Week 9: Market Indexes — How the Scoreboard Is Built

---

1. Why This Is Important

Every business newscast frames the day in index points. "The S&P 500 was up half a percent. The Russell 2000 was down 1%. Tech led on the Nasdaq 100." Those four or five numbers tell millions of investors how their portfolios "did" — without anyone explaining what an index actually is. An index is not a market. It is a recipe: which stocks to include, how much each one counts, and when to swap names in and out. Two recipes pulling from the same kitchen can produce wildly different dishes.

There are four reasons every adult investor should be index-literate.

You almost certainly own one already. Roughly $15 trillion sits in vehicles that just buy what an index tells them to buy. The S&P 500 alone has more than $11 trillion benchmarked or directly indexed to it. The fee on your 401(k) target-date fund, the ETF in your brokerage, the closet index your "active" mutual fund secretly hugs — all live or die by some index's rules. Owning an index without understanding the rules is like owning a restaurant without ever reading the menu.

The recipe quietly chooses your factor exposure. Cap-weighting tilts you toward whatever has already won. Equal-weighting tilts you toward smaller, cheaper names. Price-weighting (the Dow) tilts you toward whichever stocks happen to have a high nominal share price — a coin-flip from corporate splits decades ago. None of these are wrong, but they are not interchangeable. A 60/40 portfolio benchmarked against the wrong index can look like a hero or a clown for purely mechanical reasons.

Index plumbing creates real flows. When a stock is added to the S&P 500 or the Russell 1000, every passive fund tracking the index has to buy on a specific day. When it is removed, they have to sell. Those forced flows reliably move prices a few percent around the rebalance — and the Russell's late-June reconstitution is one of the largest scheduled trading events on earth. If you trade around index changes, this matters; if you do not, it is still useful to know that the spike on rebalance day is mechanical, not informational.

The honest version of investing accepts the index. The starting principle of this course is that genuine alpha is rare and hard to keep. The S&P 500 net of fees has beaten 80–90% of large-cap active managers over fifteen-year windows. That is not because the index is brilliant. It is because the active managers' fees and turnover compound into a headwind they cannot beat. The default move for most readers of this course is to own the index, hold the four tranches (lesson 13), and spend the saved decision-budget on tax structure (lesson 15) rather than stock picking. To do that with eyes open, you need to know what is actually in the box.

This week is the box.

---

2. What You Need to Know

2.1 Three Weighting Schemes — One Universe, Three Different Indexes

The single most important design choice in any index is how it weights its members. The same 30 stocks can produce three very different index series.

• Price-weighted. Each stock's weight is its share price divided by the sum of share prices. The Dow Jones Industrial Average is the only major index still built this way. A $400 stock counts ten times more than a $40 stock, even if the $40 stock is a much larger company. Splits, buybacks, and the choice of nominal share price (which is a marketing decision, not an economic one) all distort the result.
• Cap-weighted. Each stock's weight is its float-adjusted market cap divided by the index's total. This is the S&P 500, the Nasdaq Composite, the Russell 1000 / 2000 / 3000, the MSCI ACWI — essentially every index that runs a real fund business. The logic is sound: if NVIDIA is worth $3 trillion and a small industrial is worth $30 billion, NVIDIA represents a hundred times more of the equity-economy and should have a hundred times the weight. The cost is concentration. By April 2026, the top 10 names in the S&P 500 sum to about 33% of the index, and the Magnificent Seven alone are over 30%. You are not buying "500 stocks" in any meaningful sense; you are buying seven stocks plus 493 satellites.
• Equal-weighted. Every member gets 1/N weight on rebalance day. The S&P 500 Equal-Weight Index (tracked by the RSP ETF) is the canonical version. Equal-weight automatically tilts toward smaller and cheaper names, since the largest names are dragged down to 0.2% and the smallest are pushed up to 0.2%. Over very long horizons it has out-earned the cap-weighted version by roughly 0–2% per year, but at the cost of more turnover, more tax drag in a taxable account, and meaningful periods of underperformance (notably 2017–2024, when mega-cap tech crushed everything else).

