Lin's technology-stock portfolio returned 18% over a year, while a broad-market index gained 12%. She writes, “I beat the market by 6 percentage points. My stock selection seems to work.” The subtraction is correct. The conclusion may go beyond what it establishes.
Suppose the technology sector gained 22% that year. The result now has another interpretation. Should she compare with the broad market or the sector? Choosing whichever makes her look better would miss the point. First, understand what each comparison represents and which question her 18% is meant to answer.
What you will be able to do
- Distinguish an index, a fund that tracks an index, and a benchmark used for evaluation.
- Explain simple weighting rules and identify concentration hidden inside overlapping funds.
- Save a benchmark-selection record chosen in advance, with consistent return conventions and a policy for changes.
The required skill here is constructing a fair comparison. You only need an initial understanding of ETF prices, net asset value, and tracking differences. Real funds provide documents to inspect, but the 18%, 12%, 22%, and all subsequent numerical examples are original synthetic inputs. None is a statement about the actual performance of the named funds.
An index is a rulebook; an ETF is a product
An index combines changes in a group of securities into a calculated measure. Its rules determine which securities qualify, their weights, and when constituents or weights change. A news report saying “the market rose” is usually using an index to represent a particular part of a market. You must check that scope; “U.S.” does not mean every U.S. asset.
Think of the methodology as a scoring system. Even with the same companies, different weights can produce different scores. Unlike a casual scorecard, an index must also maintain constituents and handle corporate actions according to its methodology. You cannot freeze today's list permanently and invent a historical series from it. Read the provider's rules rather than relying on an abbreviation. [1]
An exchange-traded fund, or ETF, is a fund product whose shares trade on an exchange. You own fund shares, with investments specified by its documents. Some ETFs track indexes; others are actively managed, and some use leveraged or inverse designs. The product form alone does not establish diversification or suitability for long-term holding. The index itself remains a calculation that you cannot buy directly. [2]
A benchmark is the reference chosen to evaluate a result. An index can serve that purpose. An investable fund can also represent an alternative, but then product expenses and implementation conditions become part of the comparison. Decide whether you are evaluating market exposure, investment management, or an available alternative. Those questions do not always require the same reference.
The same stocks can produce different returns
Suppose companies A, B, and C have eligible starting market values of 600, 300, and 100, totaling 1,000. Ignore new issues, dividends, rebalancing, and fees for this period. Weighting by those values gives 60%, 30%, and 10%. If A gains 10%, B is unchanged, and C loses 10%, their contributions are positive 6, zero, and negative 1 percentage point. The result is 5%.
Now start with one-third in each company. The same individual returns produce 0% for the period. No company has changed; A simply has more influence in the first portfolio. Knowing that two portfolios “hold the same three stocks” does not establish that they have the same exposures or will have the same result.
These market values are simplified teaching inputs. Real methods may use float-adjusted market capitalization, counting the portion available for public trading under the methodology, or impose weight caps. Equal weighting generally means restoring equal weights at specified rebalancing points. Between them, price movements change the weights; the portfolio does not remain exactly equal at every moment. [1]
Read three connected parts of an index: eligibility, weighting, and maintenance. An identical sector label does not establish an identical company list. Similar lists do not establish equal weights. Those differences help explain why two apparently comparable lines diverge, before attributing the gap to skill or a mistake.
Inspect two real documents before looking at rankings
The official SPY product page identifies the S&P 500 as its benchmark and describes U.S. large-cap exposure and float-adjusted capitalization weighting. The official XLK page identifies the Technology Select Sector Index and exposure to the technology sector within the S&P 500. These are examples for identifying product scope, not recommendations to buy either fund. [3] [4]
SPY does not represent global equities or U.S. companies of every size. XLK does not necessarily include every business that someone informally considers technological. A sophisticated product does not guarantee classification in the index's information-technology sector. Replace that impression with actual constituents and classification rules before deciding what the product represents.
Find the investment objective, full benchmark name, current holdings, sector weights, expenses, and risk statements. Save a date for each document. Holdings change, fees may change, and returns, holdings, and NAV figures on one page may have different as-of dates. This chapter does not provide a current fee to memorize. It asks you to preserve the figure and version actually used in your investigation.
Owning more funds can also conceal repeated exposure. Suppose half your money is in broad fund A and half in sector fund B. A company is 6% of A and 20% of B. Its direct exposure in your combined portfolio is half of 6%, plus half of 20%: 3% plus 10%, or 13%. Adding a second product has not removed this company's concentration.
That calculation assumes direct holdings in the same company at the same date. Real investigations must reconcile security identifiers, different share classes, holding dates, and structures such as derivatives. Mark unclear portions unknown. The number of securities mentioned in marketing material cannot, by itself, establish that material concentrations have disappeared.
Your transaction price can differ from the fund's value
An ETF's per-share net asset value, or NAV, is assets minus liabilities divided by shares outstanding. Its exchange price can differ from that value. Operating expenses, bid-ask spreads, and possible taxes also affect an investor's outcome, even when a broker charges no commission. Distinguish NAV performance from the price at which your own transaction can occur. [2]
For a synthetic example, NAV is $100 while you pay $101 for one share: a 1% premium relative to NAV. Later, NAV remains $100 and the market price returns to $100. Your share has lost $1 in market value even though the underlying NAV has not fallen. Ignore fees and assume comparable valuation times. In real work, an intraday reference and end-of-day NAV are not simultaneous observations that can always be subtracted directly.
Suppose a fund's NAV total return is 7.7%, against an index total return of 8% using the same convention. The simple difference is negative 0.3 percentage point. This is a tracking difference for that period; it does not establish that the entire shortfall came from the management fee. Sampling, cash, trading, and other implementation effects can contribute.
