View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

Final Copy 10/16/99 9:00AM

WHAT ARE THE IMPLICATIONS [of efficient financial markets]

Jerry L. Jordan
President and Chief Executive Officer
Federal Reserve Bank of Cleveland

8th Annual Financial Markets Conference
Federal Reserve Bank of Atlanta

The Cloister
Sea Island, Georgia

October 19,1999

My initial response to the title of this session is that it is not clear that efficient
financial markets have any implications for the regulatory system—other than to question
its existence. As I will point out, financial crises typically are made more serious because
government intrusions into financial markets lead participants astray before, during, and
after crises. For the most part, official programs seem designed to act as sponges for
absorbing risk exposures from particular groups of economic agents. Unless the resulting
incentive to overinvest in risky projects is offset by an effective program of supervision,
risk-free agents are likely to misallocate resources. Moreover, especially before a crisis,
a government may act as though the capacity of its risk-sponge is unlimited. Only when
that capacity is tested—by calls on foreign exchange reserves, by demands on taxpayers
through the budget—does information about limits emerge. Crisis ensues.
A general point I want to make today is that events which are often labeled
financial crises are inevitable. Both government intervention and market innovations can
influence the frequency and severity of these episodes, but they cannot be eliminated.
The best we can say is that evolution toward stronger political and economic institutions
is a discovery process and the sometimes dramatic financial market adjustments labeled
“crises” are an unavoidable part of that process.
In support of this general point, I will develop five propositions.
1. Financial crises are not predictable.
2. Risk mismanagement is at the root of financial crises.
3. Government guarantees are a common element in crises.
4. Failed market innovations may contribute to crises.
5. Moral hazard is a real and serious problem.
First, financial crises are not predictable.
If they were, actions would be taken which altered the event, and it wouldn’t be a
crisis. It is the surprise contained in new, unexpected information that sets off these
episodes. If there were no surprise, there would be no basis for the sudden and

substantial changes in market prices of financial instruments that are characteristic of
Financial outcomes always have an element of uncertainty as well as an element
of risk. I choose the words “uncertainty” and “risk” deliberately. They describe two
different kinds of exposure to chance events. One, the kind Frank Knight called
“uncertainty”, is impossible to describe probabilistically because the distribution of
possible outcomes essentially is unknown. Therefore, exposure to uncertainty cannot be
hedged or insured against. Gains and losses from uncertain events are pure economic
rents. They are either borne by those exposed to them, or socialized in an act of
community greed—such as a steeply progressive tax—or in an act of community
compassion—such as Red Cross, FEMA1, or Marshall Plan assistance. We cannot help
but be surprised by the occurrence of uncertain events. This is one reason that crises are
“Risk,” on the other hand, is the word Frank Knight used for exposure to the other
kind of outcome. This exposure can be described probabilistically, typically by using a
model of some sort to filter past experience. Exposure to losses from risky events can be
bought and sold in the marketplace. An insurance company will assume your exposure to
loss from the death of a partner in return for an insurance premium. Financial institutions
accept exposure to the bankruptcy of debtors in return for a sufficiently large risk
A corollary of the unpredictability of financial crises is that no two crises are the
same. Allowing borrowers and investors to suffer the losses incurred in a crisis is a
necessary and useful way to bring about adjustments in the technology of investment.
For example, adopting regulations now that would proscribe practices that appear to have
made the 1997 Asian crisis so severe should be unnecessary because borrowers and
lenders now understand what happened and will not be surprised again. If the parties
were to repeat the same behavior, markets would discount the now-expected result,
leading to substantial risk premiums in loan and currency markets. There would be no
crisis because there would be no surprise. More likely, however, borrowers and investors
1 Federal Emergency Management Agency.


