10.1A market's movement is not an act
A crash, a bubble or a window of prices is a market's own movement: prices move, and in the record of such a movement no authority suspends, closes, lends, guarantees or fixes anything. Where a record shows only such a movement, the ledger's own reading on whether there is an act is "no". This chapter holds three kinds of such record: the price-window group of the ledger, in which the field reads "no" in twenty-eight of the twenty-nine cases and "unsettled" in the remaining one; the case of Black Tuesday of 29 October 1929, in which the ledger reads "no"; and the case of the South Sea Company of 1720, in which the ledger reads "yes" for an act but reports that the two independent readings did not settle what kind of act it was.
Why a crash is not an act. On the study's own definition, a crash is a market's own movement, and no authority suspends, closes, lends, guarantees or fixes anything in the record of it. There is no decision to name, no promise of a term, no condition of an end written by anyone, no figure announced against what was available, no one who waited for the market to do something. The fields that carry the study's question — whether the act was presented as temporary, on what condition or by what date it would end, what became of it, who bore its cost while it lasted — therefore return "not-applicable" in these records rather than a reading of the case. The absence of an act is not a defect in the record, and the fields that return "not-applicable" are not a finding about the market; the study's question simply meets a case in which it has no purchase.
What the other axes can still read in such a record. The axis on who decides reads: no case in the ledger of eighty-four carries a decider in its own text, and the two independent readings settle on who decides in fifty-six of the eighty-four cases. The axis on what is missing reads: both cases in this chapter list, among their unknowns, unverified month-and-days, occurrence times, coordinates and personal information that must not be filled in as if they were established. The axis on whether a reading could be changed and on what condition reads: neither case in this chapter names a condition on which its reading would be revisited. The axis on the time of the record against the time of the judgment reads: both records carry a decision window that is a retrospective review, keeping the facts of the original apart from the reproduction check now carried out. The axes on what kind of act, on two dated moments, on a figure, and on who bore the cost also read, question by question, for these cases. Skeptron's records are its readings of public sources; their answers are its own and dated. The full set of axes is set out in Chapter 2, along with the studio's frame of recurring sentences, which appears in these records as in every other.
What this chapter can and cannot do with such records. It can report what the records hold and what they refuse; it can say which of the study's fields return a reading and which return "not-applicable"; it can set a crash, a bubble and a window of prices beside each other on the same fields. It cannot judge a market, give a forecast, or compute a return from anything these records contain. Where no act is present, there is nothing about the end of an act to find.
10.2Black Tuesday
Skeptron's record for this case is titled "Black Tuesday 1929 and the Coming Fall", set at 29 October 1929, a public-record question in the crisis-history domain, as of 13 September 2026, checked against Federal Reserve History's essay "Stock Market Crash of 1929" on 13 September 2026. The record's scenario describes a fall of about 13 per cent on the 28th of October and about 12 per cent on the 29th, warns that the sum of the returns over the two days is not called a cumulative return, and states that the stock-price crash, the subsequent low, the banking crisis and the economic contraction differ in timing and path.
The record's second question asks how Federal Reserve History's material distinguishes the index falls of 28 and 29 October. Its answer is the sentence this chapter is required to hold exact:
It is described as a fall of about 13% on the 28th and about 12% on the 29th. The sum of the returns over the two days is not called a cumulative return.
That answer is a record reading, not a calculation: Skeptron copies what its source says and refuses to turn two dated observations into a series. The third question asks whether 29 October 1929 can be substituted for the low of 1932 or treated as the single starting cause of the entire Great Depression. Its answer states a boundary:
The stock-price crash, the subsequent low, the banking crisis, and the economic contraction differ in timing and path. No daily bar without original source material is created and inserted into technical analysis.
The fifth question computes the interval between the earlier fall of 28 October, which the record calls the Black Monday fall of the Dow index, and Black Tuesday of the 29th, and asks what retrospective confusion can be prevented by recording that the two actions are distinct stages. Its answer holds both the arithmetic and the caution:
It is 1 day. The earlier Black Monday fall of the Dow index and the later Black Tuesday 1929 and the Coming Fall are not merged into a single action. Nominal assets, cash inflows, payment obligations, and policy instruments are kept separate. Temporal precedence alone does not prove a single cause.
The record's fourth question is a calendar check, and this study does not read it as evidence about the event.