Tracking error usually concerns variation in relative returns across multiple periods. A single period's tracking difference is not the same measure. The distinction matters because one number describes a gap, while the other describes how relative returns fluctuate. Do not relabel a convenient single-period subtraction as a statistic that requires a series.
Choose the ruler for Lin's 18%
Return to Lin with a defined task: selecting stocks within the U.S. technology sector. Her portfolio holds only U.S. technology common shares, uses no leverage, and is measured in dollars. Before researching, she chooses a sector index matching the investment universe and relevant weighting characteristics as the primary benchmark. A U.S. large-cap index supplies additional context. She records both roles before observing performance.
She must also align dates, currency, dividends, costs, and external cash flows. A price index records price changes under its rules. A total-return index also handles dividend reinvestment under its particular methodology, with gross and net versions requiring attention. Comparing a dividend-inclusive portfolio with a price-only index mixes conventions into the apparent advantage. [1]
Now assume the 18%, 12%, and 22% use the same period, dollars, and dividend-reinvestment convention, and are all before tax. The portfolio's 18% is after its trading and management costs. Both indexes are theoretical, not directly investable, and have no product fees deducted. There are no external deposits or withdrawals. Relative to the broad market, the portfolio is ahead by 6 percentage points; relative to the sector, it is behind by 4.
Disclose the cost difference. This is not a fully equivalent comparison between two investable products after all their costs. The calculation establishes outperformance of one reference and underperformance of another under the stated conventions. It still does not establish stock-selection skill from a single year: different individual-stock, size, or other exposures and chance may matter. A later chapter examines risk compensation and Alpha; this one gets the comparison onto a defensible footing.
Change the question to “Was concentrating this household money in technology worth it?” A broad-market alternative may now be a more important opportunity-cost reference. The sector comparison cannot erase the option to diversify. Benchmark choice follows the research question, rather than a universal rule that comparison within the same sector is always sufficient.
If the question changes, update the benchmark record with a date and reason, preserving earlier comparisons. Do not exchange primary and secondary references after seeing which is flattering. A new question can justify a new evaluation; it cannot justify rewriting what the old result meant at the time.
Reference answers · reveal
Answer first, then inspect the reasoning
- Why can market-capitalization and equal-weight versions of the same stocks have different returns?
- An author compares a dividend-inclusive portfolio return with a price index and declares success. What should be checked first?
- Two funds each hold half your money. The same company represents 8% of one and 12% of the other. What is its combined exposure?
For the first, each stock's contribution depends on its weight as well as its return. Equal weights also drift after prices move and must be restored according to a rebalancing rule. Merely listing the same names does not describe an identical portfolio. For the second, align dividend treatment first, then dates, currency, costs, and cash-flow treatment. Do not reach a skill conclusion before repairing the comparison.
For the third, half of 8% is 4%, and half of 12% is 6%, totaling 10%. This checks whether you can look through two product wrappers to the underlying company. You cannot add the two holding percentages directly, because each applies only to the money invested in that fund, rather than to the entire portfolio.
Change the research task
You originally study a U.S. large-cap stock portfolio and select a U.S. large-cap total-return index in advance. Next quarter, the strategy moves half its money into short-term Treasury securities. The author retains the all-equity benchmark and claims excellent risk control when stocks fall. Would you retain, replace, or add a benchmark? Write a reason before reading on.
A useful response preserves the old benchmark as the original commitment and an equity opportunity-cost comparison. From the strategy change onward, add a blended benchmark matching the new allocation. Specify stock and bond weights, index versions, and rebalancing rules. Do not pretend the blended reference had always been in use, and do not infer stock-selection skill merely because half the portfolio held bonds.
The blended reference still needs supporting information. Its construction and bond maturity must fit the changed strategy. If those details are missing, record questions instead of inventing a performance result. Preserving the old reference and adding the new one makes both the original promise and the revised task visible.
Save a benchmark-selection card
Choose a portfolio question, using the fictional task if you have no account. Record the question, eligible investments, main risks, and date of selection. Choose one primary benchmark and explain the fit. Add one alternative reference and explain the different question it answers. The purpose is to make the comparison reproducible, rather than collect as many reference lines as possible.
Include the benchmark's full name and version, source and retrieval date, comparison period, currency, dividend treatment, fees, cash-flow treatment, and triggers for revision. Then ask: Did I select this ruler because I had already seen the return? Could someone reconstruct the same comparison from this card? If either answer is unclear, label the result not yet comparable and identify the missing detail.
Keep the question and the ruler together
When you see “outperformed,” ask what was beaten and why that reference suited the original task. Align periods, currency, dividends, and fees, while keeping rule-based indexes, fund NAV returns, and actual investor results distinct. Missing information can make a comparison provisional even when its subtraction is correct.
Look beyond the number of funds to weights and overlap. When a strategy changes materially, preserve the old reference and document the new one from the change date. Your evaluation can develop as the question develops, while the historical record remains intact.
References
- S&P Dow Jones Indices: Methodology Matters and Index Mathematics Methodology. Rules, weighting, and return types; calculations here are original.
- SEC: Updated Investor Bulletin: Exchange-Traded Funds, February 23, 2023. Its scope is the U.S. registered funds described there, not every exchange-traded product.
- State Street: SPY official product information.
- State Street: XLK official product information. Product scope and document fields were checked October 30, 2026; obtain current holdings, names, and fees when using them.
AI assisted with writing and checks. Human financial review and target-reader testing remain pending.