Public Enforcement of Securities Law: A comparative perspective

Date June 13, 2007
Speaker Howell E. JACKSON(James S. Reid, Jr. Professor of Law, Harvard Law School)
Commentator & Moderator SHISHIDO Zenichi(Faculty Fellow, RIETI / Professor, Seikei Law School, Seikei University)
Materials

Summary

About eight years ago I visited South Korea on behalf of the International Monetary Fund (IMF) to inspect the progress by the Financial Services Authority (FSA) in the wake of the East Asian financial crisis. At that time I gave a lecture in which I was asked how many financial regulators South Korea should hire for effective supervision of the financial services market. I had never considered this question before, so my response at the time was theoretically correct but practically unhelpful: when costs exceeded benefits then hiring should stop. On my return to Harvard I looked into this question further and realized its true difficulty.

Firstly, there is the problem of measurement: how should the costs of hiring staff be calculated? Every time a regulator is hired this puts pressure on the financial services industry, as they then have to hire people to deal with the regulator. Finding out these private costs is very difficult in itself. The real question is, "What is the incremental staff being generated by each regulator?" In addition there are adjudication costs, as is the case in the United States with litigation and enforcement - judges, lawyers, etc. - costs of disrupted transactions, as well as costs in the form of sanctions. Hence, the costs are difficult to work out.

Equally, however, the quantitative benefits of regulation are also difficult to work out, while we are told that regulators "qualitatively" protect consumers and eliminate systemic risks, etc. In the U.S. there are additional soft regulation policies, like the policing of money laundering, which are also difficult to quantify.

The underlying implication of the question I was asked in South Korea was, therefore, whether information gained through other countries' experiences (specifically, those of the U.S.) of financial regulation could be borrowed and applied to South Korea. This is also a tricky question.

There are a number of problems with this kind of comparative analysis. To begin with there are the differences in scale - of population, market size, economy, and so on; then there are the differences in the financial sectors, which would have some bearing on what the regulatory structure should look like; the differences in regulatory objectives - the U.S. has a federal system with both federal and state regulators, for example; the differences in the wealth of a country; the differences in expertise in economic and legal analysis; differences in private remedies and responsibilities; and finally, differences in compliance levels.

Therefore, I acknowledge at the outset that it is difficult to know what the optimal level of financial regulation is in any jurisdiction, and that even if there were a jurisdiction which was seen to have an optimal level, it is not clear that that could be immediately translated to other countries.

From my subsequent investigations into regulation, I estimated the total number of regulators in the U.S. to be around 45,000, at a cost of approximately $5.5 billion a year. In contrast, the FSA in Tokyo has around 2,000 people. There are then a number of ways in which to normalize these figures for scale differences. Firstly, if we normalize the number of financial regulators, the U.S. then has about 100 regulators per million, which is about five times the figure for Germany or France. It is interesting to note that there is a lot of variation across countries when looking at the normalized numbers of regulators. When looking at securities regulation costs per billion dollars of stock market capitalization, the U.S. and UK are fairly comparable but still significantly higher than either France or Germany. It is interesting to note that the amounts spent on regulators in Australia and Canada are high compared with their stock market capitalization.

When looking at the regulatory costs per billion dollars of gross domestic product (GDP) in civil law countries versus common law countries we note, firstly, that South Korea has a high level of staffing and budget for a civil law country. Japan has a much larger economy but is a relatively low-intensity jurisdiction when measured in this way.

There is a lot of interest in the question of what kind of legal system is best for securities markets. The common law systems are associated with bigger capital markets and people say this is because the civil law systems are too rigid, lacking in fiduciary duties and openness to market innovations. Contrary to prior understanding, my research has shown that common law countries actually hire many more regulators and governmental staff than the civil law countries. This means that if civil law is getting in the way of capital market developments it is not because they are hiring too many regulators; perhaps there are other aspects of the legal regime that are problematic or perhaps they in fact do not have enough regulators.

The paper I am currently working on with Prof. Mark Roe (David Berg Professor of Law, Harvard Law School) attempts to move from the description of public enforcement of budgets into a more normative question about positive and negative aspects. The motivation for this paper comes from the announcement from the World Bank which says that, "[i]n banking and securities markets, characteristics related to private monitoring and enforcement drive development more than public enforcement measures." This is based on the article "What Works in Securities Law?" by La Porta, Lopez-de-Silanes, and Schliefer ("LLS") in the Journal of Finance, which says that public enforcement is less important in the development of stock markets than extensive disclosure requirements and liability rules that make it easy to bring law suits for violations of law. This conclusion is being used to advise developing countries on their development.

