed, acquired, or demoted companies and overstates historical performance.
Last updated: September 2026
An index is a rule for turning many securities into one number . CMT Level I studies indexes in Section Nine because comparative market analysis needs a benchmark : a stock versus its sector, a sector versus the market, one average confirming another. Independent ExamFinch teaching for this CMT Level I unit stays on three exam jobs: why technicians use indexes, how major indexes are weighted, and what survivorship bias does to anyone who studies only the names that are still there. Advanced Techniques is 26% of the Level I sitting; this is still cash-market work, not yet futures construction.
Value of indexes in technical analysis
A single stock can trend while the crowd does not, or fail while the crowd runs. Indexes are how you see the crowd's tape without plotting every name.
A market proxy. The S&P 500, the Dow Jones Industrial Average, the Nasdaq-100, and the Russell 2000 are short answers to what did U.S. large-caps, 30 industrials, large non-financial growth, or small-caps do today? Intermarket work later compares those series to bonds, commodities, and currencies. You cannot do that comparison from one mid-cap chart.
Relative strength. Much of later Level I comparative work is a ratio : stock / index, sector / S&P 500, Russell 2000 / Russell 1000. The denominator has to be a published index with a known methodology. Absolute trend (price rising) and relative trend (outperforming the benchmark) are different statements. A stock can make a new high and still be a laggard if the index rose more.
Confirmation. Dow Theory, already in the Theory and History domain, used two averages—the industrials and the rails, now the Dow Jones Transportation Average —because one average can lie. The same logic scales: a new high in a cap-weighted large-cap index that is not confirmed by an equal-weight index, a small-cap index, or internals is a narrow high. Breadth tools in the internals unit are how you inspect the list underneath the index.
Communication and testing. Practitioners, journalists, and exam stems name the S&P 500 the way they name a ticker. A standardized history is also what you backtest against—if you use a history that actually reconstructs past membership, not a list of today's survivors. That caveat is survivorship bias, below.
Liquid overlays. Index futures and index ETFs (the next chapter) let traders hedge or speculate on the average itself. Volatility products such as VIX are defined on S&P 500 options; you cannot read that complex if you do not know what the S&P 500 is. This unit does not build VIX. It explains why the underlying index's construction is not a trivia extra.
Exam trap: treating the index level as a price with volume in the equity sense. Some platforms plot index tick volume or the volume of a related future. Cash index levels are calculated . The volume that matters is the volume of the constituents, an ETF, or the future you actually trade.
Weighting methods used in major indexes
Weighting is the rule that decides how much each member moves the average. Two indexes can hold similar companies and tell different market stories if the weights differ. Level I wants three methods by name, with the DJIA and the S&P 500 as the working examples.
Price-weighted: the Dow Jones Industrial Average
A price-weighted index is the sum of the constituents' prices , divided by a divisor . The Dow Jones Industrial Average (DJIA) is the textbook case: 30 large U.S. companies, price-weighted. A $400 stock has four times the influence of a $100 stock, even if the $100 company has a much larger market capitalization.
The divisor is not 30. It is a published number that S&P Dow Jones Indices adjusts so that splits, spinoffs, and substitutions do not, by themselves, gap the average. Without a divisor change, a 2-for-1 split in a $400 member would look like a crash in the Industrials. The divisor is the index-level analogue of the adjusted equity series in the prior section.
Consequences a technician must expect:
A high-priced member dominates day-to-day points even if it is not the largest firm.
A split in a high-priced member reduces that name's future influence once the divisor is reset; the company did not shrink, but its weight did.
Share issuance and buybacks change market cap a lot and the price-weighted influence little, except insofar as they move the price.
Charles Dow's original industrials and rails were also price-weighted averages. The method is old. It is not the method the S&P 500 uses. The Nikkei 225 is another widely watched price-weighted average; if a stem asks for a major price-weighted example besides the DJIA, that is the usual second name.
Capitalization-weighted: the S&P 500
A market-capitalization-weighted (cap-weighted, value-weighted) index weights each member by price × shares . The S&P 500 is the Level I example: a committee-selected index of leading U.S. large-cap companies, calculated by S&P Dow Jones Indices. The usual share count is float-adjusted : only the investable shares (free float) enter the weight. Closely held blocks do not get to dominate the index just because they exist on the cap table.
Weight of stock i ≈ (priceᵢ × float sharesᵢ) / Σ (price × float shares). A megacap can be several percent of the entire index. A smaller constituent can be a few basis points. That is why a handful of names can drive most of a year's S&P 500 return even though about 500 companies are in the list (share-class details can make the name count not exactly 500).
Consequences:
The index is a good proxy for the wealth invested in those stocks: bigger pools of capital move it more.
