> 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_01_introduction.md).

# Chapter 1: What Financial Economics Is About

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

***

## Opening Episode: The Day Oil Cost Less Than Nothing

On Monday, April 20, 2020, the New York Mercantile Exchange's May contract for West Texas Intermediate crude oil settled at minus $37.63 a barrel.

That is not a typographical error and it is not a figure of speech. A contract covers a thousand barrels, so the seller of one contract handed the buyer $37,630 *and* the oil. The previous session had settled at $18.27, which makes the day's move $55.90 — the largest one-day change in the contract's history, and the first time it had ever printed below zero.

No reading of that number as "the value of oil" survives contact with the week it happened in. Americans were still buying gasoline on April 20, at a positive price. Refiners were still running. And the June contract — a claim on the same grade of crude, at the same delivery point, five weeks later — traded near $20 all day and closed there. Two claims on the same commodity, differing only in a date, were separated by nearly sixty dollars a barrel.

So the date is where the explanation has to start, and in a physically settled futures contract a date is not a forecasting horizon. It is an obligation. The May contract stopped trading the following afternoon, and whoever was long at that point was contractually required to take delivery of crude oil at Cushing, Oklahoma, during the month of May.

Three facts about Cushing and about who was long turn an impossible price into an ordinary one.

**Storage was the binding constraint.** Cushing's working storage capacity is on the order of 76 million barrels. The Energy Information Administration's weekly series put stocks there at roughly 55 million barrels on April 17 and rising by several million a week as the pandemic collapsed demand. On paper that leaves room; in practice much of the unused capacity was already leased to somebody, and by the third week of April the price of the remaining tankage had gone vertical. Taking delivery required somewhere to put the oil, and the price of *somewhere* was, for practical purposes, unbounded.

**A large part of the long side had no use for barrels.** Retail investors had spent March and April deciding that oil was cheap and buying it the only way available to them, through exchange-traded products that hold futures and roll them forward mechanically. The largest of these, the United States Oil Fund, took in money fast enough that its position came to represent a substantial share of front-month open interest — large enough that the Commodity Futures Trading Commission's interim staff review of the episode, published in November 2020, discussed the fund's roll schedule at length. By April 20 that fund had already rolled out of May and into June, which is part of why the expiring contract was thin and part of why June did not follow it down. The investors who remained long May into the last two sessions were, disproportionately, holders who had not planned on owning oil.

**The plumbing had to be altered before the price could exist.** The exchange had modified its systems earlier that month to permit negative prices in its energy contracts, and had told the market so. Until that change, the price was not merely unlikely; it was unrepresentable.

Put the three together and the negative price stops being a statement about oil and becomes a statement about a claim. It is the price of an obligation to receive crude at a place with nowhere left to put it, held by people who could not take it, on the last day it was possible to escape. What it measures is the cost of getting out.

That is the whole book in one afternoon. A price is a fact about a particular claim, held by particular people, under particular constraints. Ask what the claim is, who holds it, and what the holder is able to do, and prices that look like arithmetic errors become legible. Skip those questions and you are left with the version of April 20 that circulated at the time, in which oil was worth less than nothing.

***

## 1.1 The Organizing Question

This book is organized around one question, asked in three parts:

**Who holds which claims, how are those claims priced, and what do the holders' constraints do to the prices?**

Each noun in that sentence is a piece of machinery, and each gets a chapter before it gets used.

A **claim** is a contract entitling its holder to payments contingent on something. That is deliberately broad: a bank deposit, a Treasury note, a share of stock, a call option, an interest in a private fund, and a pension entitlement are all claims, and they differ in what the payments are contingent on and in who owes them. The first fact of the subject is that every financial claim is somebody's asset and somebody else's liability, which means the financial system does not create wealth by issuing claims — it allocates the ownership of wealth and the risk attached to it. Chapter 2 builds the taxonomy and the map: what instruments exist, who issues them, who holds them, and how the answer has changed over fifty years.

A **price** is what someone paid for a claim, and the theory of prices in this book is one equation with many specializations. Chapter 3 derives it from the law of one price and works it fully in a two-date, two-state economy: the price of a claim is the expected value of its payoff weighted by a stochastic discount factor,

$$
p = E\[mx].
$$

Everything in Part II is a statement about what $$m$$ is. The CAPM says $$m$$ is linear in the market return; the consumption-based model says it is marginal utility growth; Black-Scholes says that if you can trade continuously, $$m$$ is pinned down entirely by the absence of arbitrage. Chapter 3 closes by asking the question the rest of the book answers: *whose* $$m$$? Marginal utility belongs to a person, and the equation is silent on which person.