The cumulative-wealth chart for the cap-weighted vs equal-weighted S&P, 1928 to 2024, is in image/week09_cap_vs_equal.png. Both lines climb. Equal-weight ends a little higher. Neither dominates in every decade.

2.2 The Mag-7 Concentration Problem

In April 2026 the S&P 500's top ten weights look approximately like this:

• Apple ~6.8%
• Microsoft ~6.4%
• NVIDIA ~5.6%
• Amazon ~3.7%
• Meta ~2.7%
• Alphabet (GOOGL+GOOG) ~3.7% combined
• Berkshire Hathaway ~1.7%
• Tesla ~1.5%
• Broadcom ~1.5%

That is roughly 33% of the index in ten companies, with the Mag-7 (Apple, Microsoft, NVIDIA, Amazon, Meta, Alphabet, Tesla) at about 30%. It is the highest concentration the S&P 500 has carried since the late 1960s' "Nifty Fifty" peak. The bar chart in image/week09_top_concentration.png shows the shape: a steep cliff for the top 10, then 490 stocks splitting the remaining two-thirds.

This is not automatically a problem. It is the index doing exactly what cap-weighting tells it to do: let winners compound. But it has three practical consequences.

Your equity sleeve is a Mag-7 bet. If you hold VOO or SPY and nothing else, roughly a third of your equity dollar is in seven tech-adjacent megacaps. The diversification narrative ("500 companies!") understates the actual factor risk.

Reversals concentrate. The 2022 drawdown was driven mainly by the same top names that drove the 2020–2021 rally; equal-weight fell less. The 2000–2002 dot-com unwind is the textbook precedent.

Equal-weight is now the obvious diversifier. Holding RSP alongside VOO does not change your sector exposure dramatically, but it does cut your top-10 concentration from ~33% to ~17%. This matters more for retirees in the spending phase than for younger accumulators.

2.3 Free-Float Adjustment — Why "Market Cap" Isn't Quite Market Cap

Every modern cap-weighted index uses the float-adjusted market cap, not the total market cap. Float is the count of shares actually available to public trading — total shares outstanding minus blocks held by founders, governments, parent companies, and other strategic holders.

A few examples, April 2026 estimates:

• Meta has two share classes; Class A has ~99% float, but Mark Zuckerberg's super-voting Class B is locked up. The S&P uses Class A only.
• Alphabet's GOOGL (Class A) and GOOG (Class C) trade publicly; the Class B held by founders is excluded. The two listed classes are both in the index.
• Berkshire Hathaway's float adjustment trims the index weight by a few hundred basis points relative to its total market cap, because Buffett's holdings are not deemed available.

The reason for float adjustment is liquidity, not fairness: an index that asks tracking funds to buy 100% of total shares would force them to demand stock that is not for sale. Float-weighting matches the "what can I actually buy?" question that index funds need to answer.

2.4 The Russell Reconstitution — A Real Trading Day

The Russell index family (Russell 3000 / 1000 / 2000) reconstitutes once a year, on the last Friday of June. The cutoff for which company sits in which size bucket is rank order on the prior May 31, with a buffer band to limit churn. Between the buffer-day announcement (mid-June) and the Friday close, two things happen:

Hedge funds estimate which names will be added to the Russell 2000 and which will be promoted to the Russell 1000, and they bid those names up.

On the rebalance day itself, every passive Russell fund — and every "closet Russell" active fund worried about benchmark deviation — must execute the change. Single-day turnover in some affected names exceeds 30% of average daily volume.

This is the largest scheduled trading event in U.S. equities. It is also one of the cleanest examples of "irrational > solvent": the price moves are mechanical, but they can persist for weeks because shorting them costs borrow fees and mark-to-market pain. The S&P 500's quarterly rebalances and the Nasdaq 100's annual rebalance are smaller but rhyme.

