I last attended the ABS annual dinner two years ago. I spoke then on the need to free up our banking sector, and to allow competition a greater role in shaping the industry. Last year, MAS announced measures to improve corporate governance and institutionalise the management of the local banks, and the first phase of a five-year programme to liberalise the banking sector. We sought to strengthen Singapore’s banking system, provide Singaporeans with quality banking services, and further enhance Singapore’s position as an international financial centre.

These reforms have brought about visible changes. The next step in strengthening the local banks is now to separate financial and non-financial businesses within the banking groups, and to further improve corporate governance through a clearer and more transparent ownership and control structure.

What ought to constitute the permissible activities for a banking group is an issue which all financial regulators have to consider carefully. If we took a purely laissez faire approach, we could leave banks to structure themselves as their owners and management saw fit. What businesses banks chose to be involved in, and what business entities chose to own a bank, would be matters for themselves to decide. Over time, banks with the right configuration and mix of activities would succeed, while those without would fail or be eliminated by the competition.

But regulators cannot leave it entirely to the market to sort out such matters, especially where the banks involved are not peripheral players but perform key roles in deposit taking and the payments system. Strong banks are important to the health of the financial system and to the entire process of credit intermediation. Regulators cannot guarantee that individual banks will never fail, but they have a duty to minimise the chances of a mishap and to maintain the integrity and overall stability of the financial system.

A clear source of potential problems in this regard is the commingling of financial and non-financial activities within the same corporate group. Many regulators regard it as good banking and sound supervisory practice to require clear separation of the two.

Firstly, separation will limit the risk of contagion from non-banking businesses to the bank. Banks with significant interests in non-banking activities are exposed not just to credit and market risks – which are the primary risks of their business, and which the banks should assess and manage as part of their core competence. These banks will also be taking on the business risks of their non-financial activities. If their non-financial activities run into trouble, the banks may feel compelled to support them beyond normal commercial considerations, to protect their own reputation. In so doing the banks may undermine their own soundness; yet if they did not extend the rescue, the problems of the affiliates may trigger a loss of confidence in the bank itself because of the close association.

Secondly, with banking and non-banking activities inter-meshed within a conglomerate, there will be a strong tendency to stretch any safety net intended for the banking system also to cover non-bank operations in the group. If failure of an associated non-bank threatens the bank and systemic stability, the regulator or government may be forced to rescue both. We have seen this happen in the recent Asian crisis. An implicit safety net for the banking system may be a necessary evil; but extending it to non-banks compounds the problem of moral hazard. It also tilts the playing field against non-bank competitors who are not associated with banks.

Thirdly, separation minimises the risk of non arms-length transactions between banks and their non-bank affiliates. When a bank lends to a related entity at more favourable terms than market, or invests in an affiliate in order to control it even though the investment does not perform, minority shareholders of the bank are disadvantaged, and so are depositors.

Some economists have argued that in less-developed financial markets, it is in fact a rational and practical strategy for a group of companies sharing common ownership or cross-shareholdings to lend to and invest in one another. Information about companies is often less than complete, and company accounts and business plans are hard to interpret. Family networks or guanxi are then efficient ways to establish a basis of trust for doing business, and to avoid the high costs of obtaining reliable knowledge about the other party’s activities and finances.

But the non-transparency of this approach can lead to serious difficulties and abuses. Cross-shareholdings diminish the influence of minority shareholders, dilute market pressure on the companies in the group, and weaken the internal discipline to appraise rigorously the risk of each investment.

The on-going reforms in many Asian countries are as much about tightening corporate governance in this regard as they are about re-capitalising weak banks. The way forward for banks is therefore not to rely on family networks or guanxi. It is to strengthen their ability to assess credit risks and investment propositions rigorously, on arms-length, commercial considerations, and to deploy their assets in a diversified, optimal manner.

