Does borrowing more make the bond index buy more?

An equity index usually gives more weight to companies with greater stock-market value. A market-value-weighted bond index follows a different quantity: the market value of eligible debt still outstanding. At the same price, an issuer that sells more bonds can therefore occupy more of the benchmark.

This formula supports a compelling criticism: the more a company borrows, the more index funds have to buy it. The claim contains some truth, but only after three corrections. The index does not admit all debt. It may cap each issuer. Finally, a bond ETF does not necessarily replicate every line to the last dollar.

The useful interpretation is neither a machine blindly rewarding leverage nor a perfectly neutral benchmark. A bond index turns a filtered stock of debt into a reference portfolio. Its filters reveal where the bias exists, where it stops and how it reaches the market in practice.

The equation beneath the benchmark

The Bloomberg Fixed Income Index Methodology, updated on 8 January 2026, defines a bond’s market value as:

market value = (price + accrued interest) × current par amount outstanding

The security’s weight is then its market value divided by the total market value of all eligible bonds. The rule is not trying to identify the best company. It measures the size of available paper within a selected universe, with the aim of building an index that investors can replicate.

The scale matters. The Federal Reserve’s Financial Accounts recorded $7.510893 trillion of bonds on the liabilities side of US nonfinancial corporate business in the first quarter of 2026, up from $7.393445 trillion in the preceding quarter. These figures cover the full nonfinancial corporate sector in the Z.1 table, not only the debt eligible for an investment-grade index. (Federal Reserve via FRED)

A small imaginary index shows the raw mechanism. Assume three issuers, all bonds priced at par and no cap. A has 20 units of debt, B has 10 and C has 5. A therefore weighs 57.1% of the total. If A issues another 5 eligible units, its weight rises to 62.5% without any change in profitability, assets or governance.

The dilution is mechanical, but it is not a free reward. A fall in A’s bond prices reduces their market value. A buyback or maturity reduces the amount outstanding. Most importantly, the issue must pass through the index gates.

LQD: six gates before weighting

LQD makes the rules concrete. The iShares iBoxx Investment Grade Corporate Bond ETF tracks the iBoxx USD Liquid Investment Grade Index. The July 2026 S&P Dow Jones Indices methodology says the index is market-value weighted, rebalanced at each month-end and capped at 3% per issuer.

A candidate bond must, among other criteria:

• be US dollar-denominated corporate credit;
• have cash flows that can be determined in advance, with several security types excluded;
• carry an average investment-grade rating, at least BBB- or Baa3 depending on the agency;
• have at least $750 million of face value outstanding;
• come from an issuer with at least $2 billion of qualifying US dollar debt in the broader universe;
• have at least three years of expected remaining life, or three and a half years for a new insertion.

Rating and outstanding-amount data are cut off three business days before the rebalance. Weights and the cap are then calculated using month-end market values. A new issue that becomes known too late waits for the next cycle. (S&P DJI, Bond Selection and Index Calculation)

The slogan that the most indebted borrower always wins breaks down here. A group can increase its debt sharply without entering LQD if its rating becomes speculative. Conversely, a strong company issuing a large, liquid bond can cross the thresholds and increase its weight up to the cap. The benchmark rewards less the leverage itself than an abundant supply of standardised, tradable debt.

As of 7 August 2026, BlackRock reported 3,151 holdings in LQD’s portfolio. That figure describes the fund, not the index exactly, and the positions can change. It still helps explain why a manager cannot treat a bond ETF like an equity basket containing only a few hundred securities. (iShares / BlackRock)

From index weight to actual buying

An index buys nothing. The iShares prospectus filed with the SEC makes the distinction explicit: an index is a financial calculation, while the fund is an actual portfolio. LQD seeks to track its benchmark, but BlackRock uses representative sampling. The fund holds securities whose aggregate profile resembles the index without owning every bond in precisely the same proportion.

Another SEC filing reproducing LQD’s prospectus language says the fund invests at least 80% of its assets in index components and at least 90% in fixed-income securities of the types included in the index. The remaining flexibility can support tracking through other bonds or derivatives.

Actual transmission therefore depends on four conditions:

1. the issue becomes eligible at the next rebalance;
2. its raw weight rises without the cap fully neutralising the increase;
3. funds and mandates linked to the benchmark have assets to adjust;
4. managers select that bond, or a close substitute, to reduce tracking error.

Net inflows provide new capital. Without inflows, a rebalance can still occur, but buying one security will often require reducing other exposures. Outflows may overwhelm the weighting effect altogether. “Passive buying” therefore describes a relative portfolio constraint, not unlimited demand.