2.5 Survivorship Bias — What the Index Throws Away

Indexes silently drop names. A company that goes bankrupt, gets acquired, or falls below the size threshold is removed; a fresh name takes its slot. The historical index series — what you see in the chart books — is a series of survivors plus replacements. The dead stocks are gone.

For long-run return claims this matters more than people realise:

• The "stocks return 9–10% per year since 1928" headline already excludes the railroads that vanished in the 1930s, the conglomerates that imploded in the 1970s, and Enron, WorldCom, Lehman, GE-as-was. Their full drawdowns are in the historical record but their post-failure performance (zero) is not blended into a "what if I had bought everything in 1928" series.
• Empirical studies put the survivorship overstatement at roughly 1–2% per year for the longest-running indexes. The Damodaran dataset used in Week 3 already corrects for this on the broad-market level, which is why its returns are a little lower than the S&P-500-since-inception headline.
• For individual stock-picking decisions, the bias is much worse. A "buy what worked in 2000–2010" strategy back-tested on today's index members will look heroic, because the actual losers from that decade aren't in the test universe at all.

The practical takeaway: always ask what got dropped, especially when someone shows you a back-test. The principle again — a clean back-test is much rarer than a clean look.

2.6 The Big Indexes — A Field Guide

• S&P 500. ~500 large-cap U.S. companies, committee-selected on profitability and float thresholds, float-adjusted cap-weighted, quarterly rebalance. Roughly 80% of U.S. investable equity by market value. Tracked by VOO (Vanguard, 0.03% expense ratio), SPY (State Street, 0.0945%), IVV (BlackRock, 0.03%). The default U.S. equity exposure.
• S&P 500 Equal Weight. Same 500 names, equal weights, rebalanced quarterly. RSP (Invesco, 0.20% expense ratio). About 7× the turnover of the cap-weighted version, which translates to a small but real tax drag in a taxable account.
• Russell 2000. The 2,000 stocks ranked roughly 1,001–3,000 by U.S. market cap. The canonical small-cap benchmark. IWM (BlackRock, 0.19%) is the main vehicle. Earnings quality is dramatically lower than the S&P 500 — a sizable share of Russell 2000 companies have negative net income — which is partly why small-cap returns have lagged large-cap since 2014.
• Nasdaq 100. The 100 largest non-financial Nasdaq-listed stocks, modified-cap-weighted with a re-weighting rule that caps individual names so a single stock cannot drift past ~24%. QQQ (Invesco, 0.20%). Heavily tech-tilted, but it is not a tech index; it is a Nasdaq-listing index.
• Dow Jones Industrial Average. 30 stocks, price-weighted, committee-selected. Cultural artifact more than a serious benchmark. DIA (State Street, 0.16%). Not recommended as a portfolio building block.
• MSCI ACWI. ~3,000 large- and mid-cap stocks across 47 developed and emerging markets, float-adjusted cap-weighted. ACWI (BlackRock, 0.32%) is the main ETF. Caveat: for a Hong Kong / Mainland China reader of this course, the only investable slice of ACWI is the ~60% U.S. weight plus the developed-non-U.S. component you can hold via the ETF. Direct exposure to A-shares, mainland developers, and most EM-listed names is either capital-controlled or operationally hostile. The course's default is to overweight U.S.-listed exposure, hold ACWI as a small diversifier if at all, and not try to engineer a "global cap-weighted" portfolio that the plumbing won't actually deliver.

2.7 Index Futures — The Liquid Truth

If you want to know where the index "really is" between cash-market opens, look at the front-month futures: ES (E-mini S&P 500), NQ (E-mini Nasdaq 100), RTY (E-mini Russell 2000), YM (E-mini Dow). They trade roughly 23 hours a day at tens of billions of notional. ETFs can trade away from net asset value briefly during stress; futures cannot drift far from the index without arbitrage closing the gap.