A **holder** is the answer. Part IV takes the question seriously and finds that it has a measurable answer, because holdings are reported: 13F filings, statutory insurance schedules, and the Federal Reserve's Financial Accounts say who owns what. A holder who is constrained — by a mandate, a capital charge, a redemption term, a benchmark, an index rule, or the simple fact of never having chosen anything — supplies size to a market without supplying price discipline to it, and that is a fact about prices, not about the holder. Chapter 20 assembles the ecology into an estimated demand system and answers Chapter 3's question in a line: everyone's $$m$$, weighted by wealth and by constraint, with the weights measurable.

The three parts of the question are one question, and that is why this is one volume rather than two. Most curricula split asset pricing from corporate finance and teach them in different rooms, which leaves a gap exactly where the interesting things happen. Firms design and issue the claims that investors hold; investors' constraints determine which claims are expensive; and firms manufacture the claims that are expensive. Chapter 23 §23.7 shows corporations lengthening or shortening the maturity of what they issue to fill gaps in the supply of government debt, and Chapter 25 §25.6 generalizes it: every financing door a firm can walk through exists because some balance sheet somewhere wants the claim behind it. Neither observation is available to a course that never puts the issuers and the holders on the same page.

Three commitments follow from the question, and they are worth stating so that a reader can hold the book to them.

**Theory is taught against the institutional map.** A model of who the marginal investor is can be checked against data on who actually holds the claim. Chapter 2's master holdings table exists so that every later mechanism can be sized before it is believed. 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 algebra is identical.

**Evidence comes with named sources.** Every table in this book states where its numbers came from, every data exercise runs on sources a reader can download, and magnitudes taken from a single study are attributed to that study rather than presented as facts of nature.

**Contested things are marked as contested.** The exemplar is Chapter 20 §20.5, which prints a table of the demand-system literature's claims with each one graded *established*, *widely accepted*, or *contested* — including the finding that the book's own flagship chapter is built on. Teaching a live literature honestly means saying which parts of it would survive a hostile referee and which would not.

> **Box 1.1 — What a physically settled contract obliges**
>
> The reason a barrel of oil could trade below zero is written into the contract, and it is worth reading once, because every later chapter's claims are also contracts with terms.
>
> The NYMEX light sweet crude oil future is *physically settled*. One contract is a thousand barrels of a specified grade, and the seller who holds a short position to the end must deliver those barrels FOB at Cushing, Oklahoma, ratably across the delivery month. The buyer who holds a long position must take them. Trading in a delivery month ends three business days before the twenty-fifth of the preceding month, which is why the May 2020 contract stopped trading on 21 April and why 20 April was the last full session in which a holder could still get out by selling rather than by arranging a tank.
>
> Every clause is doing work. "Physically settled" means the contract is not a bet on a number; it is an obligation to move a commodity. "FOB Cushing" names a point on a map with finite storage, which in April 2020 was nearly full. And the trading deadline means that a holder who had not closed out by the twenty-first faced a choice between finding storage he did not have and paying someone else to take the obligation. The price at which he would pay is unbounded below, because the cost of failing to perform is unbounded. That is the whole of it.
>
> One institutional detail completed the picture, and it is a nice one. Exchange systems had been built on the assumption that a commodity price is positive. In the days before 20 April, CME issued notices confirming that its clearing systems could represent negative prices, and that if prices went negative it would switch its WTI options from the log-normal Black model to a normal one — the Bachelier model, named for the thesis in Box 1.2. A distribution that had been a historical curiosity for a century became, for a fortnight, the only one that could price the contract at all.
>
> Chapter 8 §8.2 prices this future properly, with the cost of carry and the convenience yield that Cushing's tanks determine, and Figure 8.7 draws the curve on the day. What belongs here is only the shape of the thing: a price is a number attached to an obligation, and when the obligation binds on somebody who cannot perform it, the number goes wherever it has to.

***

## 1.2 Four Readings of a Price Move

The oil episode teaches a habit, and the habit is worth naming before it gets used, because it replaces a reflex most readers arrive with. The reflex is to read a price move as information: the number changed, therefore the world changed, therefore something is now known that was not known the day before. April 20 is a counterexample delivered at maximum volume. Nothing about crude oil was learned that afternoon. What moved was a claim, held by people who could not perform on it, in the last hours during which they could stop being its holders.

So before a price move can be read as information, four readings have to be separated. They are not mutually exclusive and they are frequently simultaneous, which is precisely why they have to be listed rather than intuited.

**News.** The claim's cash flows changed, or their riskiness did, and the price moved because the discounted value moved. This is the reading the textbook assumes, and in a narrow window around an identified announcement it is usually the right one; Chapter 7 §7.3 builds the event-study machinery that makes it testable.

