> For the complete documentation index, see [llms.txt](https://laurence-wilse-samson.gitbook.io/textbooks/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://laurence-wilse-samson.gitbook.io/textbooks/financial-economics-claims-prices-holders/part-i-foundations/chapter_02_map_of_claims.md).

# Chapter 2: A Map of Claims and Holders

*Part I: Foundations — Financial Economics: Claims, Prices, and Holders*

***

## Opening Episode: One Dollar of Retirement Saving

A hospital in Ohio runs payroll on a Friday. A nurse eight years into her career contributes six percent of her gross pay to the employer's 401(k) plan, and the payroll system withholds $180.

The first thing that happens is that a bank deposit moves. The hospital's operating account at a regional bank is debited and the plan's trust account at a large trust bank is credited. Nothing has been invested yet; a deposit has changed hands, and a deposit is a claim on a bank — the bank's promise to pay currency on demand. The regional bank settles with the trust bank by transferring reserves at the Federal Reserve, which are claims on the central bank. Two links in, and the dollar has already been two different people's liability.

The plan's recordkeeper credits the nurse's account with $180 and applies her investment election, which she has never changed, because she never made one. Under the plan's default she is in a target-date 2055 fund. Her $180 buys shares in that fund. Those shares are her claim on a pro-rata interest in whatever the pool owns.

The target-date fund owns almost nothing directly. It is a fund of funds, and its assets are shares in four other funds: a total US stock index fund, an international stock index fund, a US aggregate bond index fund, and a short-duration bond fund. At her age the glide path puts roughly ninety percent of the money in equity. So her $180 becomes about $110 of total-US-stock-fund shares, $50 of international-stock-fund shares, and $20 of bond-fund shares. Three more claims, each on a pool.

The total US stock index fund holds equity in something over three thousand companies. Her $110 buys something like seven dollars of the largest company in the market, three or four dollars of the second, and pennies or fractions of pennies of the rest. But the fund does not appear on any company's shareholder register. The shares sit in the fund's account at its custodian bank, the custodian's position sits in the books of the Depository Trust Company, and DTC holds the registered position through its nominee. The company's register records one holder for most of its outstanding stock: a nominee name, holding on behalf of participants, holding on behalf of custodians, holding on behalf of funds, holding on behalf of plans, holding on behalf of a nurse in Ohio.

The bond leg is shorter and ends somewhere else. The aggregate bond fund's largest position is US Treasury securities. Those are book-entry claims recorded in the Federal Reserve's securities system, held in the custodian's account there. Her twenty dollars of bond fund is, at the end, a fractional claim on the taxing power of the United States, plus fractional claims on the mortgage payments of American households through agency mortgage-backed securities, plus fractional claims on the cash flows of investment-grade corporations.

Count the layers. Deposit, trust account, participant account, target-date fund share, underlying fund share, custodial position, depository position, registered share — and only then the operating firm whose factories, patents, and payrolls are what all of this is ultimately a claim on. Eight or nine promises deep, and not one of them is redundant. Each layer exists because somebody was willing to pay for something the layer provides: recordkeeping, diversification, daily liquidity, legal segregation, settlement finality, custody.

![Figure 2.1: One dollar of retirement saving](https://846781005-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3EupdX99vVBoNySDtmxb%2Fuploads%2Fgit-blob-1321ba9c339eb81861301394fd6f6e02a403824c%2Ffig_02_01_one_dollar_of_retirement_saving.png?alt=media)

**Figure 2.1: One dollar of retirement saving.** The episode's chain, layer by layer, with what each layer is being paid for set in italic inside its box and what the claim actually is beside it. Read down the middle column and the chapter's first fact is unavoidable: at every step the dollar is somebody's asset and somebody else's liability, and at no step has any wealth been created. Read the italic lines and a second fact appears — none of the layers is an accident or a rent. Each is a service that somebody chose to buy, which is why the chain has not been competed away in the fifty years it took to assemble. The equity leg is drawn; the bond leg is shorter and lands somewhere else entirely, in the taxing power of the United States and the mortgage payments of American households. Hold the count in mind. Every later part of this book is about one or another of these layers, and Part IV is about what happens to prices when the layers with the most money in them are the ones that do not choose.

Two observations follow, and they organize the rest of the chapter.

First, at every intermediate step, the thing held is a claim on somebody. The nurse's asset is the plan's liability. The plan's asset is the fund's liability. The fund's asset is the company's equity — which is not, strictly, a liability, but is a residual claim that the company's own balance sheet must account for. Financial assets are not free-floating objects. They come in pairs, and the whole system is a directed graph of promises whose only unmatched endpoints are real things: buildings, machines, software, and the future output of workers.

Second, the aggregate of this graph is measurable, and somebody measures it. The Federal Reserve publishes it four times a year in a statistical release called the Financial Accounts of the United States, known by its old code, Z.1. It reports, for every major sector of the economy, what that sector owns and what it owes, instrument by instrument, and it reports the same claim from both sides — so that the amount of Treasury securities held, summed across all holders, equals the amount issued. That double-entry structure is the map. The rest of this chapter reads it.

***

## 2.1 Balance Sheets: The Minimum You Need

The measurement system deserves a chapter of its own, and gets one in the companion volume. What follows is the part a reader needs to proceed.

### The Reader's Own Balance Sheet

A balance sheet is a list of what an agent owns and what it owes, at an instant. Assets on the left, liabilities on the right, and the difference — net worth, or equity — on the right as well, so that the two sides sum to the same number by construction.

Start with the reader's own.

**Table 2.1: A Household Balance Sheet**

| Assets                        |             | Liabilities and Net Worth |             |
| ----------------------------- | ----------- | ------------------------- | ----------- |
| Checking and savings deposits | $9,000      | Credit card balance       | $3,000      |
| Retirement account (401k)     | $46,000     | Student loans             | $28,000     |
| Car                           | $12,000     | Auto loan                 | $7,000      |
| **Total assets**              | **$67,000** | **Total liabilities**     | **$38,000** |
|                               |             | **Net worth**             | **$29,000** |

*Source: Illustrative; author's construction*

Three things about this table generalize to every balance sheet in the chapter.

**Assets split into financial and nonfinancial.** The car is a real asset. It is not anyone's liability; it is a thing. The deposits and the retirement account are financial assets, and each is somebody else's obligation. The distinction matters because only the financial ones have a counterparty.

**Net worth is a residual, not a plan.** It is defined as assets minus liabilities and it moves whenever either side is revalued. If the retirement account rises with the stock market, net worth rises by the same amount, with no saving, no transaction, and no decision.

**A balance sheet is a stock, not a flow.** It is a photograph at a date. The nurse's $180 was a *flow* — saving, per pay period. The accumulated $46,000 is a *stock*. The Financial Accounts publish both, in matched form: each quarter's flow table explains the change in the corresponding stock table, up to revaluation. Confusing the two is the most common error in reading financial data, and it is the reason a headline about "record fund inflows" and a headline about "record fund assets" can be about entirely different things.

### One Agent's Asset Is Another's Liability

Put two balance sheets side by side and the deposit appears twice. The household's $9,000 asset is the bank's $9,000 liability. Do it for every agent in the economy and every financial claim appears exactly twice — once as an asset, once as a liability — while every real asset appears once.