On the ledger's own readings, this case holds no act: the reading on whether the record shows an act reads "no", and the readings on the kind of an act, on whether an act was presented as temporary, on the condition of its end and on what became of it all return "not-applicable" because no act is present. The reading on figures did not settle between the two independent readings. The field on who bore the cost is "not-applicable". Among the study's own axes, the decider is "unknown", the bearer of the wait is "unsettled", the horizon is "unknown", the constraint on reversal is "unsettled", the human's place is "unsettled", and no condition for revisiting the reading is named. The readings of the study's own hypotheses for this case are reported in Chapter 12; what matters here is that the record carries no condition on which its own reading would be revised, and no name of anyone who paid.
The reading note on Galbraith's The Great Crash 1929 (1955) is careful that the book is a narrative interpretation, not a statistical study. Where it bears on Skeptron's record: it supplies an account of the crash itself, of a kind the record does not hold. Skeptron's record holds two dated falls and a boundary; Galbraith's note gives an explanation of the boom and its break, and a judgment of the authorities' response, which is precisely the material the study's axes are not built to carry. Where it cannot bear: the note gives no dates for the banking crises or for the contraction, so it cannot check the record's refusal to merge them; and because the note describes the book as narrative rather than statistical, this study does not use it to settle what the record's figures mean beyond what the record itself says.
The record's refusal to merge the events is not a pedantic habit but the shape of a distinct literature, whose subject begins after the crash and whose dates are the panics rather than the two days in October. Where it cannot bear on this case: the reading note on Friedman and Schwartz's A Monetary History of the United States, 1867–1960 (1963) says nothing about the two days of October 1929 themselves, and this study attributes to those authors nothing the note does not give. The record and the work do not contradict each other; they are answers to different questions.
What a later reader may and may not do with this case. A reader may report the two dated falls as the record gives them, each with its own date and percentage. A reader may not sum them into a return, may not substitute 29 October for the low or for the banking crises, may not build a daily price series from the chronology, and may not say what followed on a date the record does not give. The record's own verification statement says that the action, subject and scope stated in the record are verified up to what is written there. That boundary is the whole substance of the case: a crash dated in a public record, kept apart from the events that followed it, with nothing in it that a later reader could code as a decision, a term or a bearer.
10.3The South Sea year read for its market
Chapter 6 reads the case of the South Sea Company for who bore the cost while the year's events ran. This section reads the same case for its market, on what a market reading can hold and what it refuses. The record is titled "South Sea Company where national debt absorption and a stock boom met", set at 1720, in the crisis-history domain, as of 13 September 2026, checked against the Bank of England chronology of its founding, the South Sea Company and convertibility restrictions on 13 September 2026.
The record's second question asks what public debt transaction the chronology connects the 1720 rise in the company's stock price to. Its answer:
After the South Sea Company obtained the authority to absorb part of the national debt, the stock price rose sharply and subsequently collapsed. An explanation is also left in place that the company received a trade privilege, but actual trade was extremely limited.
The record's third question asks whether the existence of a trade privilege may be read as sufficient operating revenue or as the persistence of a stock price rise. Its answer sets the pieces apart:
The patent, the debt absorption, the trading revenue, and the stock price must be verified separately. The chronology is not real daily price data, so price anchors and returns must not be invented.
On the ledger's own readings, this case reads "yes" for an act, yet the two independent readings did not settle the kind. What the record itself holds carries the point: the company obtained the authority to absorb part of the national debt, and the record gives no term for that authority, no condition or date for its end, and no name of anyone who bore its cost while it stood. Who decided, and on what horizon the act ran, the record leaves undecided.
Charles Mackay's chapter on the bubble, in Memoirs of Extraordinary Popular Delusions (1841), is the period account of this year that this study's materials carry. The transcription of the chapter available to this study was not checked against a page image, so this study describes it in its own words rather than quoting it.
The excerpt available to this study ends its account of that year with Exchange Alley in crowds and a ballad of the time, though the chapter continues beyond it. Every sum, rate and date in it is Mackay's, in the units he gives; the record before this study carries none of them, which is why the record forbids inventing price anchors and returns.
Kindleberger and Aliber's Manias, Panics and Crashes (2005), by its reading note, is a survey of financial crises over several centuries, built in its fifth edition on Minsky's model of displacement, boom, euphoria, distress and revulsion. Where it bears: the stages give a shape in which the company's authority over part of the national debt, its trading revenue and its stock price are three different things in time — which is exactly what the record's answer asks a reader to keep apart. Where it cannot bear: the note's argument turns on a lender of last resort appearing at the right moment, and this record names no lender, no rescue and no announcement of one. A mania without a rescuer is not a test of the note's demand; it is a case the demand does not reach.