**A constraint binding.** Nothing about the claim changed; something about the balance sheet holding it did, and the claim moved to a less willing owner at whatever price a less willing owner required. In March 2020 the ten-year Treasury yield *rose* through the worst fortnight of a global panic — a flight to quality running backwards — because levered basis traders were meeting margin calls and the dealers who would ordinarily have absorbed the selling had no balance sheet left to absorb it with. Chapter 19 opens there; Chapter 16 §16.5 states the mechanism once for the whole book.

**A flow.** Money arrived or left for reasons that have nothing to do with the claim — a payroll date, an index rule, a mandate — and met a demand curve steeper than the frictionless model says demand curves can be. When Tesla was added to the S\&P 500 in December 2020, every fund that promised to track the index had to own it by a particular Monday morning, in a particular weight, at whatever price the closing auction printed. The stock rose roughly 70 percent between announcement and inclusion and fell on its first day inside. Nobody in that sequence formed a view about electric vehicles. Chapter 17 opens on that Monday; Chapter 20 §20.3 turns the mechanism into a number.

**A model becoming convention.** The price is what the standard model says the price should be, because everyone is running the standard model. Implied volatilities were roughly flat across strikes until October 1987 and have sloped downward in every liquid equity index since — and the formula the slope contradicts is still the formula the slope is quoted in, Black-Scholes surviving as a coordinate system after failing as a theory. Chapter 7 §7.6 names the phenomenon performativity, Chapter 8 §8.5 works the case, and Box 26.1 shows a risk measure doing the same thing to the institutions that share it.

Now read April 20 through all four. The news reading is available and small: demand for refined products had collapsed, which is why the whole curve was low — it explains the level of oil and none of the day's move, since June sat near $20 while May printed minus $37.63. The constraint is the binding one: Cushing's remaining tankage was the capacity that ran out, and the price of an obligation to take delivery into a full tank is the price of the tank. The flow is the exchange-traded product complex, which had bought the front month on retail conviction and rolled out of it days before expiry, leaving the contract thin and its surviving longs mismatched. And the convention is the plumbing: until the exchange rewrote its systems that month, a negative price was not improbable but unrepresentable — a model constraining a price in the most literal sense available.

Four readings, one afternoon, and the only one that cannot carry the number on its own is the one that ran in the newspapers.

The question is not asked once and retired. Every chapter in Parts II, III and V carries a holder section, and every one of those sections is this question asked about a particular claim: who holds it, what binds the holder, and which of the four readings a move in its price is therefore reporting. Chapter 27 §27.5 answers the question in full, with the whole book's machinery behind it and the consequences for anyone who reads prices for a living attached. Until then the discipline is the short one — ask which of the four is doing the work before deciding what a price has told you.

***

## 1.3 The Arc of Ideas

Financial economics has a short history and a legible one. It runs in three eras, and the third is the one this book is written from.

### Prices without holders

The subject begins twice. Louis Bachelier's 1900 Sorbonne thesis derived the mathematics of Brownian motion to model the Paris bond market five years before Einstein used it for pollen, obtained an option-pricing formula, and was ignored for more than half a century. The second beginning stuck. Harry Markowitz's "Portfolio Selection" (1952) — fourteen pages, no data — replaced "pick good securities" with "choose a point on a frontier," and in doing so made a security's risk a property of the portfolio it sits in rather than of the security. Chapter 4 builds the apparatus and opens with the story of how Markowitz invested his own retirement account.

Franco Modigliani and Merton Miller (1958) applied the same kind of argument to the other side of the balance sheet: in a frictionless market the value of a firm does not depend on how its claims are packaged, because an investor can repackage them at home. The proposition is famous; what makes it useful is its assumption list, which became the research agenda of corporate finance for sixty years. Chapter 23 reads it that way.

William Sharpe (1964) turned Markowitz's decision problem into an equilibrium. If everyone solves the same problem, everyone holds the same risky portfolio, that portfolio must be the market, and expected returns line up with covariance against it. The CAPM is the first complete theory of asset prices, and it is Chapter 4's destination. Eugene Fama (1970) gave the informational side its statement, defining efficiency relative to an information set and naming the joint-hypothesis problem that has constrained every test since — Chapter 7's subject. And in 1973 Fischer Black, Myron Scholes and Robert Merton published a formula for an option that contains no expected return and no risk aversion, in the same season that the Chicago Board Options Exchange opened for business in a converted smoking lounge. Chapter 8 opens there.