The messages of our current paper are as follows: there are advantages and disadvantages to both public and private enforcement; the empirical work behind the World Bank's recommendations did not capture what is really important; our measures, which use regulatory staffing and budget, may be more useful; the World Bank should not be favoring private over public enforcement at this stage; and the World Bank should carry out more research and probably recommend both public and private enforcement.

Looking at the mechanisms of public and private enforcement, there are several mechanisms on either side. On the private side there is litigation, voting power through the proxy process and shareholder rights, and the price/exit mechanism of the market. On the public enforcement side there is rulemaking, investigation and examination, enforcement actions, and criminal prosecution. Private enforcement has the advantage of shareholders with their interests aligned with correct market outcomes, but shareholders are often dispersed, find it hard to coordinate, and may have to use lawyers who do not always work solely for the shareholders' interests. Although public enforcement does not have the advantage of alignment of economic interests and may have the weakness of corrupt officials, it does have a wider range of sanctions and many public regulators are public spirited and try to do the right thing.

An important thing to understand is that these two spheres are not distinct. Looking at the situation in the U.S., I mentioned rulemaking under public enforcement, but litigation is directly linked to this. Public and private enforcement are also linked in the sense that when a public enforcement action is taken, it is much more likely that a private suit will be brought. The market tends to penalize companies that have enforcement trouble, so the price mechanism amplifies public enforcement.

In addition, private litigation may not be most effective due to defects in the litigation process; the role of lawyers for example. In the U.S. the wrongdoers often do not pay because they do not have the resources or they are indemnified by their firms. Also, in many of the major markets the enforcement mechanisms are not purely public or private; there are hybrid entities like self-regulatory organizations (SROs) and exchanges.

The prior work in this area, as used by the World Bank, used measures of enforcement that looked to formal laws, as is usual in comparative legal analysis. These economists tried to find indices that would illustrate how tilted toward public enforcement the regulatory agencies in different countries were. They used the indicia of independence, investigative powers, remedial orders, and criminal sanctions. With coding of this type it is always difficult to know whether the correct elements are coded and whether they are coded in the right direction, for example in the case of "independence." As used in the formal indices, independence means independence from both the government ministries and the banking and insurance regulators, but there remains the question of whether independent supervision is actually better than consolidated supervision.

We are therefore comparing the above "formal law" approach with an "actual resources" approach. The basic insight of our approach is that even if there are many formal powers, without staff these do not matter. Conversely, even if formal powers are limited, if there are a large number of well-motivated staff they can find ways in which to influence the market. The weakness of our approach is that sometimes a large number of staff does not translate into effective regulation, as was the case in the former Soviet Union, for example.

We focus on two kinds of measures of real resources: staffing and budgets. There is, as I mentioned, a lot of variation between countries. The overall average of regulatory staff per million of population is 10, but a country like Australia has 40, whereas Japan, Germany, and France are all below average. Budgets, which are affected by wage levels and are normalized by GDP, also vary considerably from country to country.

When comparing the two variables - the formal powers and the actual resources - there are some cases, like France, which score very highly in terms of formal powers but have very few regulators, and cases like the Netherlands in which the opposite is true. Japan, incidentally, gets 0 in the LLS public enforcement index. There is a correlation coefficient of approximately 0.4 between these two variables overall. The goal of our analysis is to consider which of the measures of public enforcement is more strongly associated with robust capital markets.

Looking at the regulatory budgets of civil law countries versus common law countries per billion dollars of GDP, the common law countries remain statistically significantly higher. Small countries with large financial markets, like Singapore and Luxembourg, look anomalous as they have bigger financial markets than their economies would suggest.

What is the effect of public enforcement on the size and robustness of securities markets? There are many possible ways to measure this, the most common of which is the ratio of market capitalization to GDP, but trading volumes, numbers of listed firms per million of population, and levels of initial public offerings (IPOs) to GDP are all objective measures that are sometimes used. We control for a number of factors, for example through the inclusion of private enforcement variables about disclosure, liability, the wealth of countries, etc. When plotting the relationship of staff to population against stock market capitalization, the suggestion seems to be that the more financial regulators, the bigger the capital market.

