Financial Crisis and the Paradox of Under- and Over-Regulation Joshua Aizenman UC Santa Cruz & NBER SER Conference Singapore August 7 2009.

Financial Crisis and the Paradox of Under- and

Over-RegulationJoshua Aizenman

UC Santa Cruz & NBER

SER Conference

Singapore

August 7 2009

2

Paper’s agenda:Endogenous regulation cycles

Explain the tendency to under regulate in “good times,”

Explain the risk associated with overshooting the adjustment called for following a financial crisis.

Outline a regulatory structure that mitigates the above concerns.

3

Financial integration in the last 30 years, developing countries

The 1980s - “lost development decade.”

The 1990s - growing optimism about the globalization gains, trade liberalization, coupled with financial opening in LATAM (to a lesser degree in Asia).

4

5

Vigorous debate about financial opening.

Financial opening: Beneficial when the only distortion is restricting

intertemporal trade across countries, Potentially damaging in the presence of other

distortions. “Other distortions” -- moral hazard (MH), an

implicit subsidy to borrowing and investment. MH: the “too big to fail” doctrine –

allowing large borrowers to go under triggers systemic crisis

6

Financial innovations and financial deregulation in the OECD countries

Presumption – OECD’s superior financial intermediation

financial innovations are welfare improving. The spell of “the great moderation” further

reduced concerns about the downside risk. Concerns about principle-agent/ MH associated

with financial intermediation were swept aside. “As we move into a new century, the market-stabilizing

private regulatory forces should gradually displace many cumbersome, increasingly ineffective government structures.” Alan Greenspan, FED Chair, April 12, 1997.

7

The unfolding global liquidity crisis

Illustrates the universality of MH and the “too big to fail” doctrine.

Under regulating financial intermediation excessive risk taking, subsidized by tax payers’ financed bailing outs.

The magnitude of the global crisis: a surprise. IMF 4-09 estimate of global toxic assets ~ $4 trill. The high costs of the crisis the pendulum of

financial integration shifts towards reversal.

8

High persistency of financial policies.

The high co-movement of financial integration indexes across countries, and the persistence of policies towards financial openness suggest pendulum dynamics.

9

Under-regulation: possible interpretation

Prudential regulator’s success or a prolonged economic tranquillity leads to complacency lower demand for regulator’s services under regulation … financial calamity. The identity of agents that benefit from crisis

avoidance is unknown; yet regulation costs are transparent.

Crises that have been avoided are imperceptible, and are under-represented in the political discourse.

The demand for regulation declines during prolonged good times, thereby increasing the ultimate cost of eventual crises.

10

Under-regulation: the modelInteraction of two types of uncertainty

Financial reform takes place at time 0. Financial reform may lead to a future crisis. I. The crisis probability is adjusted overtime in a

Bayesian manner. Costly regulation () reduces the Probability of

financial crisis; higher lower Pr. II. Individual uncertainty: The majority is

exposed to a costly crisis with probability q < 1. The minority is exposed to the crisis with probability 1.

The majority determines the regulation.

11

Under-regulation: the modelResults

Show conditions where the regulation level supported by the majority is positive after the reform, but below the socially optimal level.

Tranquil time reduces the regulation intensity, as it reduces the posterior probability of a crisis.

Long spell of no crisis the regulation level drops towards zero; while the socially optimal regulation level remains positive.

12

Under-regulation: the modelResults, cont.

Asymmetric information aggravates the challenges facing the regulator.

Example: the public does not observe regulator’s effort.

Higher regulator effort, helping to avoid a crisis would be confused as a signal of a safer environment reducing the posterior probability of a future crisis eroding the support for costly future regulation.

13

Prudential over-regulation

Crisis resulting in unanticipated high costs may induce over-regulation and stagnation:

Parties that would bear the cost of the over regulation are under-represented in the decision making process [can be modeled applying Fernandez and Rodrik’s AER (1991) status quo bias].

14

A regulatory structure thatmitigates these concerns

Improving confidential information disclosure: needed to allow assessment of systemic risk triggered by “too big to fail” concerns.

Increasing the independence of the regulatory agency from the political process; needed to reduce under-regulating in good times.

Centralizing the regulatory process and increasing its transparency; mitigates asymmetric information problems.

Global standards of minimum prudential regulation and information disclosure, enforced by the domestic regulator; a commitment devise mitigating under-regulation in good times, reducing regulatory arbitrage between countries [see AIG…].

15

The probability of a crisis following financial reform at t = 0

The prior probability of crisis at t = 1 is .