What unites the era is an absence. In every one of these results the identity of the investor is either irrelevant or assumed away: the CAPM's investors are identical, Modigliani-Miller's can undo any packaging, and Black-Scholes's need not agree about anything except the absence of arbitrage. That absence was a feature. It is what let a fourteen-page paper with no data reorganize an industry, and what let a formula with five inputs be programmed into a handheld calculator and sold on the floor of an exchange within a year of publication — the beginning of a story Chapter 7 §7.6 tells under the heading of performativity, in which a model does not merely describe a market but becomes part of the machinery that makes its prices.

### Anomalies

Then the evidence arrived. Robert Shiller (1981) showed that stock prices move far more than the discounted stream of dividends they eventually paid — a volatility bound the data violate by a wide margin, and the sharpest single challenge to fundamental valuation (Chapter 7 §7.4). Daniel Kahneman and Amos Tversky (1979) documented that people evaluate gambles relative to a reference point, with losses looming larger than gains, which is not a small amendment to expected utility (Chapter 15). Fama and French (1992) found that size and book-to-market forecast average returns and that beta, controlling for them, did not — the finding Chapter 6 opens with, and the one Fama himself read as the death of a model of expected returns rather than of efficiency.

The anomalies themselves were the smaller half of the story. Any single one of them could be, and usually was, met with the joint-hypothesis reply: a return the model does not explain is evidence against the model, not against the market. What made the era decisive was that the replies began to require a theory of why the market did not simply erase the pattern — and the theory of expected returns has no room for such a question, because in it any mispricing is arbitraged instantly by someone with unlimited capital and no career.

The era's decisive theoretical contribution was therefore about why none of this gets arbitraged away. Sanford Grossman and Joseph Stiglitz (1980) showed that a fully revealing price destroys the incentive to acquire the information that would reveal it, so equilibrium requires prices that are informative but not perfectly so (Chapter 7 §7.2). Andrei Shleifer and Robert Vishny (1997) added the institutional half: real arbitrage is conducted by specialists managing other people's money, and their capital is withdrawn precisely when mispricing is widest, so arbitrage is weakest exactly where it is most needed (Chapter 15 §15.5).

### Holders brought back in

Shleifer and Vishny's argument is the hinge, because it makes the identity and the balance sheet of the arbitrageur part of the pricing problem. The literature that followed took the point seriously.

Markus Brunnermeier and Lasse Heje Pedersen (2009) formalized the spiral in which falling prices raise margins, forced deleveraging lowers prices further, and market liquidity and funding liquidity destabilize each other (Chapter 16 §16.5, which states the mechanism once for the whole book). Zhiguo He and Arvind Krishnamurthy (2013) wrote asset pricing with the intermediary sector's equity capital as the state variable, so that the pricing kernel belongs to a constrained institution rather than to a household — Chapter 19's body, opened by the March 2020 Treasury dislocation.

The most recent turn makes the holders the object of estimation rather than a friction on top of a model. Ralph Koijen and Motohiro Yogo (2019) estimate every institution's demand for every stock as a function of characteristics and price and let market clearing determine prices — the academic statement of this book's title. Xavier Gabaix and Koijen (2021) push on the aggregate: a dollar of informationless flow into the equity market appears to move aggregate market value by several dollars, which makes flows a first-order determinant of prices. Chapter 20 teaches both, and marks which parts of them are contested.

The arc is not a story of error corrected. Every era's machinery is still in use, and this book teaches all of it. What changed is what the equation is understood to be silent about.

![Figure 1.1: The arc of ideas, 1900-2025](https://846781005-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3EupdX99vVBoNySDtmxb%2Fuploads%2Fgit-blob-dbc9ea8fc48362ae3402643e87290513a4a9b5f9%2Ffig_01_01_the_arc_of_ideas.png?alt=media)

**Figure 1.1: The arc of ideas, 1900-2025.** The three eras of §1.3, with the chapter that teaches each entry attached to it. The dates are printed but the spacing is not proportional to them, because fourteen of the fifteen entries fall inside a seventy-year window and would be illegible on a time axis. Two features of the list are worth more than any single row. The first is the gap at the top: Bachelier derived the mathematics of Brownian motion for the Paris bond market five years before Einstein used it for pollen, obtained an option-pricing formula, and was left unread for fifty-two years — a reminder that the field's chronology is a history of what was taken up, not of what was known. The second is the direction of travel down the third column. The first era's entries are propositions about prices in which the identity of the investor is either irrelevant or assumed away; the second era's are findings that the propositions do not hold; and the third era's are all, without exception, statements about who is holding the claim and what they are permitted to do. That is the arc this book is organized around, and none of it has been discarded: the machinery of every era is still in use, and the change is in what the pricing equation is understood to be silent about.