This study reads the record's insistence that actual trade was extremely limited while authority to absorb part of the national debt was granted as the kind of separation Minsky's Stabilizing an Unstable Economy (1986) turns on — claims made against an entity whose income does not support them. The reading stops where the evidence stops: the record gives no rate, no income, no balance sheet and no date beyond the year 1720, so it cannot show which of Minsky's categories, if any, the company's financing fell into.
Garber's Famous First Bubbles (2000) is used here only through its reading note. Where it bears: the record's own boundary — patent, debt absorption, trading revenue and stock price must be verified separately — asks the same question from the archive's side: what part of a company's position is accounted for by its actual trade, and what part by its authority and its stock? Where it cannot bear: Garber's note does not decide what the Bank of England chronology states, and the chronology in the record holds no daily prices, which is why the record forbids inventing them.
The record's fifth answer names what a comparison of this case with the present would have to observe: liquidity by maturity, borrowing entities, and records of actual performance; and it says that if expected inflows are counted as available cash, or the support of other entities is combined, the plan must be recalculated. The record's own final sentence adds that the connection it makes is a comparative hypothesis, not a prophecy of the same future outcome as the past. That is the closest this case comes to naming a condition under which its reading would change — and even there, the condition is a test for a comparison the record proposes, not a term under which any act of 1720 would end.
10.4Skeptron's price windows
The reason these windows sit in their own group, apart from the crisis, credit and money histories, is set out in Chapter 2, alongside the study's own fields and hypotheses.
What the ledger reads in the group. Skeptron's price windows record observed prices up to one closing price; they hold no decision and no actor, and this study reads them only as a group. The group's readings, as "The ledger's numbers" gives them:
| What is read | Price windows (29 cases) |
|---|---|
| act | no 28; unsettled 1 |
| kind of act | not-applicable 28; unsettled 1 |
| presented as temporary | not-applicable 28; unsettled 1 |
| condition of its end named | not-applicable 28; unsettled 1 |
| outcome of the term | not-applicable 28; unsettled 1 |
| two dated moments named | not-named 25; unsettled 3; named 1 |
| figure | single-figure 20; unsettled 9 |
| bearer named | not-applicable 26; unsettled 3 |
Skeptron's price windows are twenty-nine cases with no act, recording observed prices up to one closing price, with no decision and no actor, and this study reads them only as a group.
Read across its questions, the group shows what the study's question does when the object it expects is missing. The four fields that carry the series' question — whether an act was presented as temporary, whether a condition of its end is named, what became of it, and who bore its cost while it lasted — return "not-applicable" in nearly every window, because a window of prices has no act to be temporary, to end or to be borne.
Across the twenty-nine windows none of the study's own hypotheses finds support, because none of them can be asked of a window that holds no act. The stricter of the two readings counts only what a record's own words carry; the more generous counts what the setting implies as well. On the first of the study's own hypotheses, ten windows are unclear on the stricter reading and one on the more generous, and none goes for or against it. On the second, one window is unclear on each reading and none goes for or against it. On the third, ten windows are unclear on the stricter reading and one on the more generous, and none goes for or against it. On the fifth, the counts are the same as on the first. The one verdict in this group is on the fourth: sixteen windows go against it on the stricter reading and twenty on the more generous, thirteen and nine are unclear, and none goes for it, a share of 0% on both readings. The counts for the whole ledger and the study's verdicts on its hypotheses belong to Chapter 12; what this group contributes is its own column.
Why the windows are kept apart from acts. The study's question is about what a record owes a later reader about an act's end. A price window has no act, so folding its readings into the counts on temporality, conditions of ending, outcomes and bearers would read "not-applicable" as a silence in the record when it is the study's question meeting no actor. There is a second reason, which runs the other way: the windows are the study's evidence about markets, and a price is the one object in these materials that the study never forecasts, never converts into a claim, and never extends past the closing price a record holds.
What a reader may and may not compute from these cases. A reader may read what the ledger's two independent readings say about the group, question by question and hypothesis by hypothesis, as this chapter has done. A reader may report that each window's record ends on one closing price taken after ninety days of observation, and that the ledger reads twenty of the group's cases as carrying that price as a single figure. A reader may not compute a return across the ninety days, a price at any other date, or a trend through the window; may not build a daily series from a chronology, since the Black Tuesday record states that no daily bar is created and inserted into technical analysis without original source material; and may not convert a window's closing price into a claim about an act, because no act is present.