The consequence is an identity. Summed across all sectors:

**Aggregate financial assets = aggregate financial liabilities**, and therefore **aggregate net worth = aggregate nonfinancial (real) assets**, plus the economy's net claims on the rest of the world.

The nation's wealth is its buildings, machines, land, and intangible capital. Everything else nets to zero. This is not a diminishment of finance; it is the statement of what finance does. It does not create aggregate wealth by issuing claims. It allocates the ownership of, and the risk attached to, wealth that already exists — and, by doing so, changes how much of it gets created.

### Netting and Gross

Which raises the question the identity invites: if it all nets to zero, why is the graph so large?

Consider the nurse again. Her ultimate exposure — the net position — is a share of the productive capital of several thousand companies and a share of the government's tax base. That exposure could in principle be delivered by a single claim. Instead it is delivered by nine. Each layer adds an asset for one party and a liability for another, and each therefore adds to the *gross* size of the financial system without changing anyone's *net* exposure by a dollar.

Gross positions matter for three reasons that recur throughout the book.

**Counterparty risk lives on the gross position.** The nurse's net exposure is unaffected by whether her custodian fails; her ability to get at it is not.

**Balance-sheet capacity is consumed by gross positions.** A dealer that is long a bond and short a nearly identical bond has almost no net exposure and a very large balance sheet, and regulatory leverage constraints bind on the latter. Chapter 19 makes this the center of an asset-pricing theory.

**Netting hides who bears risk.** Chapter 2's own master table will show households holding a modest and shrinking direct position in corporate equity. That is not a statement about who bears equity risk. It is a statement about the layer at which the position is recorded. Reading a netted number as an economic fact is the error §2.5 is built to prevent.

***

> **Box 2.1 — Copeland and the Invention of the Map**
>
> Before 1952 there was no map. National income accounting, built in the 1930s and 1940s, measured production and income — what was made and who earned it. It said nothing about who financed it, who held the resulting claims, or how funds moved between sectors.
>
> Morris Copeland, working at the National Bureau of Economic Research, argued that this was a gap with consequences. In *A Study of Moneyflows in the United States* (1952) he proposed a system that tracked, for each sector of the economy, not only its income and outlays but its sources and uses of funds — where the money came from and what claims it left behind. His question was explicitly a monetary one: he wanted to know how a change in the money supply reached spending, and he found that he could not answer it without a full accounting of intersectoral claims.
>
> The Federal Reserve adopted the framework, and the flow of funds accounts began publication in the 1950s. They were renamed the Financial Accounts of the United States in 2013, and the release retains its original code, Z.1. Copeland's design decision — measure both sides of every claim, sector by sector, and force them to reconcile — is why a reader today can ask "who holds Treasuries?" and get an answer that adds up.
>
> The intellectual descendants are everywhere in this book. The balance-sheet approach to monetary economics, the sectoral-balances view of macroeconomic imbalances, and the modern literature on intermediary and demand-based asset pricing all take Copeland's accounting as their substrate and add prices to it.

***

## 2.2 What Is a Claim? The Instrument Taxonomy

A **claim** is a promise, held by one party against another, that is enforceable in a specific way and ranks in a specific place. Three attributes define it, and every instrument in the chapter is a configuration of them.

**What is promised.** A fixed schedule of payments, or a residual share of whatever is left, or a payoff contingent on the value of something else.

**How the promise is enforced.** Contract law, backed by a bankruptcy process that can seize assets and replace management, is the enforcement mechanism for debt. Equity has no such mechanism; the shareholder's remedies are governance rights, not collection rights. Sovereign debt sits awkwardly: there is a contract, but no court can seize a country, so enforcement rests on reputation and market access. Chapter 10 treats willingness to pay as distinct from ability to pay for exactly this reason.

**Where it ranks.** Seniority determines who is paid in what order when there is not enough. Secured debt, then senior unsecured, then subordinated, then preferred, then common equity. The ranking is the whole content of the phrase "capital structure," and it is what Part V is about.

### Debt

A promise to pay stated amounts on stated dates. The holder's upside is capped at the promised payments; the downside runs from full repayment to whatever recovery the bankruptcy process delivers. Debt is classified two ways at once, and both classifications appear in the master table.

**By issuer**, because the identity of the promisor is most of the risk:

* **Sovereign** — Treasury bills, notes, bonds, and TIPS. The issuer taxes and, for domestic-currency debt, issues the currency of denomination.
* **Agency and government-sponsored** — debt of Fannie Mae, Freddie Mac, the Federal Home Loan Banks, and Ginnie Mae, together with the mortgage-backed securities they guarantee. Legally distinct from Treasury debt; priced as though it were nearly so.
* **Municipal** — state and local government debt, mostly tax-exempt to US holders, which shapes who holds it in a way Chapter 10 exploits.
* **Corporate** — investment grade and high yield, plus the syndicated loan market that competes with it.
* **Household-collateralized** — mortgages, auto loans, credit card receivables, student loans; issued by households, and reaching investors either on a bank's balance sheet or repackaged into asset-backed securities.
* **Financial** — the liabilities intermediaries issue to fund themselves: deposits, repurchase agreements, commercial paper, bank bonds. This category is where the map's layering shows up as an instrument class.

**By maturity and liquidity**, because a claim's usefulness as a store of value depends on how quickly it can be turned into settlement money without loss. Overnight repo, three-month bills, ten-year notes, and thirty-year bonds are promises by the same issuer with entirely different risk and entirely different holders. Chapter 9 turns this into the term structure; Chapter 16 turns it into a hierarchy of moneyness.

### Equity

A residual claim: whatever is left after the debt is paid, with no promise attached and no maturity date. Limited liability truncates the loss at zero. The holder's rights are to vote, to receive dividends if declared, and to sue for breach of fiduciary duty — not to demand payment.

The category includes public common stock, preferred stock (a hybrid in economic substance), private equity in operating firms, and the equity of financial firms, which matters disproportionately because it is the buffer that determines how much of the claim graph an intermediary can support.

### Derivatives

A claim whose payoff is defined by reference to something else: an index, a rate, a price, a credit event. Forwards, futures, swaps, and options. Two features distinguish them from the rest of the taxonomy.

They are usually in **zero net supply**. Every futures contract has a long and a short, so the instrument nets to zero across holders and appears in the Financial Accounts only in fragments. This is why the master table has no derivatives row, and why measuring the derivatives system requires other sources — a limitation the data appendix returns to.

They **unbundle risk from funding**. Buying a bond requires cash and produces exposure. Selling protection on that bond via a credit default swap produces nearly the same exposure with almost no cash. Chapter 8 prices these instruments; Chapter 26 asks what happens to the map when the risk and the funding sit in different places.

### Hybrids

Instruments deliberately built to sit between the categories, usually to arbitrage a tax, accounting, or regulatory boundary. Convertible bonds (debt with an equity option attached), preferred stock (equity in law, fixed-income in behavior), contingent convertible bank capital (debt that becomes equity when a regulatory trigger fires), and the tranches of a securitization, which slice a single pool of household debt into claims of different seniority. The existence and profusion of hybrids is evidence for a proposition Part V argues directly: the boundary between debt and equity is a legal and tax construct laid over a continuum.