> **Box 1.2 — Bachelier, 1900, and the fifty-year silence**
>
> Louis Bachelier defended *Théorie de la spéculation* at the Sorbonne in 1900. To describe the price of *rentes* on the Paris exchange he constructed a continuous random walk, derived the diffusion equation its transition density satisfies, and used the result to value options — five years before Einstein used the same mathematics for suspended particles, and more than twenty before Wiener made the underlying process rigorous.
>
> Henri Poincaré's report on the thesis is the document to read. It is admiring and guarded at once: it praises the originality, notes that the subject is remote from those the candidates usually treat, and awards a grade below the one an academic career in French mathematics then required. Bachelier spent decades in provincial posts, and the thesis went quiet.
>
> It stayed quiet for half a century, and the way it came back is the part worth keeping. In the mid-1950s Leonard Jimmie Savage, reading in the history of probability, came across Bachelier and sent postcards to economists he thought would care. Paul Samuelson got one, went to the library, and found the thesis. What followed — Samuelson's own work on warrant pricing, Osborne's on the log-normality of returns, and eventually the Black-Scholes-Merton solution — is the line Figure 1.1 draws.
>
> Two lessons, and the second is the one this book uses. The first is the ordinary one about priority: a field can mislay a result for fifty years, and the mislaying is not costless, because option pricing had to be derived a second time from nothing. The second is about what made the thesis findable in 1955 and invisible in 1905. Bachelier's model has no holders in it. He assumed a price process and derived what a claim on it must be worth, which is a way of asking the question that had to become the field's own before anyone could recognize the answer. The arc of §1.3 is the arc of that framing being adopted, pushed as far as it goes, and then partly reversed.

***

## 1.4 How the Book Is Organized

The sequence follows the question: claims, then prices, then holders, then issuers.

**The rest of Part I (Chapters 2-3) supplies the two things everything else consumes.** Chapter 2 is the map — sector balance sheets from the Financial Accounts, an instrument taxonomy, and one master holdings table showing who holds each major claim class, which later chapters cite rather than rebuild. Chapter 3 is the spine: present value, the law of one price, arbitrage, and a two-date, two-state economy worked completely, from which state prices, risk-neutral probabilities and the stochastic discount factor all fall out. It ends on "Whose $$m$$?"

**Part II (Chapters 4-8) is asset pricing, taught as a sequence of specializations of Chapter 3.** Portfolio choice and the CAPM (4); the consumption-based model and its confrontation with the equity premium, which it misses by an order of magnitude (5); factor models, the cross-section, the factor zoo, machine learning, and currency and commodity factors (6); information, efficiency, event studies, predictability, and performativity (7); and derivatives, where replication and the binomial model make risk-neutral valuation concrete and Black-Scholes is stated, interpreted assumption by assumption, and then tested against the volatility smile (8). Each chapter carries a section asking who holds the claim in question and what their constraints do to its price — leverage-constrained investors and betting-against-beta in Chapter 4, factor crowding and capacity in Chapter 6, dealer hedging demand in Chapter 8.

**Part III (Chapters 9-13) prices the major asset markets and is where holder demand becomes visible in the data.** Fixed income, the term structure, and safe assets, with the Treasury market's changing ownership as the book's cleanest demonstration that holder demand moves prices (9); credit risk across corporate, sovereign and municipal issuers, ending with insurers as the marginal holder of corporate bonds (10); microstructure and liquidity, with liquidity treated as both a holder constraint and a priced characteristic (11); equity issuance and the life of the public firm, with index inclusion as the cleanest quasi-experiment available (12); and mortgage securities, where the holder's own hedging of negative convexity moves the market the security is priced against (13).

**Part IV (Chapters 14-20) is the investor ecology, and the part the book is written for.** Households, their participation puzzles and defaults, and the concentration of risk-bearing (14); behavioral finance, seated next to households because the evidence is household evidence (15); institutional investors, the thesis chapter, where liabilities drive asset demand and constrained capital moves prices (16); the passive revolution as an ownership regime (17); private funds and the measurement disputes their returns turn on (18); banks and dealers as holders, and intermediary asset pricing (19); and the demand for assets, which estimates the ecology as a system of demand curves and answers Chapter 3's question (20).

**Part V (Chapters 21-25) turns the book around and looks at the firms that manufacture the claims.** The arc is: the factory (21, why firms exist at all, and why a security is an allocation of cash-flow *and* control rights); the calibration (22, capital budgeting, valuation, real options, Tobin's Q, and mergers); the structure (23, Modigliani-Miller and the three theories of leverage, plus payout); the control (24, covenants, boards, blockholders, activism, and governance in an age of universal owners); and the rationing (25, what firms actually do — the financing menu, the life cycle, the small-firm wall, and what the identification literature can and cannot establish). The part's thesis is that issuance responds to holder demand: firms manufacture the claims that constrained balance sheets are short of.