Finally, separating out non-financial activities will focus management attention on their core business of banking and finance. The need to concentrate on core competencies is not unique to banking. Globalisation and technology are causing many world-class companies to focus on their core competencies, and to outsource parts of their value chain for which they have no particular competitive advantage. This is especially true for banks, which are operating in an intensely competitive and dynamic environment. They need to muster all their talents and energies on the financial business, instead of dissipating them over many different businesses.

For these reasons, banks should not commingle financial and non-financial activities. However, this does not mean that banks must refrain from engaging in venture capital or merchant banking business, which involves taking an equity interest in a diversified portfolio of non-financial companies for some period. It is not uncommon for the merchant banking business of well-run, international banks to take stakes in non-financial companies for three to five years. These activities are natural extensions of the lending business of banks. They are, however, quite different from taking long-term strategic stakes in non-financial companies which tie a bank’s reputation to these companies.

In addition to the separation of banking and non-banking activities, regulators in several developed markets have also advocated dispersed ownership of banks. Often this has been achieved by legislation or via the bank's own memorandum and articles of association, which sets limits on shareholdings to prevent any single shareholder or group of shareholders from dominating the bank. The result has been a "clean" corporate structure for their banks, i.e. there are no cross-shareholdings between banks and non-banks, and there is clear separation of ownership from management.

The US is an example of a financial system which has prohibited commingling between banking and commerce, and required dispersion of bank ownership. For many years the Glass-Steagall Act restricted commercial banks even from certain other financial activities, particularly investment banking and insurance, much less non-banking business. Now that Glass-Steagall has finally been repealed, allowing for more competitive banking institutions, the US Federal Reserve Board is proposing that US financial holding companies should only undertake a well-defined range of approved activities.

In the UK, there is no legal prohibition on banks participating in non-financial activities. But in practice few of the banks do so. This is partly the result of stringent treatment of capital adequacy by the regulator – banks are required to deduct from their regulatory capital dollar for dollar of any investment they make in non-financial activities. It is also because the banks are widely held, with no dominant shareholder who has other non-financial interests. The banks therefore have little incentive to control non-financial companies through their investments.

Germany has had a different approach. German banks have traditionally held major long-term equity stakes in industrial groups, such as Daimler-Benz or Thyssen. This arrangement gave the industrial groups long-term strategic shareholders, who could look beyond the next quarter’s financial results. But it has also frozen the structure of the industrial groups, and prevented them from merging, restructuring and rationalising, as US companies have done. It has made it harder for the banks to deploy their assets flexibly, to maximise their own returns, and to become as spry and competitive as US banks.

The Germans have recognised these structural weaknesses and are taking steps to rectify them. However, one major obstacle to German banks divesting their industrial holdings has been the prohibitive capital gains tax. Now a Social Democrat government has proposed tax reforms that will remove this obstacle, and pave the way for banking and industrial conglomerates to untangle their web of cross shareholdings.

In Asia, both South Korea and Japan relied on close association between banking and commerce in the earlier stages of their industrialisation. South Korean chaebols, and the banks that lent to them, are now facing painful reforms. Poor lending practices by Korean banks disguised the deteriorating finances of the chaebols, and eventually nearly brought the whole system down in the Asian crisis.

In Japan, the pre-war zaibatsu system became the post-war keiretsu system. Each major keiretsu or industrial group consists of companies linked together through a web of cross-shareholdings, including a main bank. Keiretsu members support one another financially and in terms of direct business. The bank in the keiretsu plays a pivotal role, facilitating the financial transactions of the group.

While the keiretsu system helped Japan to rebuild and industrialise its economy after the war, many of the problems of corporate governance and unsound banks that now plague Japan are rooted in this structure. The prolonged crisis of the Japanese economy since 1990 has forced a fundamental rethink. The Japanese are now gradually starting to make basic structural changes. The banks are less willing to lend to associates without independently assessing their risks and charging the appropriate risk premiums. And both banks and non-banks within the group are no longer willing to maintain shareholdings in each other which yield low returns. The changes will enhance transparency and corporate accountability. They are essential to Japan’s economic recovery, but will have far-reaching social and cultural implications.