**Table 2.2: The Instrument Taxonomy**

| Class                         | What is promised                                                                             | Enforcement                                                                 | Typical seniority                            | Principal US issuers                          |
| ----------------------------- | -------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------- | -------------------------------------------- | --------------------------------------------- |
| Sovereign debt                | Fixed schedule, nominal or inflation-indexed                                                 | No court remedy; reputation and market access                               | Senior to all domestic claims in practice    | US Treasury                                   |
| Agency / GSE debt and MBS     | Fixed schedule; MBS pass through mortgage cash flows with prepayment option held by borrower | Contract; implicit or explicit federal guarantee                            | Senior secured or guaranteed                 | Fannie Mae, Freddie Mac, Ginnie Mae, FHLBs    |
| Municipal debt                | Fixed schedule, usually tax-exempt to US holders                                             | Contract; limited bankruptcy (Bankruptcy Code Chapter 9) or none for states | Varies: general obligation vs. revenue       | States, cities, authorities, school districts |
| Corporate debt                | Fixed schedule, with covenants                                                               | Contract; Bankruptcy Code Chapter 11 reorganization                         | Secured → senior unsecured → subordinated    | Nonfinancial and financial corporations       |
| Household-collateralized debt | Fixed schedule, secured on a specific asset                                                  | Contract; foreclosure, repossession, Bankruptcy Code Chapter 7 or 13        | Secured on the collateral                    | Households (via banks and securitization)     |
| Financial-sector debt         | Fixed schedule; often payable on demand                                                      | Contract; deposit insurance; resolution regimes                             | Deposits senior to bank bonds to bank equity | Banks, dealers, finance companies             |
| Equity                        | Residual only; no promise                                                                    | Governance rights; fiduciary duty                                           | Last                                         | Corporations, public and private              |
| Derivatives                   | Contingent payoff referencing another quantity                                               | Contract; margin and central clearing                                       | Collateralized, often senior via netting     | Dealers, exchanges, end users                 |
| Hybrids                       | Configurable: debt that converts, equity that pays fixed                                     | Contract, with the trigger written in                                       | By construction, between debt and equity     | Corporations, banks, securitization vehicles  |

*Source: Author's construction; instrument definitions follow the Financial Accounts of the United States (Z.1), Federal Reserve*

***

## 2.3 The Master Map

This section is what later chapters cite. It has two parts: what each major sector's balance sheet looks like, and who holds each major class of claim.

Two warnings about the numbers, which apply to every table in the rest of the chapter. They are round, they are approximate, and they are stated to the nearest trillion or nearest five percentage points on purpose. A student who takes them as exact is misusing them; a student who takes them as roughly right is using them correctly. The magnitudes are what matter — that Treasuries outstanding are tens of trillions and municipal bonds are a few, that foreign holders own a fifth of the US equity market and not five percent or fifty. The data exercise at the end of the chapter has the reader rebuild the exact figures from the source, which is the only defensible way to have exact figures.

The second warning concerns the household sector. In the Financial Accounts, "households and nonprofit organizations" is a **residual**: it is computed as what is left after every other sector's measured holdings are subtracted from the known total outstanding. Hedge funds, personal trusts, and domestic nonprofits fall into it. When the household row shows a large direct holding of an instrument no ordinary household owns, the residual is doing the work.

### Five Sector Balance Sheets

**Table 2.3: Condensed Sector Balance Sheets, United States, mid-2020s ($ trillions)**

| Sector                          | Financial assets | Nonfinancial assets | Total assets | Liabilities | Net worth |
| ------------------------------- | ---------------- | ------------------- | ------------ | ----------- | --------- |
| Households and nonprofits       | \~142            | \~63                | \~205        | \~22        | \~183     |
| Nonfinancial corporate business | \~37             | \~33                | \~70         | \~32        | \~38      |
| Federal government              | \~6              | \~5                 | \~11         | \~37        | \~−26     |
| Private depository institutions | \~29.5           | \~1                 | \~30.5       | \~27.5      | \~3.0     |
| Funds, insurers, and pensions   | \~89             | \~0                 | \~89         | \~74        | \~15      |

*Source: Financial Accounts of the United States (Z.1), 2026 Q1, Federal Reserve, via the FRED mirror; magnitudes approximate. The funds row's net worth is a residual struck across three different measurement conventions — pension entitlements at actuarial value, mutual fund shares at net asset value, and ETF shares carried inside corporate equities rather than in the fund-share liability row — and is an artifact of that accounting rather than an equity cushion.*

Read the rows.

**Households** are the sector that owns everything at the end of the chain. Their financial assets are roughly twice their real assets, and their net worth is on the order of $180 trillion, or something like six times annual GDP — a ratio that has risen substantially since the 1980s, mostly through asset revaluation rather than saving. Their liabilities are dominated by mortgages, with student and consumer credit behind.

**Nonfinancial corporations** are the sector the whole graph ultimately points at. Note the gap between the balance-sheet net worth in the table (\~$38 trillion, at estimated replacement or market value of assets) and the market value of the same firms' equity, which is larger. That ratio is Tobin's q; Chapter 22 makes it an object of theory.

**The federal government** has negative net worth by tens of trillions and this is not, by itself, informative, because the accounts do not capitalize the government's principal asset — the present value of future tax receipts. What the row does say is that Treasury securities are the largest single liability class in the map, and therefore that the question "who holds Treasuries?" is one of the largest allocation questions in the system.

**Depository institutions** have a balance sheet of about $30 trillion supported by roughly $3 trillion of equity: leverage of roughly ten to one, an order of magnitude more than nonfinancial firms carry. That asymmetry is the subject of Chapter 19.

![Figure 2.3: Five sector balance sheets](https://846781005-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3EupdX99vVBoNySDtmxb%2Fuploads%2Fgit-blob-37505cde968c69ca1523422d53e244b396e6248a%2Ffig_02_03_five_sector_balance_sheets.png?alt=media)

**Figure 2.3: Five sector balance sheets.** Table 2.3's rows, drawn from the current Z.1 vintage. Panel (a) is size on one scale, each bar split into the financial assets the rest of this book is about and the nonfinancial assets — houses, plant, equipment, structures — that the financial claims are ultimately claims on. Two features of the panel are the chapter's. The household bar is more than twice the next largest, and roughly two thirds of it is financial, which is the sense in which households are the sector that owns everything at the end of the chain. And the two intermediary sectors hold almost no real assets at all: a bank's building is a rounding error against its loan book. Panel (b) is the leverage sentence above, for the three sectors it compares. The depository sector carries about ten dollars of assets for every dollar of net worth against a household's 1.1 and a nonfinancial corporation's 1.8 — an order of magnitude, on the nose, and the reason Chapter 19 exists. The federal government and the funds-insurers-pensions composite are in panel (a) and not in panel (b); the panel's note says why. The levels are the 2026 Q1 vintage that Table 2.3 now carries. *Source: Financial Accounts of the United States (Z.1), sector balance sheets through the FRED mirror. Nonfinancial assets and net worth are taken as differences, so each bar balances by construction. Author's calculations.*

**The fund, insurance, and pension complex** owes nearly everything it owns to its own claimants, and that is the definition of the sector rather than a weakness of it. These institutions are pass-through structures: their liabilities are claims held by households and their assets are claims on issuers, and the sector's function is to sit in the middle and transform one into the other. Table 2.4 opens it up.