**Part VI (Chapters 26-27) closes.** Risk management spans firms and institutions — why nonfinancial firms hedge, how value at risk is built and how it fails, and how risk constraints become a pricing mechanism (26). The final chapter reprises Chapter 1's question with the book's machinery, asks what an ecology of constrained holders implies for what asset prices can tell us about the economy, and ends on the book's own thesis rather than on a cross-reference (27).

Four appendices carry the methods and the reference material: empirical methods in finance, self-contained for everything Chapters 6 and 7 require (A); data sources, each flagged free or licensed (B); accounting and financial statements (C); and a mathematical review including the Black-Scholes derivation (D).

***

## 1.5 How to Use This Book

**Prerequisites, stated plainly.** One statistics course at the level of multiple regression, and calculus through partial derivatives. That is all. No measure theory is used anywhere in the body, and none is assumed. The finance-specific empirical methods — portfolio sorts, Fama-MacBeth regressions, event studies, and the machine-learning mechanics Chapter 6 discusses — are developed from scratch in Appendix A with worked examples, so that a course adopting this book does not need a second one. Matrix notation appears in Chapter 4 and is explained where it appears. Where a chapter needs a piece of machinery, that machinery is in this book; cross-references to companion volumes always point onward for depth, never backward for a prerequisite.

**The ★ convention.** A star in a section heading marks a PhD-track section. The body of every chapter is written for the primary reader — an advanced undergraduate or masters student — and is self-contained without the starred material. A reader who skips every starred section loses generality, not the argument; starred sections may assume the starred sections of earlier chapters. The same mark appears on the harder problems and on the optional extensions to data exercises, where it means "if you have licensed data." Fourteen chapters carry starred sections; the rest keep their harder material in the problems.

**The apparatus.** Every chapter opens with an **episode** — a dated, documented event chosen because the chapter's machinery is what makes it legible. April 20, 2020 is this chapter's; the Palm/3Com arbitrage, the 1973 opening of the options exchange, the 2010 flash crash, Tesla's index inclusion, and Fisher Body's dies are others. Chapters in Parts II, III and V each carry a required **holder section** asking who holds this claim and what their constraints do to its price; if you want to know what is distinctive about this book, read those sections first. **Elsewhere in the Series** blocks mark every boundary with a companion volume, in the text where they cannot rot; the map of those volumes is in the preface. Chapters 3-26 carry **problems**; Chapters 3-13 and 19-26 print selected solutions in the chapter, and the full set is in the instructor materials. Every chapter ends with a **data exercise** built on the free-data-first rule: the baseline version runs on sources anyone can download — Ken French's library, FRED, the Financial Accounts, EDGAR and 13F filings, Shiller's data — with starred extensions for readers with licensed access.

**Three paths through the book.** The full course is Chapters 1-27 in order. The PhD-preparation track is that plus every starred section and Appendix A. The practitioner path is Part I read lightly, then Parts III and IV in full, then Chapter 26. Part IV is deliberately modular: its hard prerequisites from the first half of the book are Chapters 2-4 and 6-7, which makes it assignable on its own in an asset-management elective.

**For instructors.** The apparatus that belongs to whoever assigns this book rather than to whoever reads it — the complete list of starred sections, two thirteen-week syllabus maps, and the chapter-dependency table with the diagram that draws it — is collected in the **For Instructors** section of the front matter. A reader working alone can ignore it; a reader building a course should start there.