For a long-term reader, the reason to know this is not to trade futures (you almost certainly should not). It is to understand that:

• Pre-market price moves quoted on financial TV are usually the front-month future, not the index.
• During a crisis (March 2020, August 2024), the futures often print prices that the cash ETFs only catch up to at the next open. The futures are not lying — they are just open while the underlying is closed.
• For very large allocations (institutional scale), futures plus T-bills are often a cheaper way to hold S&P 500 exposure than holding SPY, because the financing rate embedded in the future is sometimes below the all-in cost of the ETF. The tax-via-options/margin angle sits on the same hook.

2.8 Build Your Own Index — Try the Interactive

The interactive demo for this lesson, interactive/week09_index_builder.html, is a sandbox. It gives you 30 representative S&P 500 components with realistic prices, share counts, and trailing 12-month returns. You toggle between cap-weighted, equal-weighted, and price-weighted, and watch:

• the composition pie redraw (concentrated for cap, flat for equal, price-distorted for price-weighted), and
• the 12-month index return recompute.

The point is to see, with your hands on the dials, that the same 30 underlying stocks produce three different return numbers — sometimes diverging by several percent — purely because of the recipe. That is the lesson.

---

3. Common Misconceptions

"The Dow tells me what the market did." The Dow is 30 names, price-weighted by an arbitrary nominal-price quirk. On any given day it can disagree with the S&P 500 by 50+ basis points for purely structural reasons. Treat it as a cultural data point, not a portfolio benchmark.

"The S&P 500 is diversified across 500 stocks." It is diversified across the names of 500 stocks. By April 2026 it is concentrated in the Mag-7 to the tune of ~30%, with the top 10 at ~33%. Your equity dollar is doing a lot less spreading than the headline implies.

"Equal-weighting always beats cap-weighting." Long-term yes by a small margin, but with multi-year stretches of underperformance — most recently 2017–2024 — and higher turnover. It is a different bet, not a free lunch.

"Index funds are passive." The fund is passive to its rules, but those rules are a committee's active choice. The S&P committee actively decides who gets added. Russell's reconstitution is mechanical but the timing creates predictable trading. There is no truly passive index — there is only "the active decisions are baked in upstream."

"Adding to the S&P 500 is just paperwork." Empirically the announcement causes a 3–8% pop in the added stock between announcement and effective date, sustained for weeks. Deletions cause a corresponding drop. This is one of the most reliably documented anomalies in equity finance.

"NASDAQ = tech." NASDAQ is a listing exchange. The Nasdaq 100 is heavily tech because tech companies historically chose to list there, and they have grown into the size that dominates a cap-weighted index. PepsiCo and Costco are also Nasdaq-100 names.

"Survivorship bias only applies to active funds." It applies to the indexes themselves. The 1928 S&P series silently substitutes new names for old ones; the dead names' post-failure trajectory does not appear.

"MSCI ACWI gives me real global exposure." It gives you global exposure in a backtest. Operationally, for a non-U.S.-resident investor working through a Hong Kong or Singapore broker, large parts of the EM weight are not easily, cheaply, or tax-efficiently held. Only what is actually investable counts.

"Float adjustment is a minor technicality." For names with concentrated insider holdings (META, BRK, founder-led companies) it is the difference between a 4% index weight and a 1.5% index weight. It is not minor.

"My active manager beats the index." They might, in any given year. Over fifteen years, after fees, roughly one in seven large-cap active U.S. equity funds beat the S&P 500. The base rate on alpha is bad — that is the alpha-is-rare principle, top of the list.

---

4. Q&A Section

Q1. Should I own RSP instead of VOO to avoid the Mag-7 concentration? A. Probably not the whole sleeve. A common compromise is 70–80% VOO plus 20–30% RSP, which trims your top-10 concentration from ~33% toward ~25% without giving up the natural cap-weighted compounding. RSP also has 7× the turnover and an extra 17 bps of expense ratio. In a tax-deferred account the cost is small; in a taxable account it eats into the diversification benefit.