**Table 2.4: The Fund, Insurance, and Pension Complex ($ trillions)**

| Institution type                            | Approximate assets | What it owes, and to whom                                     |
| ------------------------------------------- | ------------------ | ------------------------------------------------------------- |
| Mutual funds (long-term)                    | \~20               | Redeemable shares, mostly to households and retirement plans  |
| Exchange-traded funds                       | \~10               | Shares tradeable on exchange; created and redeemed in kind    |
| Money market funds                          | \~7                | Stable-value shares to households, corporations, institutions |
| Life insurers (general + separate accounts) | \~9                | Policy reserves, annuities, guaranteed products               |
| Property-casualty insurers                  | \~3                | Claims reserves, largely short-tailed                         |
| Private pension funds (DB and DC)           | \~15               | Retirement entitlements to participants                       |
| State and local government pensions         | \~6                | Retirement entitlements, backed by sponsor covenants          |
| Federal government retirement               | \~2                | Entitlements to federal employees and military                |

*Source: Financial Accounts of the United States (Z.1), Federal Reserve; magnitudes approximate*

The nurse's $180 passed through three rows of this table on its way from a payroll system to a factory.

### The Master Holdings Table

This is the chapter's core object; Figure 2.2 draws it as a matrix shaded by dollar size.

**Table 2.5: Who Holds What — Major Claim Classes by Holder Sector, United States, mid-2020s ($ trillions)**

| Claim class                                       | Total outstanding | Households & nonprofits | Nonfin. business | Govt. incl. Fed | Banks & depositories | Funds (MF, ETF, MMF) | Insurers & pensions | Rest of world |
| ------------------------------------------------- | ----------------- | ----------------------- | ---------------- | --------------- | -------------------- | -------------------- | ------------------- | ------------- |
| Treasury securities                               | \~27              | \~4                     | —                | \~5.5           | \~2                  | \~5.5                | \~2.5               | \~8.5         |
| Agency & GSE-backed securities (incl. agency MBS) | \~12              | \~1                     | —                | \~2.5           | \~3                  | \~2                  | \~1.5               | \~1.5         |
| Corporate & foreign bonds                         | \~16              | \~2                     | —                | —               | \~1                  | \~4                  | \~5                 | \~4           |
| Municipal securities                              | \~4               | \~1.5                   | —                | —               | \~0.4                | \~1                  | \~0.5               | —             |
| Corporate equities                                | \~60              | \~25                    | \~0.5            | \~0.5           | —                    | \~15                 | \~4                 | \~12          |
| Mutual fund and ETF shares                        | \~30              | \~21                    | \~1              | \~1             | —                    | —                    | \~6                 | \~1           |
| Deposits at US depositories                       | \~19              | \~12                    | \~3.5            | \~0.7           | —                    | \~1.5                | \~0.5               | \~1.5         |
| Pension entitlements                              | \~33              | \~33                    | —                | —               | —                    | —                    | —                   | —             |

*Source: Financial Accounts of the United States (Z.1), Federal Reserve; magnitudes approximate. Em-dash denotes a position too small to register at this rounding. Rows do not sum exactly to totals: minor holder sectors are omitted and every cell is rounded.*

![Figure 2.2: The master map: claim classes by holder sector](https://846781005-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3EupdX99vVBoNySDtmxb%2Fuploads%2Fgit-blob-22dfc81eb97c31011c2812ea9585fec5c719495c%2Ffig_02_02_master_map.png?alt=media)

**Figure 2.2: The master map: claim classes by holder sector.** Table 2.5 as a claim-class x holder-sector matrix, cells shaded by dollar size on a log scale, with the rest-of-world column outlined as the large one the text flags. Each cell carries the amount held in trillions of dollars with the share of its row below it; the colour key runs along the foot of the figure. Households and nonprofit organizations is a Z.1 residual — hedge funds, personal trusts and domestic nonprofits sit inside it — so that column overstates ordinary households. Holder columns aggregate published Z.1 sectors: federal and state and local employee retirement funds are in insurers and pensions, not in government and central bank, and Z.1 classifies ETF shares within corporate equities, so the mutual fund share row is L.224 alone. An em dash is a position below $0.05 trillion or one Z.1 does not publish; rows do not sum to their totals because minor holder sectors are omitted; and the columns must not be added down, since pension entitlements are a claim on assets that appear in the rows above. *Source: Financial Accounts of the United States (Z.1), 2026 Q1, Board of Governors, via the FRED mirror of Z.1 release 52; one series per cell, from rows L.210, L.211, L.213, L.212, L.223, L.224, L.204-L.205 and L.117-L.120; author's calculations.*

Three features of the table are worth naming before it is put to work.

**The last row is a claim with exactly one holder.** Pension entitlements are, by construction, held entirely by households: a pension entitlement is the household's claim on a pension fund, and it exists nowhere else. It is also the second-largest single row in the table. This is the sharpest illustration in the chapter of the layering point: the $33 trillion of entitlements is *not* an additional $33 trillion of wealth. It is a claim on the fund's assets, which are themselves rows further up the same table. The map counts both, correctly, and a reader who adds the column has double-counted.

**Instruments differ enormously in how concentrated their ownership is.** Municipal bonds are held almost entirely by US households, US mutual funds, and US insurers, for a single reason: the tax exemption is worthless to a foreign holder, a pension fund, or an endowment, so those holders do not appear. Treasury securities, at the other extreme, are held by everyone in the table and by foreign central banks besides. Chapter 10 turns the muni clientele into a spread puzzle; Chapter 9 turns the Treasury clientele into a theory of safe assets.

**The rest-of-world column is one of the largest in the table.** Foreign holders are the single largest holder class of Treasury securities and hold roughly a fifth of US corporate equity. No account of who funds the US government or who bears US equity risk that stops at the water's edge is complete. See Box 2.2.

***

## 2.4 Fifty Years of Reallocation

The map is not stable. Three reallocations dominate the postwar record, and every one of them is a change in *who holds* rather than in *what is issued*.

![Figure 2.4: From direct to intermediated ownership](https://846781005-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3EupdX99vVBoNySDtmxb%2Fuploads%2Fgit-blob-d08fbbc93d988c432898449a005e18eac0146d93%2Ffig_02_04_direct_to_intermediated.png?alt=media)

**Figure 2.4: From direct to intermediated ownership.** The first of the three reallocations, as shares of all US corporate equities outstanding. In 1945 households held ninety-five percent of the equity market in their own names; by 2012 they held thirty-five, and the difference went to funds, pensions and insurers first and then increasingly to the rest of the world. Three things this drawing makes clear that a pair of columns cannot. The rotation is not a story about households owning less equity — their claim on the corporate sector is larger than it has ever been — but about the number of intermediaries between them and it, which is what Chapter 2 is counting. It is not monotone: the direct share bottoms in 2012 and has risen since, which is index funds and brokerage accounts rather than a return of the stock-picking household. And the foreign band is the fastest-growing thing in the picture, which is why the "rest of the world" column of Table 2.5 is the one the later chapters keep having to look at. Shares are computed against all corporate equities, because the household line in the Financial Accounts is a residual defined on that base; the grey band is the sectors not named — banks, nonfinancial corporations, closed-end funds and government. *Source: Financial Accounts of the United States (Z.1), table L.223, corporate equities by holder, via the FRED mirror; author's calculations.*

