The cheapest protection against disaster may be offered when nobody believes it is needed. Payments arrive, assets find buyers and borrowers renew their loans without difficulty. The record offers a succession of reassuring observations. Against that record, the person demanding a large margin for danger can look expensive, timid or simply mistaken. The question of the bubble begins here: whether a market can learn from its apparent safety in a way that makes it less safe, and whether the price paid for bearing risk can conceal the exposure being built beneath it.
A lower yield, by itself, proves very little about that exposure. Consider an imagined asset paying a certain five every year forever, beginning a year from now. With a constant discount rate of five percent, its present value is one hundred; with a rate of three percent, it is about one hundred and sixty-seven. Nothing in this example has become less dependable, and no resale fantasy is needed to explain the increase. For an uncertain asset, changed expectations about its payments can also alter its price. The observer who sees a rise has to distinguish these possibilities from a reduction in the compensation investors require for a given risk. Price is observable; the beliefs and alternatives making it acceptable are a harder object of inquiry.
The danger appears when the explanation for buying depends increasingly on the continued willingness of others to pay more. A building can provide rent, a business can produce earnings and a security can promise a payment; a buyer may nevertheless justify the purchase chiefly by the price expected from the next buyer. Resale is an ordinary and useful possibility. It becomes a fragile foundation when a payment coming due requires a favorable resale, and when many owners rely on that same exit at the same time. The question then concerns the whole chain of financing, including what each participant must do if the next transaction fails to occur.
Hyman Minsky organized this inquiry around the relation between cash receipts and commitments. His hedge-financing unit expects its cash flows to meet contractual payments. His speculative unit can meet interest from income but needs to refinance principal. His Ponzi unit needs additional borrowing or asset sales even to meet interest. These are financing positions, whose safety depends on what arrives and what must be paid. The term Ponzi in this classification does not by itself allege fraud, and the hedge position remains vulnerable if its expected receipts fail. What changes across the positions is the extent to which meeting today's obligations requires a further transaction.
Minsky's further claim is that prolonged prosperity can alter those positions. Financing that once seemed adventurous becomes familiar when it works repeatedly. A borrower learns that debts have been renewed; a lender sees collateral that has risen in value; an intermediary discovers a profitable way to make another advance. The commitments can grow around the conditions that validated the earlier commitments. In the hypothesis, stability helps induce a move toward arrangements that depend more heavily on continued favorable financing. The mechanism gives confidence an economic consequence: it changes the claims that have to be met, rather than merely changing how their holders feel.
Leverage makes the apparent record especially persuasive. Imagine an asset bought for one hundred using twenty of the buyer's own money and eighty of fixed debt. Ignore interest and other costs. If the asset rises to one hundred and ten, the owner's stake rises from twenty to thirty: a fifty-percent gain from a ten-percent increase in the asset. If it falls to ninety, the stake falls to ten: a fifty-percent loss. An owner who financed the same asset wholly from her own resources would gain or lose ten percent. The asset's history can be identical while the histories of its holders differ dramatically. A sequence of favorable observations can therefore display large rewards for the financial structure that will also magnify a reversal.
The reversal can change the structure as well as the valuation. A lender may reduce the amount it will advance against a given asset or decline to renew a loan. The owner then needs additional funds or a sale. If many owners face the same requirement and buyers capable of valuing the asset have limited funding too, selling can depress prices and weaken other balance sheets. Brunnermeier's account of the 2007–2008 crisis distinguishes this market-liquidity problem from the difficulty of obtaining funding and shows how they can reinforce each other. The price decline becomes an input into the next financing decision, which can provoke another sale.
This history also reveals why apparent diversification needs an account of the common exposure. Brunnermeier describes structured mortgage claims whose valuations benefited from optimistic assumptions about regional house-price relationships, alongside banks' reliance on short-term funding and liquidity backstops. Risks that looked dispersed could return through related assets and promises. That is a bounded explanation of the architecture of one crisis. It does not turn every new security into a deception or every favorable rating into proof of carelessness. It identifies a question to ask of an apparently quiet return: what common event could disturb several of its protections together?
A cheap asset after a crash need not be a safe asset. Its price may offer a greater expected return under a particular forecast while the buyer faces greater uncertainty about funding, counterparties or the forecast itself. Brunnermeier identifies several reasons lending conditions can tighten after prices fall: heightened uncertainty about future fluctuations, concern about the quality of collateral offered and estimation procedures that respond to recent losses. The danger has not necessarily passed with the exuberance. Some dangers have changed form, and others have been created by the attempt to escape the first ones.
Nor does a crash, observed afterward, prove that every earlier buyer misunderstood the odds. A bad outcome can occur under a well-specified distribution; a revised account of future payments can change a sensible valuation. Establishing mispricing requires more than pointing to the last price and declaring the earlier one foolish. It requires an account of the information, feasible alternatives and required compensation available then. The claim that prices always measure danger soberly is too strong for the mechanisms just examined. The claim that they are always mood gives up the harder work of explaining when information, financing and incentives have made them informative or misleading.
The volume can therefore ask a more useful question than whether the crowd is brave or mad. What supports the payments, how long can the holder wait, and which promises fail together? The answers may expose fragility even when recent losses are small. They may also show that an expensive asset has dependable earnings and financing that can withstand a difficult interval. The comparison disciplines both exuberance and the pleasure of denouncing it. A warning is valuable when it identifies an exposure that the reassuring record has left out; its value does not come merely from being darker than the prevailing story.
The calm can change the commitments built upon it, and those commitments can change what the next disturbance will do.
The bubble brings the volume's two meanings of risk into contact. There is the uncertainty people believe they are being paid to carry, and there is the vulnerability their way of carrying it creates. A price can reflect the first imperfectly while helping enlarge the second. When the claims finally have to be met, confidence alone cannot supply the resources. The closing plate turns to that reckoning: how the losses should be shared when neither private courage nor a well-priced contract can protect every person from the dangers of the common world. The argument remains open.