would not repeat the behavior that brought them grief the last time. Instead, faced with
markets that extract more accurate premiums, they would adopt new practices—that is,
innovations—some of which would work and some of which would not.
The general point is that lessons learned in a crisis lead to changes in behavior
that prevent a repeat of the conditions that led to the crisis. The discipline exerted by
global financial markets is beneficial in that it erodes local resistance to more efficient
domestic markets. This is what the president of Korea had in mind when he said recently
that there is a “silver lining” to the Asian currency crisis. The restructuring and
reforming of the banking institutions now occurring in Asia will leave them better off. It
would have taken much longer to implement these reforms without the “crisis
atmosphere.” As a result, these nations may soon have better credit risk analysis, better
asset and liability management techniques, be less subject to politically connected bank
lending, and develop both internal audit and external supervision that is essential to sound
banking. 2
Second,financial crises typically resultfrom risk mismanagement.
The existence of risky situations is not a source of crises, for rational economic
agents will have incorporated into their decision-making calculations about potential
outcomes. Instead, crises come from new information that changes the calculations of
many agents simultaneously, forcing them to re-evaluate their positions and adopt new
strategies. The incidence of crises is not predictable because the occurrence of these
changing probabilities is uncertain. Nonetheless, I believe that the phenomenon of crises
is related, in part, to the way risk is managed, or, more precisely, mismanaged.
Risks can be redistributed across the world’s population, but cannot be reduced or
eliminated. Similar to a well-known principle of physics, risk cannot be eliminated, but

2 Or they may not. Suppose, instead, that these governments heed the advice of those who would throw
more sand in the gears of cross-border financing to discourage susceptibility to capital flight. A turnaround
tax on capital withdrawn within a few months or a year of entering a country is a popular suggestion. The
approach might work, just as capital controls seem to have worked for most nations during and after WWII,
but at the cost of a less efficient global allocation of capital. If regulation were to be part of the reform
process, it would make more sense to remove regulations that discourage long-term capital flows than to
add regulations that discourage short-term capital flows.


only transferred from one party to another. In the examples I used earlier, exposure to
loss from a partner's death can be shifted through an insurance company to a diversified
set of policyholders. Exposure to loss from lending can be shifted through a bank and the
FDIC to all insured depositors or general taxpayers. Exposure to exchange rate changes
can be shifted to someone else through the organized exchanges or an over-the-counter
transaction or by government to general taxpayers.
The important point is that while individual parties can take actions to reduce,
minimize, or avoid their own risk exposure, they can do so only when another party is
willing and able to bear such risks. This is where government behavior can become a
problem. Usually, with the most sincere and honorable of intentions, governments seek
to reduce, minimize, or eliminate the exposure to risk of some constituent.
A myriad of government intrusions in the marketplace—often in the form of
controls or guarantees—have altered the incentives of participants to accept risk
exposures in affected markets. The stated justification for regulation often is a contention
that a market failure or imperfection is present. It is important to understand that the
perceived failure or imperfection often was introduced by another government intrusion.
In that circumstance, then, regulation or supervision is an effort to recreate the
competitive situation that would have been present without the initial government
involvement in the market.
Banking supervision is a prime example of this effort. One of the most familiar
government intrusions into financial markets of this century has been a governmentsupplied guarantee called deposit insurance. Much of the regulation governing firms
covered by deposit insurance is rationalized by the need to neutralize the moral hazard
introduced by the insurance. The challenge always has been to construct government
regulations that do not undermine the effectiveness of market regulation—the discipline
that comes from having to compete.
A third, and broader point is that the prior presence of guarantees by
government has been a common element underlying many episodes labeled “crises.”
Any guarantee provided to anyone by some level of government alters behavior of
market participants. Most often, such guarantees are introduced as a well-intentioned