**Table 2.6: Three Reallocations, c. 1975 vs. c. 2025 (approximate shares)**

| Measure                                                         | c. 1975 | c. 2025 |
| --------------------------------------------------------------- | ------- | ------- |
| Households' *direct* holdings, share of US corporate equity     | \~60%   | \~40%   |
| US funds, insurers, and pensions, share of US corporate equity  | \~20%   | \~35%   |
| Rest of world, share of US corporate equity                     | \~4%    | \~18%   |
| Defined benefit share of private pension assets                 | \~70%   | \~20%   |
| Mutual fund and ETF shares, share of household financial assets | \~2%    | \~18%   |
| Depository institutions, share of US financial-sector assets    | \~55%   | \~25%   |
| Bonds share of nonfinancial corporate credit-market debt        | \~45%   | \~65%   |
| Finance and insurance, share of GDP                             | \~4.5%  | \~8%    |

*Source: Financial Accounts of the United States (Z.1), Federal Reserve; GDP share from Greenwood and Scharfstein (2013) and BEA; magnitudes approximate*

### From Direct to Intermediated Ownership

In the 1950s, an American who owned stock owned it in certificate form, registered in her own name, chosen by her or her broker. Households directly held the overwhelming majority of US corporate equity. By the mid-1970s that share had fallen to about sixty percent; today it is around forty percent of a market many times larger, and the residual-sector caveat means even that figure overstates ordinary households, since hedge funds and personal trusts sit inside it.

The share did not migrate to a single new owner. It went partly to US institutions — mutual funds, ETFs, pension funds, insurers — and partly abroad. What changed is that the *decision* to hold a particular company was separated from the *bearing* of that company's risk. The nurse in the opening episode bears equity risk in three thousand firms and has chosen none of them. Chapters 12, 16, and 17 develop the consequence: when the holder of record is an agent operating under a mandate, the mandate becomes an input into prices.

### From Defined Benefit to Defined Contribution

In 1975 a private-sector worker with a pension typically had a *defined benefit* claim: a promise from the employer of a stated income in retirement, calculated on years of service and final salary. The employer bore the investment risk and the longevity risk; the pension fund's asset allocation was the employer's problem.

Today most private-sector retirement wealth sits in *defined contribution* plans and individual retirement accounts. The employer's promise is to contribute; everything after that — the asset allocation, the market risk, the sequence-of-returns risk, the risk of outliving the money — belongs to the worker. The DB system did not vanish. It survives in the public sector, where state and local plans still hold several trillion dollars against benefit promises, and in a shrinking legacy corporate stock now managed largely to be de-risked and transferred to insurers.

![Figure 2.5: Defined benefit to defined contribution](https://846781005-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3EupdX99vVBoNySDtmxb%2Fuploads%2Fgit-blob-33e34c0b9bb83ba172b66e1afbc05dc0a96a2f7c%2Ffig_02_05_defined_benefit_to_defined_contribution.png?alt=media)

**Figure 2.5: Defined benefit to defined contribution.** The second of Table 2.6's three reallocations, in levels and in shares, from the Financial Accounts alone. Panel (a) shows what grew: US retirement wealth rose from under three trillion dollars in 1985 to about thirty-eight, and every band in the stack rose with it — the defined benefit promise did not shrink in dollars. Panel (b) shows what changed. The dashed line separates the promises from the account balances, and the promise falls from 87 percent of the total in 1985 to 43 percent now, crossing below half in 2023. The two panels together are the point: nothing was taken away from anybody, and the composition of who bears the investment risk was nevertheless reversed inside forty years. One measurement note. The three defined benefit bands are *entitlements* — an actuarial liability of a fund to its members, computed at a discount rate the fund does not choose — while the two lower-risk-bearing bands are *assets* at market value. The stack is therefore not a single accounting object, and the reason it is still the right picture is that both measures answer the same question: how much retirement wealth is a claim on somebody else's promise, against how much is a balance worth whatever the market says today. Chapter 14 §14.3 takes the household consequences and Chapter 16 §16.2 takes the fund's. *Source: Financial Accounts of the United States (Z.1), pension entitlement and retirement account rows through the FRED mirror. Author's calculations.*

The pricing consequences are large and run in both directions. A DB plan with a defined liability has a natural demand for duration — long bonds that match long promises — and Chapter 9 shows how that demand shapes the long end of the yield curve. A DC saver in a target-date default has a demand that responds to birthdays and payroll dates. Substituting one for the other has changed the composition of demand for every long-dated claim in the map.

### From Banks to Markets and Funds

In 1975, an American corporation that needed money borrowed it from a bank, and the bank funded itself with deposits. The depository sector held roughly half of all US financial-sector assets. Today it holds around a quarter, and the money it does not hold sits in mutual funds, ETFs, money market funds, insurers, pension funds, and the securitization structures that convert loans into securities.

Corporate borrowing followed. Bonds are now roughly two-thirds of nonfinancial corporate credit-market debt, against something under half fifty years ago, and the loan market that remains is substantially originated to be distributed rather than held.

The most useful thing to notice about this shift is what it did *not* do. The bank did not disappear from the chain; it changed position in it. A mortgage that in 1975 sat on a savings institution's balance sheet, funded by insured deposits, today more often sits inside an agency MBS pool, held by a mutual fund, whose shares are held by a retirement plan — with a bank still present as originator, servicer, custodian, or repo counterparty. Chapter 19 asks whether this makes the system more robust or merely relocates the fragility, and Chapter 13 follows the mortgage in particular.

Underlying all three is a fourth fact: gross intermediation grew faster than net saving. The finance and insurance share of GDP roughly doubled over the period, while the household saving rate did not rise. More layers were added between the saver and the real asset. Greenwood and Scharfstein decompose the growth and find it concentrated in asset management and in the fee income associated with household credit — which is to say, in the business of running the layers rather than in the business of financing new capital. Whether the additional layers earn their cost is the question Chapter 27 closes on.

***

## 2.5 Who Issues, Who Holds, and Why It Matters

A table is only useful if the reader knows which direction to read it in. Table 2.5 rewards two readings.

### Reading Across: Who Funds This Issuer?

Take the Treasury row. Foreign holders are the largest single class, at roughly $8.5 trillion; the Federal Reserve and other government accounts hold something over $5 trillion; funds — money market funds especially, at the short end — hold a comparable amount; the household residual, which at this point contains a great deal of hedge-fund basis-trade inventory, holds several trillion; banks and insurers hold the rest. This is not a stable partition. Each of these holders buys Treasuries for a different reason and would sell for a different reason, and the composition has changed radically twice in twenty years — once through quantitative easing and once through its reversal. Chapter 9 makes this the book's cleanest demonstration that holder demand moves prices.

Take the corporate bond row. Insurers and pension funds together hold roughly $5 trillion, more than any other domestic sector, and their holdings are governed by risk-based capital rules and by the accounting treatment of held-to-maturity assets. When Chapter 16 says that insurers are the marginal holder of corporate credit, Table 2.5 is what says how big that position is — and therefore how much a change in insurers' regulatory treatment of a rating category can be expected to move spreads.