In Singapore, the local banks (other than DBS) started as family-owned banks and built up a successful financial business on that basis. Over time they have merged and consolidated. Today the founding families remain the principal shareholders, and are in several instances also involved in the management of the bank.

Besides banking, the principal shareholders have also built up diverse non-banking businesses, especially in property. The local banks participate in these non-financial activities of the group, both by lending and through equity ownership. The banks and non-banks in the same group have developed significant cross-shareholdings. A mixed conglomerate structure has evolved.

The commingling of financial and non-financial activities will potentially expose Singapore banks to the same problems that banks elsewhere have experienced. We have not run into serious problems, so far. This is because the principal shareholders have run the banks prudently and properly. The MAS has also been stringent in monitoring related party transactions. Strict laws govern such transactions; for instance, bank directors are jointly and severally liable to indemnify the bank for losses arising from unsecured loans to related companies.

But problems will manifest themselves only in times of stress, as the Japanese and Korean experiences have shown. For the future, we need to restructure the banks’ participation in non-financial activities before potential vulnerabilities become actual problems. We have the luxury of acting now while the banks are sound; we must not delay or put off changes which are essential for their long-term well-being.

The banking environment in Singapore will only become more open and competitive, and less forgiving of weaknesses or under-performance. Banks must concentrate on their core competencies and build strong and institutionalised management structures to compete in a globalised industry. Ultimately, their improved performance will benefit all depositors and shareholders, and strengthen the financial system as a whole.

The Government has therefore decided to require the local banking groups to segregate their financial and non-financial activities, and unwind the cross-shareholdings between the two. This will mean changes to their present corporate structures. These measures will apply to all local banks – OCBC, UOB and OUB, as well as DBS and KTB.

There will be four key elements in the separation between the financial and non-financial entities: ownership, cross-shareholding, management, and name-sharing.

MAS will require all the financial activities in each banking group to be grouped together, either under the bank or under a non-operating financial holding company (FHC). The non-financial activities must be segregated from the banking group and divested. They can be sold to third parties, or alternatively to the principal shareholders of the bank, so that the principal shareholders own them directly, and not through any of the entities in the financial arm of the group. MAS will regulate the bank or FHC, but it will not regulate the non-financial activities.

Currently, all the local banks are listed, and so are many of the non-banks in the groups. The banks and non-banks already have separate sets of shareholders. In fact, the minority shareholders, i.e. the shareholders other than the principal shareholder, typically account for about 70% of the ownership of the bank.

MAS will require this separation of ownership to be retained. The bank or FHC in the existing banking groups must continue to be listed on its own. For example it cannot be subsumed as a 100%-owned subsidiary of a listed top-level holding company, which owns both the bank as well as the non-financial entities.

Having distinct groups of shareholders for the financial and non-financial activities will strengthen market discipline on the bank, and facilitate clearer accountability of its performance to minority shareholders. It will help to ensure that the separation of the financial and non-financial activities is substantive and not just pro-forma.

There are many ways for banking groups to reconfigure their corporate structure. In the end, what counts as a FHC is a matter of substance, and not of the form of the organisation chart. MAS reserves the right to declare any company which owns a substantial share of a bank, directly or indirectly, as a FHC. Such an FHC will be subject to MAS regulation and will not be allowed to own non-financial activities.

The financial entities will not be allowed to own, directly or indirectly, the shares of non-financial firms affiliated to the same principal shareholder.

Within the financial arm, shareholdings should only be in one direction. We will not allow two companies in the financial arm to have mutual shareholdings in each other. For example if a bank takes a stake in an insurance company or securities firm, then the insurance company or securities firm would not be allowed to hold a stake in the bank. The Companies Act already prohibits subsidiaries of banks from owning shares in the bank. We will extend this principle to all entities below the bank or FHC, whether or not they are subsidiaries of the bank or FHC.