effort to shelter someone from exposure to the costs of some type of risk. But, those
costs and risks are real; as I mentioned earlier, they are not eliminated by the presence of
government-supplied guarantees. They merely have been shifted to someone else,
usually in a very diffused or opaque way.
When government offers guarantees, one of two possible problems is created.
Because market participants engage in behavior that is conditioned on the existence of
the guarantee, government has become a third contractual partner to any transaction or
agreement. In so doing, government has caused someone else—often the general
taxpayer—to bear a risk or cost, usually unknowingly. Then, when unforecastable events
force government to make good on the guarantee, general taxpayers belatedly are
informed that they must incur a wealth loss as a result of risk having been shifted to them
without their prior knowledge and agreement. As long as economic agents are unable to
fully internalize the potential costs of government promises, efficient financial markets
are not complete.
The alternative problem is that withdrawal of a guarantee has the effect of
breaking a contract, thereby imposing losses on someone. Abandoning an exchange rate
peg is a good example of this breach of contract. Many of the events that have come to
be labeled “financial crisis” or “banking crisis” involve the breaking of an explicit or
implicit contract, or the withdrawal by government of the offer to make new contracts—
that is, to provide guarantees—for the future. Because past behavior was influenced by
past guarantees, their withdrawal must be reflected in surprise adjustments of relative
prices. If these price changes are large and occur within a short time interval, the event is
labeled a crisis.
Exchange-rate crises of the early 1970s involved the withdrawal of the U. S.
guarantee that foreign governments could exchange dollars for gold at a ratio of 35 to 1.
Most exchange-rate crises since then have simply been reflections of unsustainable
government promises to redeem their own fiat currency for a foreign currency at pre­
announced rates.
A related aspect of crises concerns the quality of a nation’s legal and financial
infrastructure. The surprise that triggers a crisis is some new piece of information about


the inability of debtors to meet the terms of financial contracts. Contracts are drawn up
according to a nation’s rules and regulations.3 If those rules are lacking, unclear, or
capable of manipulation, then the opportunity for surprise is greater than if the rules are
clear and stable.
I don’t mean to imply that all crises are purely domestic, for there is little point in
trying to distinguish between an international crisis and a domestic crisis. In the context
of late Twentieth century global capital markets, all financial markets have an
international element. In fact, global financial asset portfolio managers can be thought of
as a new class of voters—stateless citizens of the world, empowered to roam the globe
casting votes for and against economic policies of the 200 or so nation-states. An
emerging common and ultimately hopeful element of "crises"—Mexico in '94-95, Asia
in '97, and maybe even some major industrialized countries in the future—is that various
types of government guarantees and promises—on deposits, on foreign investments, on
domestic pensions—may be on the "endangered species list" because so many are being
revealed as impossible to honor. That is what the current political turmoil in Germany is
all about. To the extent that global capital market vigilantes are becoming more effective
in evaluating the true costs of national governments’ promises, there should be less
opportunity for big surprises to bring on financial crises.
Fourth, crises may originate in failed market innovations.
The collapse of Long Term Capital Management last year raised questions about
the relationship between financial innovations and crises. Innovations are tests of new
ways of doing things. Some succeed, but, just as surely, others fail. Innovation is bound
to bring profits to some and losses to others. That’s the way the market system works.
Governments of emerging market economies want to realize very high growth rates by
introducing private enterprise. They don’t plan on losses on investments, but they will
have them. Unfortunately, government authorities in these countries often take actions—
3 An interesting exception has been proposed by Howell E. Jackson in “The Selective Incorporation of
Foreign Legal Systems to Promote Nepal as an International Financial Services Center” forthcoming in an
Oxford University Press symposium volume on regulatory reform. His suggestion is that firms be
permitted to establish operations in Nepal if they agree to abide by their home countries’ regulations and
submit their Nepalese operations to their home countries’ supervision.