***

## Elsewhere in the Series

* **The history of economic thought as a subject rather than as motivation** — *History of Economic Thought*. Section 1.3 selects the fifteen entries that lead to the holder-side turn and is silent about almost everything a history has to carry: the schools, the controversies, the institutional settings, and the ideas that went nowhere. That volume owns the genre; this book borrows from it only in §1.3 and in the chapter-opening episodes.
* **Performativity — what it means for a model to stop describing a market and become part of the machinery that makes its prices** — *Philosophy of Economics*, Chapter 11. Section 1.2's fourth reading of a price move states the mechanism in a paragraph and Chapter 7 §7.6 gives it a section, which is as far as a finance text should take it.
* **All cross-border positions** — *International Finance*. The dollar's international role, capital flows, currency risk, and the foreign holder who appears in every column of Chapter 2's map and is treated here only as a holder.
* **The macroeconomics this pricing question sits inside** — *Institutionalist Macroeconomics*. Chapter 2 marks the measurement boundary and Chapter 19 the credit-supply one. What this chapter marks is smaller and prior: financial economics is a part of macroeconomics and not an alternative to it, and the four readings of §1.2 are readings a macroeconomist would recognize.
* **Empirical methods** — Appendix A, and deliberately not a companion volume. Section 1.5 promises that a course adopting this book needs no second one, and the promise is kept inside these covers: portfolio sorts, Fama-MacBeth, event studies and the Chapter 6 machine-learning mechanics are all developed there from scratch. **The mathematics** — Appendix D.
* **The question this chapter asks, answered rather than posed** — this book, Chapters 20 and 27. Part IV is where the holder side stops being a way of framing the subject and becomes the thing that is estimated; §27.5 returns to §1.2's four readings and states which of them the book has taught the reader to tell apart, and which remain genuinely hard.

***

## Summary

1. **A price is a fact about a claim and its holders, not a shadow of intrinsic value.** On April 20, 2020, an expiring oil contract settled at minus $37.63 while the contract five weeks later traded near $20. Nothing about oil explains the gap; delivery obligations, full storage at Cushing, and the composition of the long side do.
2. **Before a price move can be read as information, four readings have to be separated.** News, a constraint binding, a flow, and a model becoming convention (§1.2). April 20, 2020 runs on the last three; the news reading explains the level of oil and none of the day's move. Every chapter's holder section asks which of the four applies to its own claim, and Chapter 27 §27.5 answers the question in full.
3. **The organizing question has three parts.** Who holds which claims (Chapter 2), how are they priced (Chapter 3 and Part II), and what do the holders' constraints do to the prices (Parts III-V). The parts are not independent, which is why the book is one volume.
4. **Everything in asset pricing is a specification of** $$m$$ **in** $$p = E\[mx]$$**.** The CAPM, the consumption-based model, factor models and Black-Scholes are four answers to the same question, and Chapter 3 derives the equation before any of them appears.
5. **The equation is silent about whose marginal utility it contains, and that silence is the book's subject.** Chapter 3 §3.7 asks "whose $$m$$?"; Chapter 20 §20.6 answers it — everyone's, weighted by wealth and by constraint, with the weights measurable from holdings data rather than assumed.
6. **The intellectual history runs in three eras.** Prices without holders (Bachelier, Markowitz, Modigliani-Miller, Sharpe, Fama, Black-Scholes-Merton); anomalies (Shiller, Kahneman-Tversky, Fama-French, Grossman-Stiglitz, Shleifer-Vishny); and holders brought back in (Brunnermeier-Pedersen, He-Krishnamurthy, Koijen-Yogo, Gabaix-Koijen). No era's machinery has been discarded.
7. **The book makes three commitments:** theory checked against the institutional map, evidence with named sources and reproducible free-data exercises, and explicit marking of what is contested — with Chapter 20 §20.5's table as the standard.
8. **The ★ convention makes the reading tracks enforceable.** The body serves an advanced undergraduate or masters reader; starred sections carry the PhD-track material and can be skipped without loss of argument. Fourteen chapters carry starred material; the full list is in the For Instructors section of the front matter.
9. **Part IV is modular by design.** Its hard prerequisites are only Chapters 2-4 and 6-7, which is what makes the second of the For Instructors section's two syllabus maps a thirteen-week course rather than a second year.

***

## Key Terms

* **Claim**: A contract entitling its holder to payments contingent on a state of the world. Every financial claim is simultaneously somebody's asset and somebody else's liability
* **Holder**: The party that owns a claim. In this book an analytical category rather than a descriptive one, because a holder's identity, size, and constraints are inputs to the price
* **Constraint**: Anything that makes a holder's demand for a claim depend on something other than the claim's risk and expected return — a mandate, a capital charge, a redemption term, a benchmark, an index rule, a tax status, or inertia
* **Holder section**: The section required in every chapter of Parts II, III and V asking who holds this claim and what their constraints do to its price. The structural device that makes the subtitle a method rather than a slogan
* **★ convention**: A star in a section heading marks a PhD-track section, in a problem marks a harder problem, and in a data exercise marks an extension requiring licensed data. The body of every chapter is complete without any of them
* **Stochastic discount factor (**$$m$$**)**: The random variable that prices every claim through $$p = E\[mx]$$. Introduced in Chapter 3 §3.5; the object every asset pricing model in Part II specifies and every holder in Part IV owns a version of

***

## Readings

There is no required reading for this chapter. What follows is the short list of papers this book returns to most often — the ten that appear in more than one chapter's required reading, or whose argument is used in more than one part. Each is listed with the chapter that teaches it.