Take the equity row. Households directly hold roughly $25 trillion of a roughly $60 trillion market. Direct household equity ownership is also among the most concentrated asset holdings in the economy: the top decile of the wealth distribution holds the large majority of directly held stock, and the top one percent holds a substantial share of that. The aggregate row conceals this entirely — an aggregate cannot show a distribution. Chapter 14 develops the concentration point with the Federal Reserve's distributional accounts and argues that it is a pricing input rather than a distributional footnote: if the marginal holder of equity risk is a wealthy household with an unusual consumption process, the consumption-based models of Chapter 5 are being tested against the wrong consumer.

### Reading Down: What Does This Holder Own?

Take a defined benefit pension fund. Its column, disaggregated, shows long-dated bonds, corporate credit, public equity, and increasingly private assets — an allocation driven not by a view on relative value but by a liability whose duration is measured in decades and whose discount rate is set by accounting and funding rules. Chapter 16 argues that the liability side is doing the work: the fund's asset demand is a derived demand, and understanding it requires reading the claim it has issued, not just the claims it holds.

Take a money market fund. Its column is Treasury bills, repo, and short agency paper — a portfolio constructed to support a promise of stable value and same-day redemption. Chapter 16's hierarchy-of-money section explains why that promise pins the fund to the shortest and safest instruments in the map, and why an attempt to reach for a few extra basis points inside that mandate has twice been a systemic event. A stablecoin issuer's column is close to a copy of this one, made without the rulebook or the backstop; Box 2.3 puts it and the other digital claims on the map.

Take a bank. Loans, Treasury and agency securities, reserves at the Fed — funded by deposits that can leave on any business day. The mismatch is the business model and the vulnerability, and Chapter 19 prices it.

### The Standing Instruction

Every later chapter in this book opens from Table 2.5.

When Chapter 9 asks who the marginal buyer of duration is, the answer starts from the Treasury row. When Chapter 10 says insurers are the marginal corporate-bond holder, this is the table that says the position is roughly $5 trillion and therefore that the claim is worth taking seriously. When Chapter 12 discusses index inclusion as a demand shock, the size of the shock is a fraction of the mutual fund and ETF entry in the equity row. When Chapter 17 argues that the passive complex has become a price-insensitive owner, the fraction of the equity row it occupies is the measure of how much that matters. When Chapter 20 estimates a demand system, the holdings in this table are, quite literally, its left-hand side.

The instruction is therefore practical: when a later chapter makes a claim about a holder, come back to this table and check the magnitude before accepting the mechanism. A mechanism operating on a $200 billion position and a mechanism operating on a $15 trillion position are different claims about the world, even when the theory is identical.

***

> **Box 2.2 — The Rest of the World as Holder**
>
> Foreign investors hold roughly $8.5 trillion of US Treasury securities and roughly $12 trillion of US corporate equity — in the first case the largest single holder class in the market, in the second something close to a fifth of the entire US stock market. The composition differs by instrument and by holder type. Treasury holdings are split between official holders, chiefly foreign central banks and sovereign wealth funds accumulating dollar reserves, and private holders, chiefly foreign insurers, pension funds, and asset managers reaching for yield and safety. Equity holdings are overwhelmingly private and increasingly arrive through global index products in which the US carries a provider-set weight, so that a European pension fund holds Apple for reasons having nothing to do with Apple. What these holders share is that their demand responds to variables — reserve accumulation, currency hedging costs, home-market yields, index reclassification — that appear nowhere in a domestic asset pricing model. The mechanics of all of it, including the dollar's international role, the hedged-investor basis, and the cross-border capital flows that produce these positions, belong to *International Finance*. This book takes the positions as given and asks what they do to prices.

***

> **Box 2.3 — Digital Claims on the Map**
>
> "Crypto" names a technology, and this book is organized by claims rather than by technology. Read the assets that way — what is promised, by whom, ranking where — and most of them sort onto the map.
>
> A *stablecoin* is a short-term claim on a portfolio of reserves, redeemable at par on demand. That is an uninsured money market fund share by construction: the promise is stable value and same-day redemption, and the assets behind it are Treasury bills, repo, and bank deposits — the money market fund column of §2.5 with the fund rulebook and the official backstop removed. The larger issuers have become substantial holders of Treasury bills, which puts them in the Treasury row of Table 2.5 whether or not the accounts carry a line for them. The de-pegs of 2022 and 2023 came from that structure and not from the ledger: one par promise had no real reserve behind it, another's reserve sat partly in a bank that failed, and holders redeemed at once. That is the Reserve Primary Fund with different nouns. Chapter 16 §16.4 places the instrument on the hierarchy of money; §16.5 supplies the run.
>
> An *unbacked cryptoasset* is the opposite case: a claim on nothing. No issuer's balance sheet stands behind it, no promise is enforceable against anyone, and it ranks nowhere, being a residual with nothing underneath. That is why it is absent from the master map's issuer-holder grid — a row requires an issuer — and why a book about claims and their holders treats it here and nowhere else.
>
> *Tokenization* changes the rail, not the claim. A Treasury bill recorded on a distributed ledger is a Treasury bill: same issuer, same promise, same row. What changes is settlement speed and collateral mobility, a market-structure question of the kind Chapter 11 handles. A *central bank digital currency* would be the genuinely new entry — a direct liability of the central bank held by the public: the top of Chapter 16's hierarchy, now reserved for banks, opened to households.
>
> The box is a box because the claims are novel and the analysis is not.

***

> **Box 2.4 — The Household Sector Is a Residual**
>
> Section 2.3 warns twice that "households and nonprofit organizations" is computed rather than measured. The warning is easy to read past, so here is what it means in practice.
>
> The Financial Accounts build most instrument tables from the issuer's side and the institutions' side. The total amount of an instrument outstanding is known because somebody issued it and reports the amount: Treasury debt from the Treasury, corporate bonds from the issuers and the underwriters, mutual fund shares from the funds. The amounts held by identifiable institutions are known because those institutions file — call reports from banks, statutory schedules from insurers, 13F filings from large managers, N-PORT from registered funds. The household row is then the subtraction: total outstanding, less everything the filers reported holding. It is not a survey of households, and no household is ever asked.
>
> Three things follow, and each of them changes how a number in this chapter should be read.
>
> **Every measurement error somewhere else lands here, with the sign reversed.** If an institutional sector's holdings of corporate bonds are understated by fifty billion dollars, households are overstated by fifty billion dollars. The residual is where the accounts put their ignorance, and it is the largest row in the table.
>
> **Entities that nobody files for live here.** Domestic hedge funds, private trusts, personal holding companies, and domestic nonprofit organizations are all inside the line. The sector's name says the last of those out loud and is silent about the rest. A levered trading sector of considerable size therefore sits, invisibly, inside the row a reader is most likely to picture as retirees and index funds — which is the whole motivation for the Federal Reserve's Enhanced Financial Accounts project, whose point is to break out from the residual the sectors the published accounts cannot see. Chapter 18 §18.6 is about holders that this table can only see as an absence.
>
> **The row moves when the accounting changes, not only when the world does.** Break out a new sector and the household line falls by the whole amount, revised backward through history, with no household having sold anything. A time series of the residual is therefore a joint record of portfolio behavior and of what the statistical agency learned to measure, and the two are not separable from the published series alone.
>
> The reading rule is the one §2.3 gives: **when the household row shows a large direct holding of an instrument no ordinary household owns — commercial paper, agency securities, foreign bonds, syndicated loans — read the residual, not the household.** Where household holdings must be right rather than merely large, use a source that asks households directly: the Survey of Consumer Finances, which Chapter 14 §14.2 uses for exactly this reason, and the Distributional Financial Accounts, which reconcile the survey to the aggregate and are what make Chapter 14 §14.6's concentration statements possible. Appendix B §B.1 sets out the construction in full and §B.3 lists the alternatives.