Ideally, we should also disallow the non-financial entities from owning shares in the financial entities. However, we do not start from a clean slate. Requiring the banks to divest themselves of the non-financial activities will already be a major structural change. Therefore, in the present exercise MAS will allow the principal shareholders to continue owning indirect stakes in the banking groups through the non-financial entities that they control. MAS will review this arrangement later.

To avoid conflicts of interest, the management of the financial entities should be separate from the management of its non-financial affiliates. There should be no sharing of executive directors and management staff, such as the chief financial officer and chief operating officer. On the board of directors of a regulated financial entity, a majority of the directors should not be holding directorships on the boards of the non-financial affiliates as well.

Mixed conglomerates usually leverage on common brand building. Many of the non-financial affiliates of local banks share the name of their parent banks. Financial and non-financial entities alike have built up goodwill in the brand name over the years.

But reputational and contagion risks can result from sharing of names and corporate badges between the bank and its non-bank affiliates. Where an unregulated entity sharing a name fails or falls into disrepute, the public may attribute its problems to the bank, causing a bank run.

MAS will therefore disallow the sharing of names and logos between the financial and non-financial entities. We are considering amending the Banking Act, to prevent the use of names or acronyms by companies that may confuse the public, or give the impression that a company is related to or associated with any Singapore-incorporated bank.

MAS will determine the scope of activities which the bank or FHC will be permitted to undertake. Besides financial activities, they will include activities which have clear synergies with the bank’s financial business. We will not have a fixed list of permitted activities. Instead we will exercise flexibility to allow banks to undertake business ventures that are related to their core competencies, such as e-commerce activities that are changing the banking industry worldwide. These include providing payment services in support of e-commerce, and leveraging on banks’ IT competencies to provide IT services, especially in building infrastructure and applications for B2B e-commerce.

Banks will also be allowed to undertake equity portfolio and venture capital investments. We will define an equity portfolio investment to be one where the bank and its financial affiliates do not have control over the investee company, e.g. by being the largest shareholder. An equity portfolio investment should generally not exceed 10% of the investee company’s capital. Furthermore, each equity portfolio investment will be subject to a single security limit of 2% of the bank’s capital funds. We will amend Section 31 of the Banking Act to remove the existing aggregate limit on portfolio investments, and replace it with this new single security limit. We will also exclude debt instruments from Section 31 because they are generally less risky than equities.

Venture capital investments will be treated separately. MAS is drafting separate rules to allow for venture capital investments by banks, and will consult them on these guidelines soon.

The local banks have significant property holdings. This is partly because over the years the banks have grown together with the property business of their principal shareholders, and also because property has been seen as a good and safe investment in Singapore. Some of these property holdings were acquired many years ago, and have been kept on the banks’ books at book values which are well below their present market values. This has provided a comfortable cushion of reserves for the banks against uncertainties. But it has also failed to reflect the true cost of holding these properties and therefore reduced the pressure on the banks to optimise returns on assets.

Property assets that earn competitive returns have their place as one component of the investment portfolio of the local banks. MAS will therefore allow the banks to hold properties, solely for investment purposes, and subject to a limit aimed at preventing excessive risk concentration. How much property a bank actually holds within these prudential limits is for its management and shareholders to decide.

However, banks should not be engaged in developing or managing hotels or other properties, whether directly or through companies they control. Banks are not property companies. MAS will therefore require banks to divest all properties held for development, and shareholdings in property development companies beyond the portfolio investment limits.

The aggregate net book value of permitted property holdings will be subject to a revised Section 33 limit of 20% of a bank’s capital funds (reduced from the existing 40%). The 20% limit translates into less than 5% of each of the local bank’s total assets currently. It will apply on the basis of both bank and group capital. As with the present Section 33, properties held for the purpose of the bank’s business premises will be excluded from this limit. But whilst the present Section 33 applies only to properties directly held by banks and their subsidiaries, the new limit will also include properties held by companies in which the banking group has shareholding of 10% or more.

Most of the local banks have insurance companies within their groups. These insurance companies present a slight complication. They invest two separate types of funds: the equity in the company belonging to its shareholders, and the insurance funds managed by the company but belonging to the policyholders, for example the life fund of life insurance companies.