Q2. Why do active managers struggle to beat the S&P 500? A. Three structural reasons. First, fees: average large-cap active is ~70 bps versus 3 bps for VOO, so the manager is starting 67 bps behind every year. Second, the index's cap-weighting lets winners compound automatically; managers tend to trim winners (rebalancing discipline) and reinvest in laggards, which underperforms in trending markets. Third, the index does not pay capital gains taxes; the active fund's turnover does. Compounded over fifteen years, the gap is enormous.

Q3. What happened to "the next Apple" stocks that didn't make it? A. They left the index. Polaroid, Eastman Kodak, Sears, Lehman, GE (downsized out), Bear Stearns, MCI, Enron, WorldCom, Pacific Gas (in and out), AIG (in 2004, out in 2008, back later) — every one of those was once a multi-percent S&P 500 weight. Their losses are reflected in the historical series only up to the day they left. After that, the slot is taken by a fresher name and the dead one's continued zero is not in the line.

Q4. Can I invest directly in the S&P 500 itself? A. No. The index is a calculation. You can buy ETFs (VOO, SPY, IVV) that hold the underlying stocks, mutual funds that track the index (VFIAX), or futures (ES) that settle to the index. All three are wrappers around the same recipe; the wrapper you choose is a fee, tax, and access decision.

Q5. How does the S&P 500 committee decide additions? A. Eligibility rules — U.S. domicile, market cap above the floor (~$18 billion in April 2026), at least four consecutive quarters of GAAP profit, public float ≥ 50% — get you onto the candidate list. From there the committee picks based on sector balance and replacement need (a deletion creates the slot). The committee's discretion is the reason Tesla took years longer than expected to be added, and is the reason the Nasdaq Composite (rules-based on listing exchange) and Russell 1000 (pure market-cap ranking) move differently from the S&P at the margin.

Q6. Why isn't the Hang Seng Index in this lesson? A. The investable-universe constraint. For a Hong Kong / mainland reader of this course, Hang Seng exposure via 2800.HK is technically holdable, but the underlying is dominated by mainland-listed banks and developers whose accounting and political risk are categorically different from anything U.S. investors price. The course's stance is: if you live in HK, buy a small Hang Seng tracker for currency-of-spending hedging, but do not pretend it is a substitute for the U.S. index. The S&P 500 is the engine; HK exposure is hedging the gas tank.

Q7. What is the Mag-7 weight likely to be in five years? A. Nobody knows, and that is the honest answer. Two historical analogies: the Nifty Fifty top weight peaked in 1972 at a similar concentration and was cut roughly in half by 1974–75; the dot-com mega-caps peaked in early 2000 and the index spent a decade rotating into other names. Both unwinds were painful. A third possibility — the one cap-weighting bets on — is that AI productivity gains keep these companies' earnings growth ahead of the rest of the index, in which case the concentration grows. Hold both VOO and RSP if you don't want to bet on which.

Q8. How do I benchmark a 60/40 portfolio? A. The cleanest benchmark is 60% VTI (or VOO) and 40% AGG (or BND), rebalanced quarterly. You compute that benchmark's return, then compare to your actual portfolio. If you want to be more sophisticated, blend in a small International allocation (5–15% of the equity sleeve via ACWX) to match what you actually hold. Comparing a 60/40 against 100% S&P will make you feel terrible in good equity years and falsely heroic in bad ones; that is the wrong benchmark.

Q9. Why don't professional traders just trade the Russell rebalance? A. They do, and the trade is well-known enough that the easy money has been arbitraged out. The remaining risk premium is for taking bidirectional inventory (you don't know exactly which names will be added until two weeks before, and you carry borrow costs to short the deletions). Hedge funds with the right infrastructure still earn a few hundred basis points a year on this — but they pay for it with risk and complexity that retail traders cannot replicate.

Q10. What is the single most important index for me to understand? A. The S&P 500. It is the benchmark for ~$11 trillion of U.S. equity capital, the default in every 401(k), the comparison every active fund is judged against, and the cleanest expression of the alpha-is-rare principle: own the index, accept the market return, and spend your scarce decision budget elsewhere — on the four tranches, on tax structure, on not panicking. Everything else in this lesson is texture around that core fact.