***

## Elsewhere in the Series

* **The Flow of Funds as a measurement system, its national-accounts context, sectoral balances, and the macroeconomics of financial imbalances** — *Institutionalist Macroeconomics*, Chapter 3. This chapter keeps the balance-sheet primer a reader needs to proceed and points onward for the accounting system itself.
* **All cross-border positions**: the dollar's international role, reserve accumulation, currency hedging, global capital flows, and the mechanics behind Box 2.2 — *International Finance*, especially Chapters 2, 11, and 14.
* **The map read as an ecology — who holds which claim, and what their constraints do to prices** — this book. That reading is the subject of Part IV and the organizing question of the whole volume.
* **Data sources, free and licensed, including the Z.1 file structure** — Appendix B.
* **Accounting and financial statements: the balance-sheet identity, book value against market value, and how the three statements articulate** — Appendix C, which assumes no accounting course and takes §2.1 as its only prerequisite.

***

## Summary

1. **Every financial asset is somebody's liability.** Financial claims come in pairs. Summed across all sectors, financial assets and liabilities cancel, so aggregate net worth equals the economy's real assets plus its net foreign position. Finance allocates the ownership of wealth and the risk attached to it; it does not create wealth by issuing claims.
2. **Gross positions are not net positions, and both matter.** The nurse's $180 passes through eight or nine layers of claim, each adding to the gross size of the system without changing her net exposure. Counterparty risk, balance-sheet capacity, and regulatory constraints all bind on the gross figure, which is why the netted number in an aggregate table is never a statement about who bears risk.
3. **A claim is defined by what is promised, how it is enforced, and where it ranks.** Debt promises a schedule and is enforced through bankruptcy; equity promises a residual and is enforced through governance; derivatives promise a payoff contingent on something else and net to zero in supply; hybrids exist because the debt-equity boundary is a legal and tax construct laid over a continuum (Table 2.2). Digital claims are classified on the same three attributes, which is what Box 2.3 does with them.
4. **The Financial Accounts (Z.1) measure both sides of every claim.** The release, descended from Morris Copeland's 1952 moneyflows project, reports for each sector what it owns and owes, instrument by instrument, so that holdings summed across holders equal the amount issued. The household sector is a residual and contains hedge funds, personal trusts, and nonprofits.
5. **Five sector balance sheets carry most of the map.** Households hold net worth on the order of $180 trillion; nonfinancial corporations show a balance-sheet net worth well below the market value of their equity; the federal government's largest liability is the largest single claim class in the system; banks run roughly ten-to-one leverage; and the fund, insurance, and pension complex owes nearly all of what it owns to its own claimants because it exists to sit in the middle (Tables 2.3 and 2.4).
6. **Table 2.5 is the master holdings table and the object later chapters cite.** Rows are claim classes, columns are holder sectors. Ownership concentration varies enormously by instrument — municipal bonds are held almost entirely by tax-paying US holders, Treasuries by everyone — and pension entitlements are a $33 trillion claim class with exactly one holder sector, which is the clearest demonstration in the chapter that the map counts layers rather than wealth.
7. **Fifty years produced three reallocations, all on the holder side** (Table 2.6): households' direct equity ownership gave way to intermediated and foreign ownership; defined benefit pensions gave way to defined contribution plans and IRAs; and bank-held loans funded by deposits gave way to market-based finance and funds. Bonds rose from under half to roughly two-thirds of nonfinancial corporate credit-market debt.
8. **Gross intermediation grew faster than net saving.** The finance and insurance share of GDP roughly doubled while household saving did not rise, with the growth concentrated in asset management and household-credit fee income — more layers between the saver and the real asset, whose cost Chapter 27 evaluates.
9. **Read the table across for issuers and down for holders.** Across: who funds the government, who bears equity risk, who holds corporate credit. Down: what a pension fund's liability does to its asset side, why a money market fund holds what it holds, why a bank's mismatch is its business model.
10. **Aggregates hide distributions.** Households' $25 trillion of directly held equity is concentrated in the top decile of the wealth distribution. That concentration is a pricing input, not a footnote, and Chapter 14 develops it.

***

## Key Terms

* **Claim**: A promise held by one party against another, defined by what is promised, how the promise is enforced, and where it ranks in seniority
* **Balance sheet**: A statement of what an agent owns and owes at an instant, with net worth as the residual that forces the two sides to balance
* **Sector**: A grouping of economic agents with similar function used in the national and financial accounts — households and nonprofits, nonfinancial business, government, depository institutions, funds, insurers and pensions, and the rest of the world
* **Instrument**: A standardized class of claim (Treasury security, corporate bond, deposit, equity share) reported as a row in the accounts and identified by its issuer, promise, and seniority
* **Intermediation**: The insertion of one or more balance sheets between an ultimate saver and an ultimate borrower, each layer transforming maturity, liquidity, credit risk, or denomination
* **Financial Accounts of the United States (Z.1)**: The Federal Reserve's quarterly statistical release reporting sector balance sheets and intersectoral claims for the US economy, published since the 1950s and known by its release code
* **Flow versus stock**: A flow is a quantity per unit of time (saving, issuance, fund inflows); a stock is a quantity at an instant (wealth, debt outstanding, assets under management). The Financial Accounts publish matched flow and stock tables, with revaluation reconciling the two
* **Residual sector**: A sector whose holdings are computed as the difference between known totals and all other measured sectors — in Z.1, households and nonprofits, which therefore absorbs hedge funds, personal trusts, and measurement error
* **Seniority**: The order in which claims are satisfied when assets are insufficient; the content of the term "capital structure"

***

## Readings

### Required

* Board of Governors of the Federal Reserve System. *Financial Accounts Guide* (online, `federalreserve.gov/apps/fof/`), together with the current Z.1 release, *Financial Accounts of the United States*. *The guide indexes every table by code and explains what each series measures and how it is estimated. Read the guide's description of the household sector as a residual before using any household figure in this chapter.*