Portfolio investments made using the shareholders’ funds of the insurance companies will be subject to the same single security limit of 2% of the group’s capital funds and the limit of 10% of the investee company’s capital. The portfolio limits for the banking group as a whole will therefore include the investments using these shareholders’ funds of their insurance companies.

MAS recognises that the insurance funds are different in nature from the equity of the company or the group because they belong to the policyholders and not the shareholders. Therefore in principle investments using the insurance fund should be treated separately from the investments of the company and the group. Nevertheless the insurance fund should also be used only for portfolio investments, and not for the purpose of controlling other companies in the group, especially non-financial companies, whether by itself or together with other investments by the group.

Currently, some banking groups have used these insurance funds as the vehicle for owning and controlling their investee companies. For example the bank itself may own only 5% of an investee company, but the insurance fund of the insurance company may own another 15%, giving a total stake of 20%. This is not satisfactory, because it presents the same problems of contagion, non-transparency and dilution of management focus as the bank itself investing directly in these companies.

MAS will study the appropriate investment limits of insurance funds for both bank-related and independent insurers. The limits set should protect the interests of policyholders through adequate diversification, and ensure that the investments do not amount to controlling stakes. These measures will be developed in consultation with the industry, and are expected to be finalised by the end of the year.

Whatever the precise portfolio investment limits for the insurance funds, MAS will take the governance and deployment of these insurance funds into consideration, when assessing whether the banking group has de facto control of an investee company, and thus whether an investment meets the portfolio investment criteria for the banking group.

The separation of the financial and non-financial activities will be a major change for the banking groups. There will be significant corporate restructuring and divestment of assets. We must do this in an orderly manner, over a reasonable period. There is no reason to force a fire sale, which would weaken the banks and diminish confidence, and would be against the interests of shareholders and depositors.

Local banks will be given three years to complete the restructuring and comply with the new requirements. The three years will commence from the passage of the relevant legislation, which should take place by the end of the year. MAS will work with the banks on a phasing and schedule, so that the restructuring and divestments can take place in an orderly manner and be completed within the three-year deadline.

The divestment of non-permitted assets will have significant tax implications. In order to avoid uncertainty as to how IRAS will treat these transactions, and to minimise the tax impact on the banks of changes which have been mandated by MAS, the Government has decided on a set of one-off administrative concessions to facilitate this exercise.

All gains arising from disposals of shares and real properties which have been held for more than 10 years will be treated as non-taxable capital gains. For assets held for less than 10 years, IRAS would look at the facts of individual cases to decide whether capital or income treatment is appropriate. The Government will also waive stamp duty on sales of all shares and properties where the gains are considered capital in nature.

These concessions will not apply to gains arising from the disposal of shares which have previously been claimed as trading stock for tax purposes, or gains on disposals of real properties by banks’ property development companies. However, where income tax is payable arising from a mandated sale of assets that does not qualify for capital gains tax exemption, the Government will give the banks three years to settle the amount.

This package will apply only to assets required to be disposed of under the restructuring exercise, and will only cover disposals made within the three-year period allowed by MAS for the restructuring.

MAS has discussed our plans with the local banks. They agree with the policy direction proposed. They have suggested some modifications to MAS’ proposals, which wherever possible MAS has incorporated.

Our approach is pragmatic, not purist. Our circumstances are different from other countries. We need to strike the right balance, to find the framework that suits our economy, and which recognises the historical developments that have brought about the present situation. The principal shareholders remain a source of strength to the banks, and it is not our intention to conscribe their contributions to the future banking business.

The separation of financial and non-financial activities will be another major step forward in strengthening the local banks. As one analyst wrote in March:

I urge all the banks to continue to upgrade their capabilities and build up their management teams, in order to prepare themselves for the new opportunities and challenges arising from the rapid changes of technology and the market. MAS will continue to work with the local banks to foster closer co-operation and strengthen the culture of openness as we liberalise the banking industry.