I. Sovereign wealth funds: what does it mean?

When this author introduced the term “sovereign wealth funds” five years ago, he fully expected to see these institutions grow in number, size, and market presence.Footnote 3 What he did not expect, however, was the sheer level of controversy and the divergence of opinions surrounding SWFs that emerged in the subsequent debate, often precluding a calm and rational analysis of the benefits and risks presented by these entities. One particular reason for this was the confusion about the term: different commentators meant different things when they talked about SWFs and there was no universally accepted definition.Footnote 4 This led to inconsistencies as to which funds were being included in the analysis and why. There was also no consensus on how much money they managed: even the very rough estimates one found in the public domain would vary by up to US$1 trillion. Growth projections for these funds were even more suspect, based on very arbitrary assumptions and containing huge estimation errors.

Of course, the funds themselves were, and continue to be, far from homogenous. There is no such thing as a “typical” sovereign wealth fund, as they differ in structure, governance, policy objectives, risk-return profiles, investment horizons, eligible asset classes, and instruments, not to mention levels of transparency and accessibility. In order to make the debate meaningful and the policy response fair, it is important to get the basics right and to develop a shared analytical framework that would be as inclusive and comprehensive as possible, while at the same time being objective and rule-based. It is in this spirit that the author tried to define and describe SWFs back in May 2005. Below we revisit the original definition and then contrast it with several alternative approaches. We then consider the current definition of an SWF, as formulated in the Generally Accepted Principles and Practices (GAPP), also known as the Santiago Principles. We consider the pros and cons of this definition and potential ways of further improving it.

II. The original definitionFootnote 5

In the original article, SWFs were defined as sovereign-owned asset pools, which are neither traditional public pension funds nor reserve assets supporting national currencies. There are three key components to this definition, two of which help us define the broad SWF universe by specifying what these funds are not, rather than what they are. Specifically, the funds are: (1) sovereign owned and managed; (2) assets which are not prudential monetary reserves; and (3) assets which are not traditional pension funds.

The first component is fairly straightforward. The only point to add is that the concept of sovereignty is typically associated with national or federal authorities, but in this context it can also include subnational and subfederal entities.Footnote 6 The second and third points allow us to broaden the universe to cover all sources of sovereign wealth. Central banks and monetary authorities maintain a certain amount of foreign exchange reserves for prudential policy purposes. Such reserves cannot and should not be viewed as sovereign wealth. However, if they end up accumulating considerable “excess” reserves—substantially above what might reasonably be required for traditional policy purposes—then this excess portion constitutes sovereign wealth, whether or not the authorities choose to acknowledge and manage it as such.Footnote 7 It may not represent “net sovereign wealth”, insofar as assets are supported by local currency-denominated sterilization or other debt on the liability side of the balance sheet, but it is sovereign wealth nonetheless.

Some commentators have suggested a separate, if overlapping, category of sovereign pension funds.Footnote 8 It may, in fact, be more helpful to distinguish between various national pension schemes in terms of sources and ownership of the underlying pension money. If the main source of the fund is individual and employer contributions on behalf of current and future retirees, and if the underlying beneficiaries are effectively the legal owners of the fund, then the entity cannot under any circumstances be considered an SWF. The fact that a sovereign institution happens to be the agent entrusted with managing the fund does not make it meaningfully different from similar subsovereign public or private pension plans. If, on the other hand, the source of the money is not pension contributions, but taxes, privatization proceeds, and other payments out of the general budget, and if the fund is effectively owned not by the underlying beneficiaries, but by the taxpayers, then it can legitimately be considered a SWF—which just happens to be managed to meet a particular future pension liability on behalf of the sovereign state.Footnote 9

The above definition defines sovereign wealth in the broadest possible way. As long as the assets in question are sovereign-owned and do not represent either prudential monetary reserves or classic public pension money, they must be sovereign wealth—regardless of whether they are domestic or foreign, equity-like or debt-financed, earmarked for current or future generations, highly liquid and broadly diversified, or relatively illiquid and concentrated.

The broad and universal nature of this definition is its biggest strength, but it is also its most serious weakness. While it captures all sovereign investment vehicles in whichever shape or form they may exist, it fails to elucidate the key features and distinct characteristics of the funds in question. In other words, it is just too vague to be applied on its own. Therefore, to make this definition more structured and analytically robust, one needs to augment it with a simple, rule-based classification system to clearly describe the different types of SWFs within its universe. In a subsequent article this author proposed to use liability profiles of SWFs as the basis for such a classification.Footnote 10

Liability profiling essentially requires asking two questions: first, what are the sources of funds, and second, what are the intended uses of funds? On the former question, a fund can represent an equity-like claim by the sovereign on the underlying assets. This is typical of commodity-based SWFs, such as those in Norway, Russia, and most of the Middle East, which represent net savings by their respective governments. On the other hand, an SWF may be funded through the issuance of local currency-denominated debt, which is more typical of entities created from excess foreign exchange reserves of respective central banks, with the Republic of Korea and China being classic examples.