### Recommended

* Copeland, M. A. (1952). *A Study of Moneyflows in the United States*. National Bureau of Economic Research. *The origin document. The introduction and the statement of the accounting framework are what to read; the empirical chapters are of historical interest.*
* Mehrling, P. (2011). *The New Lombard Street: How the Fed Became the Dealer of Last Resort*. Princeton University Press, Chapter 1. *The balance-sheet way of seeing, stated compactly. Read it as the argument that you cannot understand a financial event without drawing the balance sheets of everyone involved.*
* Godley, W. and M. Lavoie (2007). *Monetary Economics: An Integrated Approach to Credit, Money, Income, Production and Wealth*. Palgrave Macmillan, chapters 1-2. *Sectoral-balance accounting used as a modeling discipline: because every financial asset is somebody's liability, sector surpluses must sum to zero, and a model that violates the identity is rejected before it is estimated. Section 2.1's identity, taken as a constraint rather than an observation.*
* Merton, R. C. (1982). *Finance Theory*. Unpublished lecture notes, MIT Sloan School of Management, chapters IV and V. *Chapter IV, "On the Role of Business Firms, Financial Instruments and Markets," derives the instrument taxonomy of §2.2 from what each instrument does rather than from what it is called; chapter V does the same for intermediation, and is §2.5's argument in its original setting.*
* Greenwood, R. and D. Scharfstein (2013). "The Growth of Finance." *Journal of Economic Perspectives* 27(2): 3-28. *Decomposes the doubling of finance's GDP share into asset management and household credit, and asks what society got for it. The empirical backbone of §2.4's last paragraph.*
* Philippon, T. (2015). "Has the US Finance Industry Become Less Efficient? On the Theory and Measurement of Financial Intermediation." *American Economic Review* 105(4): 1408-1438. *The complementary measurement: the unit cost of intermediation, and the finding that it has not fallen.*
* Bodie, Z. and R. C. Merton (2000). *Finance*. Prentice Hall (later editions, with D. L. Cleeton, as *Financial Economics*). *The standard institutional map in its textbook form, organized by sector and function. Read the first two chapters as the treatment §2.2's claim-first taxonomy defines itself against.*
* Investment Company Institute, *Investment Company Fact Book* (annual, free online). *The fund industry's own statistical compendium; the cleanest free source for the fund and retirement rows of Tables 2.4 and 2.6.*

***

## Discussion Questions

1. **Trace a claim chain.** Pick a financial asset you or your family holds — a bank deposit, a pension entitlement, a brokerage position, an insurance policy. Draw every balance sheet between you and the real asset your claim ultimately rests on, naming the institution at each layer and what each layer provides. Then ask, for each layer: if this institution failed, would your net exposure change, or only your access to it? Which layers could be removed, and what would be lost?
2. **What netting hides.** Table 2.5 shows households directly holding roughly $25 trillion of a roughly $60 trillion equity market, and Table 2.6 shows that direct share falling for fifty years. A commentator concludes that American households have reduced their exposure to equity risk. State precisely what is wrong with the inference, identify the rows of Table 2.5 that repair it, and describe the accounting you would do to measure households' *total* equity exposure, direct and indirect. What would remain unmeasured even after you did it?
3. **Why gross grew faster than net.** Between 1975 and today, the finance and insurance share of GDP roughly doubled while the household saving rate did not rise. Offer three candidate explanations — one in which the additional layers are productive, one in which they are rents, and one in which they are a response to a change in what savers demand. What evidence would distinguish them? Which does the Greenwood-Scharfstein decomposition support, and how strongly?
4. **A claim class with one holder.** Pension entitlements are $33 trillion held entirely by households, and they are a claim on assets counted elsewhere in the same table. Explain to a reader who has just added the column why the total is not the wealth of the United States. Then construct the opposite error: describe a policy question for which counting entitlements *separately* from the underlying assets is exactly the right thing to do, and say why.
5. **Redraw the map for another country.** Choose a country with a different financial structure — Germany, Japan, or China. Without looking up the figures, predict how its version of Table 2.5 would differ: which rows would be larger, which holder columns would dominate, which claim classes might not exist at all. Then find the country's flow-of-funds or financial accounts release and check. Where were you wrong, and does the error reflect a difference in institutions or a difference in measurement convention?

***

## Data Exercise: Rebuild the Map

Everything in this exercise runs on free public data. This is the chapter's flagship exercise, and the tables it produces are the ones later chapters ask you to have.

**Part A — Who holds Treasuries (free data).** Download the current Z.1 release from the Federal Reserve's website, either as the full PDF or, better, as the CSV data package. Locate table **L.210, Treasury Securities**.

1. Extract the level of Treasury securities outstanding and the holdings of each sector for the most recent quarter. Rebuild the Treasury row of Table 2.5 with exact figures. Confirm that holdings sum to the total, and identify what the accounts do with any discrepancy.
2. Build the same row for every fourth quarter since 2005. Plot each holder sector's share over time as a stacked area chart. Mark on the chart the beginning of the first large-scale asset purchase program, the beginning of balance-sheet reduction in 2017, the March 2020 expansion, and the resumption of runoff in 2022. Describe what the Federal Reserve's row does, and — the harder question — describe which other sector's row moves in the opposite direction each time.
3. The household row of L.210 rose substantially after 2021. Consult the Financial Accounts Guide on how the household sector is constructed, and explain why practitioners read a large part of that increase as hedge fund positions rather than as household ones. What independent data would you use to check this? (The Office of Financial Research and CFTC position data are the usual sources; Chapter 9 uses them.)

**Part B — One sector's balance sheet (free data).** Locate table **B.101, Balance Sheet of Households and Nonprofit Organizations**.

1. Rebuild Table 2.3's household row exactly: financial assets, nonfinancial assets, liabilities, net worth. Then decompose financial assets into their five largest components and nonfinancial assets into real estate and consumer durables.
2. Construct the ratio of household net worth to GDP annually since 1960, using the BEA's GDP series from FRED. The ratio rises substantially from the mid-1990s. Decompose the change into contributions from (i) saving, using the corresponding flow table F.101, and (ii) revaluation of existing assets. Which dominates, and in which sub-periods?
3. Repeat step 1 for **B.103, Nonfinancial Corporate Business**. Compute the ratio of the market value of corporate equity outstanding to the balance-sheet net worth. This is a version of Tobin's q; note its current value and its 1980 value, and hold the number for Chapter 22.

**Part C — The DB-to-DC shift (free data).** Using Z.1 tables **L.117 through L.120** (private and public pension entitlements, and the DB/DC split reported within them):

1. Plot defined benefit and defined contribution entitlements for private plans, in nominal dollars and as shares of the total, from the earliest available date to the present.
2. Add state and local government pension entitlements to the plot. The public-sector line behaves differently from the private-sector line; describe how, and offer an institutional explanation.
3. Add IRA assets, available from the ICI *Fact Book* or from the Z.1 supplementary tables. The combined DC-plus-IRA line is the one Chapter 14 uses. State what fraction of US retirement wealth now sits in vehicles where the household bears the investment risk directly.

**Part D ★ — Concentration (free data, harder).** The Federal Reserve's **Distributional Financial Accounts (DFA)** distribute the Z.1 household balance sheet across wealth percentiles quarterly.

1. Download the DFA and compute the share of directly held corporate equity and mutual fund shares owned by the top 1 percent, the next 9 percent, the next 40 percent, and the bottom 50 percent.
2. Repeat for pension entitlements and for real estate. The three assets have very different distributions. Describe them, and explain in one paragraph why an asset pricing model calibrated to *aggregate* household consumption may be calibrated to the wrong household.
3. State the DFA's chief methodological limitation — that it distributes the aggregate using survey shares from the Survey of Consumer Finances rather than measuring the distribution directly — and say what that implies for the precision of your answer to step 2. Chapter 14 builds on this exercise.