Costly investment in regulation at time zero, per agent reduces the probability to ; ;

To fix ideas,

0

0,1P

0,1P Q (0) 1; 1Q Q 0( ); ' 0Q Q Q

00

0

2exp( )

1 expQ

16

The time line

No crisis: (see Appendix). 0,1 1,2P P

17

Two groups, FE (fully exposed) and PE (partially exposed)

18

Claims 1 & 2With positive regulation, the planner’s

regulation level is above the optimal level set by the partially exposed (PE) group [the majority].

In an equilibrium with positive regulation; the regulation rate increases the higher the perceived crisis probability, P; the higher are crisis costs, the higher is the effectiveness of regulation.

These factors increase the likelihood of positive regulation.

19

A simulation: regulation level () and the expected marginal benefit of

regulation, at time t = 0 [S: planner’s curve, PE: majority’s curve]

S

PE

20

Claim 3: good luck spell High efficacy of regulation

the regulation set by the majority [PE group] in time t = 0 is positive, but below the socially optimal level. The under-regulation drops with the share of exposed agents, q.

Tranquil time reduces the regulation intensity. A long spell of no crisis regulation level drops towards zero;

yet the socially optimal regulation level remains positive.

21

S

PE

A simulation: regulation level () and the expected marginal benefit of regulation, a long spell of “good

time” [S: planner’s curve, PE: majority’s curve]

22

Claim 4: less informative prior[assuming a simple discrete binomial pdf]

Less informative prior regarding the probability of a crisis the faster would be the drop in regulations induced by a “good luck” run.

Thus, good luck runs are especially damaging in the aftermath of unprecedented financial reform, when the public is exposed for the first time to new financial instruments.

23

Claim 5: asymmetric information

The public gets noisy signals about the regulator's effort Higher effort which avoids a crisis would be confused as a signal of a safer environment reducing the posterior probability of the crisis, below the level of symmetric information.

Would erode the future support for costly regulation. May increase the ultimate crisis cost.

24

Over-regulation hazardA crisis that leads to a cost of a higher

order of magnitude than the anticipated one, may induce a pendulum shift from under-regulation to over-regulation.

25

Costly over regulation

With repressed financial intermediation entrepreneurs that would have benefited from financial intermediation, are under-represented in the decision making.

This happens with individual-uncertainty regarding the incidences of successful investment [analogues to Fernandez and Rodrik (1991)].

Crisis outcome: increasing the risk of Over-regulation a static economy; the benefit of crisis avoidance would come with a large cost of stagnation, a cost that is under-represented in the political discourse.

26

Regulatory changes needed to deal with the challenges associated with under and

over-regulation

Information gathering: a necessary condition for effective regulation is mandatory periodic confidential reports of the balance sheet exposure of all financial institutions above a minimum size, operating in the domestic market. In the US, the BEA collects detailed

confidential info. about the real sector, but no comparable disclosure of the financial system…

27

Independence of the regulatory agency from the political process and various

pressure groups

Due to principle-agent problems, the regulator’s independence is needed to avoid “regulatory capture.” Interested parties prefer under-

regulation as a way to facilitate excessive risk taking subsidized by the tax payers. Rajan and Zingals (2003) and Rajan (2005).

28

Centralizing the regulatory process

A fractured regulatory process each agency focuses on its narrowest task, viewing “ the big picture” as beyond its mandate.

Evaluating systemic risk requires combining all the pieces of the financial puzzle.

Centralized regulatory process minimizes regulatory arbitrage.

29

Adopting global standards of minimum prudential regulation and information disclosure, enforced by the domestic

regulator.

Global minimum standards increase the costs of deregulation -- a commitment devise.

Minimum prudential standards mitigates “regulatory arbitrage” across countries:

Under-regulation attracts capital inflows in search of higher returns induced by the implicit subsidy provided in more under-regulated countries.

AIG ‘s under-regulation allowed it to sell under-priced insurance contracts to European institutions, arrangements that were subsidized by US tax payers.

30

The models results are robust to various extensions, including

Lobbying model instead of a ‘median voter.’ Non Bayesian agents, as may be the case if

people pay insufficient attention to low probability risks of disaster before a crisis happens, and too much attention right after. See Viscusi (2009) and Tversky and Kahneman.

Continuous distribution of agents with heterogeneous income and exposure, where the median voter determines the regulation.

31

Concluding remarks

The challenge: to adopt Goldilocks regulations, mitigating the temptation to under-regulate in spells of good time, preventing over-regulation in the aftermath of a financial crisis.

The risk of not meeting these challenges: Some affected countries will opt to reduce their

financial integration; Some will overshoot the regulatory adjustment

inducing lower future growth; Others will remain exposed to the hazard of

replaying crisis dynamics in the future.

Thank you!