On the latter question, one can think of four different types: (1) contingent liabilities, typical of stabilization funds; (2) fixed future liabilities, typical of national pension reserve funds; (3) mixed or endowment-type liabilities (i.e. perpetual capital with interim outflows); and (4) open-ended liabilities (i.e. effectively no contractually attached liabilities).

Contingent liability funds are typically set up to smooth budget revenues and expenditures, help sterilize excess liquidity, and protect the economy from overheating, “Dutch disease”,Footnote 11 and boom-bust cycles. Their liabilities are contingent on volatile and unpredictable commodity prices, and as such these funds will typically be managed with objectives similar to central bank reserves: safety, liquidity, and return. The focus of these funds is on nominal returns denominated in foreign currency.

Fixed future liability funds refer to entities such as France’s Fonds de Réserve pour les Retraites (FRR), Ireland’s National Pension Reserve Fund (NPRF), Australia’s Future Fund, and New Zealand’s Superannuation Fund. These SWFs were set up to meet a fixed liability of the sovereign twenty to thirty years into the future, which happens to be a projected shortfall in the public pension system. These funds resemble a large pension plan with a very young workforce and very few, if any, current retirees. At inception, it is extremely flexible in what it can do on the asset side of the balance sheet, but as it matures, it becomes much more constrained.

Mixed liability funds are essentially endowment-type future generation funds, like the ones in Norway and Russia. Their liabilities are mixed in the sense that they have a contractually fixed obligation to make regular payments into the budget—the so-called “fiscal rule” or “spending rule”. But at the same time they are open-ended funds with no targeted terminal value. They may not have quite the same degree of freedom on the asset side compared to a newly launched fixed liability fund but, as time goes by, they maintain their relatively high degree of freedom, while the fixed liability fund is increasingly constrained.

Finally, open-ended liability funds are essentially the various investment authorities and government investment corporations. These funds do not have readily identifiable or contractually defined obligations: they have neither formal spending rules nor liability shortfall targets. In theory, these funds have the longest investment horizon, the maximum capacity for risk taking, and the broadest possible latitude and degree of freedom on the asset side of their balance sheets. Quite simply, they can afford to do with their money whatever they want.

To summarize, the main benefits of the above approach to defining and describing SWFs are twofold. First, it allows for the broadest possible demarcation of the SWF universe—there is no risk of leaving any fund behind simply because the definition happened to be too narrow. Second, by introducing a classification system based on fund liability profiles, the universe can be broken down into logical and internally consistent segments, which makes the analysis of SWFs tractable.Footnote 12 The downside, however, is that this approach to defining SWFs is “messy”, in that it does not produce a short, snappy, universally applicable, and simple-to-use definition. As we shall see in the examples below, alternative approaches to defining SWFs tried to achieve better precision at the expense of narrowing the universe.

III. Alternative definitions

In a speech in June 2007, Clay Lowery, in his capacity as Acting Under Secretary for International Affairs at the US Treasury, offered the following definition of a sovereign wealth fund: “I will use the term [sovereign wealth fund] to mean a government investment vehicle which is funded by foreign exchange assets, and which manages those assets separately from official reserves.”Footnote 13

While this definition is reasonably succinct and precise, it is open to interpretation and manipulation. For example, a fund like Temasek Holdings of Singapore is not funded by foreign exchange assets, so technically, according to this definition, it is not an SWF. Or consider Korea Investment Corporation (KIC); while operationally it is separate and independent from the central bank, the latter maintains a call option on the reserve assets it contributed to KIC at the time of its formation in 2005. So, technically, one could argue that the bulk of KIC’s assets are not entirely separate from the Bank of Korea’s official reserves and hence may not constitute sovereign wealth according to this definition.

In October 2007, the International Monetary Fund made its first attempt to define SWFs in the Global Financial Stability Report: “SWFs can generally be defined as special investment funds created or owned by governments to hold foreign assets for long-term purposes.”Footnote 14 This is an improvement on the US Treasury’s definition, in that separation from official reserves is no longer a limiting condition. However, the focus on foreign assets potentially leaves out any sovereign fund that happens to be invested exclusively or primarily in local markets.Footnote 15 However, it is well known that during the first two decades of its existence Singapore’s Temasek invested almost exclusively in domestic assets. Many sovereign funds in Asia and the Middle East continue to do so today.

Stephen Jen, a widely quoted financial analyst and a knowledgeable authority on SWFs, offered this take on what constitutes an SWF in one of his research reports: “In my view, a SWF needs to have five characteristics: (1) Sovereign; (2) High foreign currency exposure; (3) No explicit liabilities; (4) High risk tolerance; (5) Long investment horizon.”Footnote 16 While this definition is very precise, it severely limits the universe of sovereign funds by excluding all conservatively run stabilization funds with low risk tolerance and short investment horizon and all sovereign funds that are invested primarily in domestic assets. It also effectively excludes all national pension reserve funds, since they have explicit future liabilities.