with the best of intentions—designed to eliminate or minimize losses or risks to
investors, especially foreign investors. This distorts the evolution of their market
In the case of hedge funds, massive doses of computer technology are combined
with finance theory and on-line, real-time global information gathering to exploit
perceived anomalies in the prices of related financial instruments. This is a valuable role
to play in perfecting markets, and it can be a very profitable role if all goes according to
plan. However, as we have seen in a few cases, it can be a very unprofitable role if
market prices do not move toward the assumed past relationships. The resulting losses
can be multiplied manyfold to the extent such funds are leveraged. As long as there is
uncertainty about market relationships, surprises are capable of inflicting big losses—
perhaps big enough to be called crises.
Often, what gets missed in considering new vehicles for risk bearing is that they
are methods of redistributing risk. Far from increasing risk in the financial system, as
some have thought, they in fact serve the primary purpose of unbundling risk into its
components and, thus, allowing each type of risk to be managed separately. This allows
for specialization in risk bearing, as particular types of risk can be transferred from one
party, who is averse to that risk, to another party who is less averse to that risk. In this
way, financial innovations lower the cost of bearing risk by transferring it to those most
willing, and presumably able, to bear it. However, some financial innovations don’t
work and investors lose money. Rarely, the losses are big enough—Tulip mania comes
to mind—to warrant the label “crisis.”
Fifth, moral hazard is a real and serious problem.
“All that’s very well,” you will say, “but haven’t many market participants come
away from the 1997 Asian crisis with an entirely different lesson?” Namely, “Don’t
worry, ‘cause the IMF or someone else will pull our chestnuts out of the fire next time—
just as they did last time.” Not only are financial rescues likely to impede effective
reforms, but they may also induce back-sliding by encouraging an inattention to risk,
which will make the next crisis worse than the last one and thus induce more pernicious
moral hazard through even stronger rescue efforts. Notice that moral hazard involves the


intervention in the market by a government-type authority that relieves certain market
participants of a risk exposure by assuming that exposure itself—that is, on behalf of the
general taxpayer. Especially in cases where the risk transfer is only implicit, a big danger
is that the authority fails to create a supervisory mechanism to manage or control the risk
it has assumed. At least the current U.S. deposit insurance and lender-of-last-resort
institutions maintain an active supervisory presence in the banking system. Without a
supervisory presence, rescue efforts may promise relief in the short run, but will be
incapable of keeping that promise in the long run.
One approach to reducing the perceived need for government agencies to mount
rescue operations is to strengthen private market facilities for monitoring and dealing
constructively with potential crisis situations. I’ll mention just two examples of
proposals that have been getting attention. One would improve the transparency of the
financial positions of debtors through more consistent and thorough financial disclosure.
Creditors could then derive earlier warnings of trouble and force actions to limit losses.
Greater transparency is a desirable feature of efficient financial markets in its own right.
It is not clear that it would do much good in preventing the kinds of surprises that trigger
financial crises; however, it might aid in crisis resolution.
Another approach would include “collective action clauses” or other such
provisions in debt contracts that would facilitate action by private creditors to reschedule,
restate, or otherwise resolve situations in which debtors need relief. Especially with a
growing share of credit going directly to private firms rather than through sovereign
borrowers, creditors should be better equipped to protect their own interests without the
intervention of a government or international agency.
Of course, arming private investors will not provide an incentive for them to use
new provisions as long as government agencies appear willing to intervene and achieve
settlements on more favorable terms than produced by private negotiations. Given that a
“just say no” intervention policy is politically difficult, judgment must be exercised in
drawing the line between productive and unproductive intervention.
Critical to any successful reform is a reduction or elimination of uncertainty
arising from official responses to international market disruptions. Governments and

international agencies might clarify for markets what actions they would be prepared to
take, and under what circumstances. This could reduce incentives for creditor runs on
sovereign borrowers. Moreover, it would increase the accountability of policymakers for
their actions and, in turn, reduce the attractiveness of ex post creditor bailouts. This is by
no means as easy as it might sound. More fruitfully, perhaps, agencies might explore a
structured format for pre commitment, comparable to that made familiar in bank capital
In summary, financial crises are inevitable, I’m afraid. They are not predictable.
Typically, they arise from financial mismanagement by governments, sometimes in
response to failed financial innovations. Moral hazard compounds the inevitability of
crises by encouraging private inattention to risk exposures without any assured
supervisory offset.