* Markowitz, H. (1952). "Portfolio Selection." *Journal of Finance* 7(1): 77-91. — Chapter 4
* Sharpe, W. (1964). "Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk." *Journal of Finance* 19(3): 425-442. — Chapter 4
* Modigliani, F. and M. Miller (1958). "The Cost of Capital, Corporation Finance and the Theory of Investment." *American Economic Review* 48(3): 261-297. — Chapter 23
* Fama, E. (1970). "Efficient Capital Markets: A Review of Theory and Empirical Work." *Journal of Finance* 25(2): 383-417. — Chapter 7
* Black, F. and M. Scholes (1973). "The Pricing of Options and Corporate Liabilities." *Journal of Political Economy* 81(3): 637-654. — Chapter 8
* Grossman, S. and J. Stiglitz (1980). "On the Impossibility of Informationally Efficient Markets." *American Economic Review* 70(3): 393-408. — Chapter 7
* Mehra, R. and E. Prescott (1985). "The Equity Premium: A Puzzle." *Journal of Monetary Economics* 15(2): 145-161. — Chapter 5
* Fama, E. and K. French (1993). "Common Risk Factors in the Returns on Stocks and Bonds." *Journal of Financial Economics* 33(1): 3-56. — Chapter 6
* Shleifer, A. and R. Vishny (1997). "The Limits of Arbitrage." *Journal of Finance* 52(1): 35-55. — Chapter 15
* Koijen, R. S. J. and M. Yogo (2019). "A Demand System Approach to Asset Pricing." *Journal of Political Economy* 127(4): 1475-1515. — Chapter 20

Read in that order, the ten are the arc of §1.3 in primary sources. Each chapter's own required list is two papers long and is where the reading for a course should actually come from.

For the history behind that arc rather than the sources themselves, one book: Mark Rubinstein, *A History of the Theory of Investments: My Annotated Bibliography* (Wiley, 2006). It works through the field's results in the order they actually first appeared, and its corrections to the standard attributions — de Finetti before Markowitz on mean-variance, Treynor before Sharpe on the equilibrium model, Williams before Modigliani and Miller on the irrelevance of packaging — are a useful antidote to reading §1.3 as a sequence of unique discoveries by the people whose names are attached to them.

***

## Discussion Questions

1. **The multiplier in the oil pit.** Return to the opening episode after Chapter 20. Section 20.3 defines a price multiplier as the change in market value produced by a dollar of informationless flow, and reports estimates for the aggregate equity market. State what the analogous object is on April 20, 2020, and why it is not the same kind of quantity: identify what was inelastic, what the binding capacity constraint was, and over what horizon the elasticity was effectively zero. Then say whether a demand-system estimate of the kind Chapter 20 teaches could have been fitted to that market, and what data it would have required.
2. **Three commitments, one chapter.** Pick any chapter in Parts II-V and audit it against §1.1's three commitments. Where does it check theory against the institutional map, where does it name its sources, and where — if anywhere — does it assert something as settled that Chapter 20 §20.5's grading scheme would call contested? A chapter that passes all three trivially is probably a chapter you have read too quickly.
3. **Whose** $$m$$**, before the answer.** Before reading Part IV, write down your own answer to Chapter 3 §3.7's question in two sentences: whose marginal utility do you expect prices the US equity market, and what evidence would tell you whether you are right? Keep it. Compare it with §20.6 at the end of Part IV, and be specific about which part of your original answer the evidence changed.
4. **Which era are you in?** Take a market event from the past year that was widely discussed. Explain it three times: once in the language of §1.3's first era (a change in fundamentals or in required returns), once in the second (a behavioral bias or a limit to arbitrage), and once in the third (a change in who was holding the claim and under what constraint). Which explanation would be easiest to falsify with data you could obtain for free, and what does your answer suggest about why the third era arrived when it did?

***

## Data Exercise: Start with the Map

This chapter sets no exercise of its own. Chapter 2's — *Rebuild the Map* — is the standing exercise for the whole book, and it is worth starting before Chapter 2 is assigned. It runs entirely on the Federal Reserve's Financial Accounts (Z.1), which are free, quarterly, and downloadable as a CSV package, and it asks you to reconstruct the master holdings table of §2.3 row by row: who holds Treasury securities and how that has changed across two rounds of quantitative easing, what the household sector's balance sheet actually contains, and how much of American retirement wealth now sits in vehicles where the household bears the investment risk directly. Every later chapter that makes a claim about a holder is making a claim about a cell in that table, and the habit this book most wants a reader to acquire is checking the magnitude before accepting the mechanism.