Our findings show that public enforcement does seem to be associated with more robust capital markets. Private enforcement remains important - particularly the disclosure variable, which tends to also have a positive relationship. The liability variable has a much less strong relationship, however. We ran many regressions using our variables and then compared the results of our public law indices against the formal indices to see which did better in so-called horse races. Our conclusion is that using the real resources variable is statistically stronger for public enforcement than using the existing public law indices.

There are always weaknesses in this type of empirical analysis, however. One potential problem is that of outliers affecting the statistical analysis. In the case of this study, the jurisdictions of Hong Kong and Luxembourg - small economies with big capital markets - are the outliers. When using the budget to GDP regression, if the two outliers are removed the positive relationship remains, but the line is not as steep.

The second problem is that of causation. Causation is not demonstrated by the above relationships. It is possible that capital markets become large and then they hire a lot of regulators, not that regulators cause capital markets. We believe that this actually works in both directions. It could also be the case that there is an independent variable - the number of retail investors, for example - which causes there to be more regulators.

There are a number of ways to solve causation, the best of which is to watch changes over time, but it might well be a hopeless task to try and go back and look at the changes over the last 10 years. Going forward, we will be able to collect the data, so in 5 or 10 years we will be able to know more.

In order to obtain some information about causation we used the (admittedly imperfect) corrective mechanisms of instrumental variable and two-stage regression analysis, which suggest some support for causation running from public regulators to capital market outcomes, but one must be cautious about interpretation of this data.. On a more anecdotal level, one thing which does seem to suggest against large capital markets causing public enforcement is the fact that large numbers of regulators are not hired in boom markets, but rather after financial crises.

There are two measures of capital market performance which our public enforcement variable does not say much about. The first is dispersed ownership. It seems that public regulators do not change the dispersion of ownership, i.e. the corporate structure of countries. Disclosure and liability rules do seem to have an effect on dispersion. It is not clear whether dispersion is necessarily good or bad. Conceivably, dispersion of ownership represents a political decision that public regulators do not get involved in.

The other area which we were hoping to find an effect in was the technical measures of stock market performance, e.g. bid-ask spreads, synchronicity, or volatility of stock markets. Puzzlingly, we could find no strong correlation between either public regulatory intensity or private law indices with technical measures.

In this paper we set out to show that public enforcement is important. Despite the weaknesses of the paper, I think that it at least shows that the World Bank should be more skeptical about the preeminence of private enforcement and should withdraw its recommendations.

In terms of areas for future research, perhaps it would be better to do this type of analysis looking at regulatory outputs (e.g. policy research, regulations, criminal sanctions) rather than the regulatory inputs of budget and staffing, as they may be better measures of regulatory intensity.

I have been looking at enforcement actions in different countries, beginning with the U.S. In the years 2002-2004 there were 3,630 securities enforcement actions in the U.S. resulting in $5.2 billion in penalties imposed on the market by public officials, with an additional $3.5 billion in private litigation awards. In 2004, the UK had 90 sanctions with $40.5 million in penalties, and Germany had 149 sanctions. Looking at 2004, the U.S. had far more enforcement actions per trillion dollars of market capitalization than the UK or Germany. The number of enforcement actions for Japan is steadily increasing and is approaching the value for the UK.

Comparing the enforcement penalties per billion dollars of market capitalization, there is a huge differential between the U.S. and the UK even though their normalized staffing numbers are similar. The penalties for Hong Kong are tiny and I imagine that the value for Japan would be low also. It is difficult to know how to interpret these sanctioning differentials.

There are clearly differences in enforcement strategies, and the issues of the differences in compliance levels and informal guidance may also be quite relevant. The investigation-to-sanction ratio is interesting, with the website of the US Securities and Exchange Commission (SEC) proudly stating that they obtain formal sanctions in 90 percent of their investigations, whereas the FSA in the UK aim to resolve the majority of investigations without formal sanctions, but rather through the payment of penalties in settlements. What is being sanctioned is also very different in different countries. The U.S., for example, has many more issuer sanctions than most other jurisdictions. Also, understanding where sanctions are coming from and how they should be allocated in the modern global market is becoming very complicated.

