4.1The claim
Chapter VII of Lombard Street describes a system with one banking reserve. The other banks kept what they called reserves, but on Bagehot's account those were deposits at the Bank of England, which the Bank lent out as it lent any deposit. When a panic came, the pressure of every kind of borrower fell on the same store. The measure of that store was a ratio of reserve to liabilities, and the chapter shows the people who spoke for the Bank giving a ratio without agreeing on what it meant.
The shared-measure danger that Daníelsson finds in modern risk models (Daníelsson 2002) has an older relative in Bagehot: the ratio by which the single holder judged its own reserve. This study does not equate the two. In Bagehot the sameness of the many holders is a fact of the system, not a product of their measure; the ratio shows how little the one holder could say about what its own figure meant.
4.2The only large store of unused cash
Bagehot reprints an Economist article written about a speech by the Bank's Governor. The article claims that the Bank of England keeps the sole banking reserve of the country, the only considerable store of unused cash in it. The claim concerns the reserve held against deposits, not the one held against notes; the article sets aside the question of who should issue notes.
The article adds that there had been no universal agreement on this. Great authorities had been unwilling to admit it, but had not formally argued against it, because they would have had to name some other great store of unused cash, and they could not. The article reads the Governor's speech as saying that the other banks need not keep a reserve of actual cash of their own to help them through a sudden panic, and as acknowledging a duty on the Bank to support the banking community. It welcomes that plainness, and asks that the country know exactly who is to keep the banking reserve.
For this study the point is the place of the store. The cash is unused, and it is kept in one place.
4.3Reserves that were already lent out
Bagehot reports the reply of Mr. Hankey, whom he describes as one of the most experienced directors of the Bank. Hankey called the Economist's doctrine the most mischievous ever put forward in banking in the country. As he stated it, the doctrine was that the Bank should keep money available at all times for bankers who had rendered their own assets unavailable. Hankey held that the Bank should keep its banking deposits in the most available securities, with about a third of them in cash, and in a sudden pressure bear its full share of a drain. He thought it impracticable, and if practicable inexpedient, for the Bank to keep money unemployed for an emergency.
Hankey's bankers whose assets have become unavailable are, in other words, the gap that the second record names for one trading account: assets owned, but not usable when needed. This study makes that link, not Hankey.
Bagehot answers in four points. First, the question is what exists, not what ought to exist: the Economist had said only that a single reserve was the system there was. Second, Hankey should have shown some other store of unused cash from which advances in a panic could be made. The reserves of the London bankers are not such a store. They are cash that is used, not unused: part of the Bank's deposits, and lent as such. Third, the joint stock banks of London did not keep a third of their liabilities in cash, and the Bank of England kept a different kind of reserve from the others. If the two were to be managed on the same principle, Hankey should have said which was to move towards which. Fourth, in most panics the largest advances from a banking reserve go not to bankers but to the mercantile public and to bill-brokers, and by the country's system all extra pressure falls on the Bank. In the worst part of the latest panic he reports, fresh money could not be borrowed, even on the best security and even on Consols, except at the Bank. No other lender stood ready for new borrowers.
In the passages this study read, the third point is the nearest Chapter VII comes to the subject of a shared measure, and it points against one: the Bank and the joint stock banks hold different kinds of reserve. Bagehot makes it in passing, and this study does not build on it.
4.4A ratio with no stated meaning
How large should the one reserve be? Bagehot says that the Bank had no adequate rule recognised for managing its banking reserve. Mr. Weguelin, the most recent Governor to be examined, had said that it was enough to keep between a quarter and a third of banking liabilities as a reserve. Bagehot writes that no one would now be content with a reserve near a quarter. Hankey named about a third as the proportion to aim at, but did not say whether a third was a minimum below which the reserve should never fall, or an average about which it might rise and fall.
On this reading the measure of the one reserve was a ratio from ordinary banking whose meaning in a panic was left undefined. The ratio was stated by individuals, a Governor and a director, and the chapter says only that no rule was recognised. It does not say that a board had adopted the ratio as a rule.
4.5The board and its received rules of thumb
Why did the rule stay vague? Bagehot looks to how the Bank was governed. He describes the directors as a board of plain, sensible, prosperous English merchants, from whom no one could expect great knowledge of economic science or original views on banking. A board can scarcely ever make improvements, he says, because its policy is set by the opinions of its average members, who are never ready for sudden improvements. Such a board acts on what it thinks safe, that is, on the accepted rules of the mercantile world of its time, and so, Bagehot writes, the directors of the Bank had acted nearly uniformly.
He gives an earlier example of the habit at work, and excuses the directors of that time: they spoke according to the best mercantile opinion of England, and the City and the House of Commons approved. They believed that notes issued only at a fixed rate, and only on the discount of good bills, could not be depreciated. The checks worked for a time, because good bills grew slowly. Bagehot says that they failed in time, and that the theory behind them was nonsense, though for a time their operation was powerful and excellent.
This study reads the passage as the nearest thing, in the passages it read, to a measure held in common. It is a rule of thumb held in common not across many banks but across the one board that held the one reserve. Bagehot does not call it a measure, and a single board is not a population of banks. The reading shows only this: a rule shared by those who decide can work while its conditions hold, and then fail.
4.6The record: one measure for all banks
Synthetic scene: the record below is a synthetic setting that WSTiger publishes for study, a modern platform summary with no outcome, and it reports no real supervisor or bank.
WSTiger dates the record "1996 · synthetic setting". Its scenario:
A hypothetical bank supervisory authority recommends that large banks set trading limits using the same statistical risk measure. When volatility spikes, the measured values of all banks rise at the same time and the limits are breached.
The question it sets:
Should the recommended measure be used as it is, should in-house stress scenarios be reviewed alongside it, and should responses to limit breaches be set differently for each bank?
The three options, as the record lists them:
Use the recommended measure as it is Review in-house stress scenarios in parallel Differentiate responses to breaches
WSTiger's example judgment, a dated reading, not a finding:
This example separates the measurement accuracy of individual banks from the risk that all banks react to the same signal. It estimates the scale of market sales that would emerge when limit breaches occur simultaneously.
The thread the record belongs to, "When everyone uses the same model", draws this insight:
When everyone uses the same model and data, decisions that are rational for each party move in the same direction at the same moment and become a risk to the market as a whole.
The scenario puts the uniformity in the measure, not in a panic: values rise "at the same time" because the measure is the same. Beside the example judgment this study sets Bagehot's report from 4.3: in the worst part of the latest panic he reports, no one but the Bank lent to new borrowers. The record estimates how much would be sold, and says nothing about who would buy. Chapter 5 returns to this.
4.7What the record adds and where it stops
The record puts, in the words of a decision, a danger of the kind Daníelsson describes: a measure acted on by many firms changes what it measures (Daníelsson 2002; MacKenzie 2006). The record does not cite either author, and the link is this study's. One option, to differentiate responses to breaches, would put variety back at the level of the response.
Yet none of its options names a holder apart. The options are about the measure, a parallel review, and the responses to breaches. The record does not say who decides among them, and it does not ask, as Bagehot did, who keeps the store that everyone turns to.
The case's contribution: Bagehot's pressure on one reserve came from the system, while the record's sameness comes from the measure, before any panic, and the record names no holder who would stand against it.