It is a poor proxy for the typical stock . Equal-weight and small-cap series exist because the median member is not the megacap.
Mechanical cap-weighting adds weight to winners as they rise and reduces weight on losers as they fall. That is not a tactical call; it is arithmetic.
Reconstitution and float updates change weights even when your chart of one name is quiet.
The Nasdaq-100 is a modified cap-weighted large-cap growth index (capping rules exist so that one name cannot run away with the whole product). If a stem says S&P 500, answer float-adjusted cap-weighted , not equal-weighted and not price-weighted.
Equal-weighted
An equal-weighted index assigns each constituent the same weight at each rebalance—about 0.20% each in a 500-name index, before the next drift. Between rebalances, winners grow above equal weight and losers shrink below it; the periodic rebalance sells relative winners and buys relative losers to restore equality.
The S&P 500 Equal Weight Index is the usual comparison series against the headline S&P 500. Equal weight:
Gives smaller members more influence than they have in the cap-weighted parent.
Usually has a different sector mix (in recent years, less concentration in a few Information Technology megacaps).
Needs higher turnover and has more exposure to less liquid names , which matters for capacity and for how clean the chart is.
Equal weight is not a claim that every company is equally important to the economy. It is a claim that the technician wants the typical constituent , not the capitalization-weighted wealth portfolio.
Worked three-stock comparison
Three names, invented so the arithmetic is visible. Ignore float for the toy cap-weight (treat all shares as investable).
- Stock Price Shares Market cap Price-weight (price / $550) Cap-weight (cap / $1,600) Equal-weight
- A $400 1 million $400 million 72.7% 25.0% 33.3%
- B $100 10 million $1,000 million 18.2% 62.5% 33.3%
- C $50 4 million $200 million 9.1% 12.5% 33.3%
Stock A dominates the price-weighted average because it is expensive per share. Stock B dominates the cap-weighted average because it is the largest firm. Equal weight refuses both stories and listens to A, B, and C the same at rebalance. If a stem gives prices without shares, you cannot compute cap weights. If it gives caps without prices, you cannot compute DJIA-style weights.
- Method What gets the most influence Major example Mechanic a technician must remember
- Price-weighted Highest price DJIA (30 stocks); Nikkei 225 Divisor absorbs splits and substitutions
- Float-adjusted cap-weighted Largest investable cap S&P 500 Megacaps dominate; float, not full shares
Equal-weighted Each name the same at rebalance S&P 500 Equal Weight Rebalance restores equality; more small-name influence
Survivorship bias
Survivorship bias is the error of studying only the securities that survived some filter—still listed, still in the index, still large enough—while ignoring the ones that failed, were acquired, were demoted, or were delisted . The sample of survivors is not the sample of everything that was tradable at the time.
How it shows up in index work:
Building a 20-year chart of today's S&P 500 members and treating that basket as the S&P 500 of 2006. Companies that went bankrupt, were bought out, or shrank out of the index never appear. The backtest overstates returns and understates the pain of names that died.
Studying classical patterns only on current large-caps that look good on a 10-year log chart. The graveyard of failed breakouts is not on that screen.
Quoting the average performance of funds that still exist after a decade, without the funds that closed.
Official index total-return series from a serious provider typically reconstruct historical membership : when a company left in 2009, it is in the 2008 return and out of the 2010 return. That is not survivorship-free in a philosophical sense—the index itself kicks out failures—but it is the honest history of that rule set . The bias appears when you rebuild the rule set from the ending list.
Related traps:
Look-ahead membership. Using a constituent list that would not have been known on the historical date.
Delisting returns. Ignoring the last gap to zero or to the acquisition price.
Index additions. Names are often added after a strong run; a naive study of adders going forward is a different question from a study of the index's own published history.
The quantitative-methods unit later in Advanced Techniques returns to survivorship bias in testing . The definition to memorize here is already enough to fail a stem that treats today's 30 DJIA names as the only industrials that ever mattered.
Exam habits for this unit
Name the method before you interpret the chart. A new high in the cap-weighted S&P 500 can be a megacap event; check equal-weight or a mid-cap index before you call it a broad tape. If a question gives a toy table of prices and shares, compute both price weights and cap weights—the contrast is the point. If a question describes a study of current members' past 20 years, name survivorship bias .
- Key Takeaways
- Indexes compress a market or slice of a market into one series for RS, confirmation, and breadth
- DJIA: 30 names, price-weighted, divisor-adjusted for splits and substitutions
- S&P 500: float-adjusted cap-weighted large-cap U.S. companies
- Equal weight: same weight at rebalance; more small-name influence; needs rebalancing
- Survivorship bias: studying only who is still in the index overstates the historical record