I have today looked mainly at the question of the establishment of financial markets in developing countries, but perhaps more relevant for Japan is the question of acceptance of financial regulations by other developed countries. This issue is at the forefront in U.S.-EU relations, with the SEC having to consider whether alternative accounting rules should be accepted in the U.S., i.e. EU issuers with International Financial Reporting Standards (IFRS) versus U.S. Generally Accepted Accounting Principles (GAAP). When considering this question, the SEC will also be looking at whether enforcement of accounting standards by regulators is somewhat analogous to that in the U.S.

In addition, there are currently proposal being considered at the SEC to allow foreign exchanges to enter the U.S. with remote terminals without being subject to U.S. regulation, which would allow foreign regulation of exchanges and broker-dealers. The acceptance of substitute compliance would be a dramatic change in the SEC's position. In making this decision one thing that they will look at will be whether the enforcement efforts and the technical measures of stock market performance are comparable to those in the U.S. Hence, this will be an important issue to the SEC's international division and I am very appreciative of having had the opportunity to be here in Tokyo and learn more about Japanese enforcement.

Questions and Answers

Q: When assessing the regulatory structure of Japan, did you take into consideration that the regulatory service is part of the civil service, hence its LLS independence rating of 0? Secondly, with reference to your discussion on U.S.-EU regulatory dialogue, what is your view on the prudential carve-out in relation to the World Trade Organization (WTO) and the General Agreement on Trade in Services (GATS)?

A: I was surprised that Japan had a LLS rating of 0. It cannot be just due to the independence issue, although when the coding was done regulation was even more centralized in the Ministry of Finance. I understand that the coding was done prior to the expansion and reform of remedial powers and criminal provisions, so the problem is that their assessment is not up to date. Having said that, many papers have been written based on and correcting their work, so although I think their original work was flawed, it has certainly opened up the whole area for discussion.

On the WTO question: in the financial services dialogue, the U.S. position with respect to exchanges has been challenged as protectionist by those in Europe. The SEC sees it as investor protection and therefore very much locates it within the prudential exemption. Whilst the SEC defense is credible, the Europeans are also correct that it has the effect of helping the New York Stock Exchange gain foreign listings for European firms. I am not a WTO expert but I do think that this kind of a regulation would sustain a prudential exemption defense; nevertheless, concerns over protectionism are very much in the background.

Q: Given the state-based nature of insurance regulation in the U.S., is the estimate in your report of 13,056 insurance staff not quite low? As for Japan coming at the lower end of the scale for regulatory intensity, does this not have something to do with Japan being a civil law country that used to rely on ex ante regulations and rule-based systems?

A: I will take another look at the insurance data in due course, but it is the case that insurance and banking are the areas in which the U.S. is most out of line with the rest of the world, precisely due to the state system. Comparatively, the portion of the FSA dealing with insurance in the UK or Japan, for example, is tiny. It depresses me to think that I have underestimated the figures for U.S. insurance regulation as the U.S. is already a very significant outlier.

Please bear in mind when looking at these distributions that we are now only in the preliminary stages of acquiring a sound economic basis for judging what kind of regulation actually helps. As it stands, empirical evidence for the value of regulation is minimal. Most practitioners have confidence in the fact that what they do has a positive effect, but finding evidence of that is difficult. Some empirical work has been done that suggests insider trading regulation has an effect on the market, but even in this case we do not know whether it is worth the cost.

When looking at Japan with its lower-than-average figures for securities regulation, we see that there are some other civil law countries that are somewhat comparable. My feeling is that Japan's capital markets are more robust than those other civil law countries. It is quite possible that there are good reasons for Japan to be in that particular position, just as there may be good reasons for the U.S. to be a significant outlier.

Q: Do you feel that the legal system of a country has a bearing on the effectiveness of its private enforcement?

A: When considering private enforcement, one has to take into account a country's legal system and the structure of private litigation. Interestingly, some jurisdictions have attempted to increase the use of their legal system on the criminal side, but found it difficult to change their views of their judges. There is also the question of whether a country has the mechanisms to bring class actions.

The area of private remedy is broader than just litigation. There are intermediate solutions that do not depend on traditional courts or class actions: arbitration mechanisms for client disputes in the U.S. and the public mechanism for private compensation of the ombudsman in the UK have both been seen to be effective.

Class actions should probably be the last line of regulation, not the first. Many in the U.S. want to move away from class actions, but it is very difficult as the industry and the attitudes of judges are very much geared towards class actions. Therefore, yes, the legal system of a country does play a big role.

*This summary was compiled by RIETI Editorial staff.