What this group lends to the series' question, and what it does not. It lends a clean contrast: a record can be complete in its own terms — ninety days, one asset, one closing price, a dated observation — and still hold no decider, no act, no bearer and no end to name. It also shows that "not-applicable" is not "absent": the ledger's verdicts name the difference between a case that goes against a hypothesis, a case that supports it, a case that is unclear, and a case where the hypothesis does not apply. What the group cannot lend is anything about an emergency act's term, its condition of ending or its cost: with no act, there is no term to check and no one who waited while an act stood.
10.5Instruments that observe
The note on Donald MacKenzie's An Engine, Not a Camera: How Financial Models Shape Markets (2006) says that MacKenzie traces the development and use of option pricing theory, including the Black-Scholes-Merton model, and how it changed trading practice; that he also discusses the crash of October 1987 and the collapse of Long-Term Capital Management, where models and their wide use affected outcomes; and that the book is a study of how theories enter into practice.
Where it bears on this chapter. This study reads that argument as extending to records of prices read through instruments, in which what the instruments show shapes what the market does. The same caution applies to this study. The ledger's readings are not copies of the cases either; they are made with categories, and the readings that did not settle mark the places where the instrument returns no single answer. On the price windows, nine of the twenty-nine cases did not settle on the figure and three on each of two other questions, one on each of the five questions that ask about an act; on the South Sea case the two independent readings did not settle what kind of act the grant of authority was, and the ledger's numbers give neither the values they read nor a basis for that disagreement. Chapter 2 sets out the two independent readings and the two levels at which a count may be taken; what this chapter adds is their appearance in cases with no act at all.
Where it cannot bear. MacKenzie's note speaks of equity and derivatives models of the twentieth century and of two episodes in which models and their wide use affected outcomes. It says nothing of crypto-assets, of Skeptron's records or of this ledger. This study uses it to say what an instrument is — a thing that reads and in reading shapes — and not to explain any price window. The fit is partial, and saying so is part of using the work honestly.
What a record of observation without a decider lends to the series' question.
First, a boundary on speech. The Black Tuesday record states that the sum of two daily falls is not called a cumulative return, and that no daily bar is created and inserted into technical analysis without original source material. The South Sea record states that price anchors and returns must not be invented because the chronology is not real daily price data. These are statements about what a later reader may say of a market — the same shape as the study's question, which asks what a record owes a reader about an act. Where an act is absent, the record still owes the reader its limits.
Second, a contrast with the act-bearing cases. Where the study's field reads "no", the fields that carry the series' question do not fill with guesses: they return "not-applicable", though the other questions still give a reading from the record — Black Tuesday's "dates apart" is "not-named" and its figure reading did not settle, and the price windows carry single figures with their dates usually not named — and the ledger keeps the count so that the group's column can stand beside the columns of the crisis, credit and money histories without pretending the questions were asked of an actor.
Third, a check on method. The frame's recurring sentences appear in these records too — including the sentence that announcement, decision, implementation and results must not be merged into a single event — and the frame is Skeptron's method, set out in Chapter 2, not a finding about the cases. Its presence in a record with no act shows the method at work in every case, not a property of markets. This chapter counts no frame sentence as a finding and no link between cases as a pattern.
Fourth, a limit. Neither case in this chapter names a condition on which its reading would be revisited, and the question of who bore the cost reads "not-applicable" for one of them and "unsettled" for the other. From a record with no act at all, this study cannot read anything about who waited, what an act cost, or what would change an emergency act; and the South Sea record, which the ledger reads as showing an act, gives no term for it, no name of anyone who bore its cost and no settled answer on who decided. The study's verdicts on its hypotheses are Chapter 12's; the condition of an end, read across all the cases, is Chapter 11's; and the question of what evidence an act's success would require at all is set out in Chapter 1.
What such a record cannot lend. It cannot lend the term of an act, the condition of its end, a horizon, a bearer or a decider. It cannot show what any act promised, who paid while it lasted, or whether it was undone. It cannot say whether an emergency act worked, because that question needs acts, and the study has never claimed that its records hold what a judgment of success would need. What it can lend is sharper than it first appears: an instrument for keeping the study's own question honest. The series asks of emergency acts what a record must contain for a later reader to judge their end. The crashes, the bubbles and the windows are the cases where the question returns "not applicable", and those verdicts are part of the count, not an exception to it.