Bond ETF plumbing adds another layer of distance. Using 2020 data, the BIS found that creation baskets for US corporate bond ETFs covered about 20% of holdings, while redemption baskets covered about 35%. This flexibility helps funds manage thousands of bonds with uneven liquidity. It also means that creating ETF shares cannot be equated with buying every benchmark bond in proportion.

Six events with six different effects

A useful reading separates amount outstanding, price, eligibility and fund flows.

The rating boundary requires a careful distinction. The iBoxx methodology permits a bond downgraded below investment grade to leave even during its minimum-run period. This does not mean every investor must sell on the same day. In its May 2020 Financial Stability Report, the Federal Reserve noted that no general regulation required investment-grade funds to sell every fallen angel, although some funds might reduce their positions. Our guide to reading credit ratings explains this boundary.

Every benchmark contains an opinion

Market-value weighting has a strong justification: it reflects market size and capacity. An index based on available amounts is easier to follow than a portfolio that overweights small, scarce bonds. Bloomberg identifies this replicability as a central reason for the method.

The same rule nevertheless concentrates the portfolio on issuers supplying a great deal of eligible paper. It cannot tell whether the borrowing finances a destructive acquisition, a share buyback, a profitable data centre or the prudent refinancing of a maturity. Fundamental analysis must return to the balance sheet, maturity schedule and use of cash, as described in l0g’s guide to reading corporate debt.

The equity comparison clarifies the paradox. Capitalisation-weighted equity exposure rises when the value of equity rises. Market-value-weighted bond exposure can rise when the volume of eligible debt rises. In both cases the benchmark measures the existing market. It does not promise that the measured quantity is an economic virtue.

Other constructions exist. Bloomberg lists capped, fundamental, target-allocation and risk-weighted indices. Each replaces one weakness with a different choice: a cap reduces concentration but departs from the market; fundamental weighting requires a selection of indicators; risk weighting depends on a model and unstable data. Perfect neutrality is not available.

Federal Reserve research on the shift from active to passive investing also finds mixed effects. Some passive strategies may amplify volatility, while evidence connecting index investing to comovement in prices and liquidity remains inconclusive. The methodology establishes a constraint. On its own, it does not prove a price effect.

The l0g view

Yes, issuing more eligible debt can increase a borrower’s weight in a market-value-weighted bond index. That is an arithmetic consequence of the amount outstanding.

No, the index does not finance the most indebted company without limit. The price can fall, the rating can exclude, thresholds can block and a cap can bound the exposure. For the iBoxx index tracked by LQD, the 3% issuer limit is explicit.

Finally, a higher weight is not a universal market order. Funds use investment rules, flexible baskets and representative sampling. Establishing that a specific issue created actual buying pressure requires the inclusion calendar, weight changes, assets linked to the benchmark, fund holdings and transaction evidence. The equation supplies a channel. It does not prove causality on its own.

The most useful conclusion fits in one sentence: a bond index does not select the best borrowers; it organises the debt that has passed its filters.

---

Sources

• S&P Dow Jones Indices, iBoxx USD Liquid Investment Grade Index Methodology, July 2026: selection, thresholds, calendar, weighting and 3% issuer cap.
• Bloomberg Fixed Income Index Methodology, 8 January 2026: market-value formula and alternative weighting schemes.
• Federal Reserve, Financial Accounts Z.1, nonfinancial corporate business, first quarter 2026: corporate bonds on the liabilities side.
• SEC, iShares prospectus, 2025: distinction between index and fund, representative sampling and index-related risks.
• SEC, description of LQD taken from its prospectus, 2026 filing: investment objective and minimum asset proportions.
• iShares / BlackRock, LQD, data as of 7 August 2026: benchmark and number of fund holdings.
• BIS, “The anatomy of bond ETF arbitrage”, March 2021: divergence between holdings and creation or redemption baskets.
• Federal Reserve, “The Shift from Active to Passive Investing”, revised 2020: documented effects and uncertainties of passive indexing.
• Federal Reserve, Financial Stability Report, May 2020: potential pressure around fallen angels and the absence of a general mandatory-sale rule.

Method and limitations

• Data cut-off: 11 August 2026. The article uses no current security price, yield, valuation multiple or market ratio.
• The A, B, C example is entirely imaginary. It shows raw weighting before caps, taxes, transaction costs and price moves.
• The Z.1 figure covers US nonfinancial corporate business. It measures neither global bonds, investment grade alone nor assets actually held by LQD.
• BlackRock’s holdings count describes the fund’s portfolio as of 7 August 2026. It should not be read as the exact number of index constituents.
• The BIS study uses 2020 data and documents the general architecture of bond ETFs. Its basket-to-holdings overlaps are not presented as current LQD measurements.
• This article describes an index-construction mechanism and possible transmission channels. It does not estimate the causal effect of a particular bond issue on its price or funding cost and is not investment advice.