In and of themselves, none of these limitations necessarily invalidate the proposed alternative definitions. As long as the analysts using the terms are fully aware of these limitations and make sure to choose the correct interpretation for their needs, such differences need not present a problem. However, once we move into the realm of legal definitions and statutes, one has to be extra careful in how one defines an SWF.

IV. The iwg-swf definition (aka the “santiago principles” definition)

In October 2008, a group of SWFs, working closely with the IMF, formulated and agreed upon an SWF definition which then formed the basis for the Santiago Principles and related official documents. This has now become the consensus definition, for at least two reasons. First, it was developed and affirmed by the International Working Group of Sovereign Wealth Funds (IWG-SWF): in other words, it was produced by a forum of SWFs working on a set of principles for SWFs. It cannot get more genuine and authoritative than this. Second, the definition strikes a reasonably good balance between precision and breadth. It also explicitly offers a fund classification by sources of capital:

The Santiago Principles further provide three key specifications that must be in place for a fund to be legitimately classified as an SWF. The three key elements that define an SWF are:

Interestingly, the Santiago Principles also specify which market participants should not be classified as SWFs:

This definition provides a clear and solid framework for any formal discussion of SWFs. However, there are at least three specific groups of sovereign investors who are not covered by this approach, whose activities may—potentially—present similar policy and regulatory challenges that the SWF debate is trying to address.

The first group are sovereign funds that are currently outside the scope of the formal IWG-SWF definition simply because they invest exclusively or primarily in domestic markets.Footnote 20 Many such entities have been set up in developing countries around the world: Khazanah Nasional Berhad in Malaysia, State Capital Investment Corporation in Vietnam, Samruk-Kazyna National Wealth Fund in Kazakhstan, Mumtalakat Holding Company in Bahrain, Mubadala Development Company in Abu Dhabi, Oman Investment Fund, to name just a few. Some of these sovereign funds are already active investors overseas, while others can be expected to follow this trend in the future. The explicit exclusion of such funds from the Santiago Principles framework looks artificial.Footnote 21

The second group are central banks and monetary authorities with excess foreign exchange reserves. The core of these assets will be managed as liquidity, not sovereign wealth, to facilitate any balance-of-payments problems and to effect currency interventions. However, there is a strong possibility that the excess portion of reserves, however defined and measured, will be increasingly viewed and managed as sovereign wealth. Arguably, central banks and monetary authorities in China, the Republic of Korea, Singapore, Saudi Arabia, Hong Kong, and some other countries and territories have already been managing a portion of their overall assets along these lines.Footnote 22 As central banks in emerging market economies continue to accumulate foreign exchange reserves at a rapid clip, it would not be unreasonable to expect, over time, more reserve managers to begin managing at least a portion of their assets in this way.

The third group are investment funds owned by sovereign rulers, primarily in the Middle East. They may run into perception problems when investing overseas and might therefore benefit by being included in the category of SWFs. While in theory the Santiago Principles are correct to explicitly exclude assets managed for the benefit of individuals, in practice the line between sovereign wealth and the private wealth of the sovereign can be quite blurry.

If policy-makers and regulators around the world considering SWF-related matters want to avoid any misunderstandings and unintentional mistakes, let alone potential regulatory arbitrage or intentional abuse of the system, they would do well to consider working with the SWF community on broadening the scope and sharpening the focus of definitions and qualifications contained in the Santiago Principles.

Specifically, they may want to suggest that the recently established International Forum of Sovereign Wealth Funds,Footnote 23 together with the IMF, consider inviting the first group of institutions to sign up to the Principles and join the SWF community formally. As for the other two groups, it may be constructive to establish ongoing consultations, with a view to potentially inviting these entities to participate in the work of the Forum as “affiliated members”.

V. Definitional challenges with transparencyFootnote 24

The biggest issue that most policy-makers and commentators seem to have with many SWFs is their perceived lack of transparency. Coupled with the recent string of high-profile private-equity-style transactions, this has led to increased pressure on SWFs to become more open and transparent and to subscribe to a code of best practice. In fact, this push for transparency was the major impetus behind the Santiago Principles initiative. Given the increased size and market presence of SWFs, and their expected growth trajectory, it is not surprising that many countries on the receiving end of sovereign wealth flows feel the need to understand more and get more comfortable with these funds. A meaningful increase in transparency and the amount and quality of information disclosed by SWFs will go a long way towards accommodating this need. However, the issue of transparency and public disclosure is much more nuanced and complex than may appear at first glance.

While there is considerable room for SWFs to increase their overall levels of transparency, utmost care must be exercised when determining exactly what information can and should be disclosed. When the political establishment in the West demands more transparency from the SWF community, the latter should respond by requesting that these demands be couched in much more detailed and concrete terms. Exactly what information would be required and why? At what level of detail? What would be the frequency and timing of such disclosure? Who are the intended recipients of this information and how do they propose to use it?

Once the discussion gets down to this level of detail, some interesting—and much more fundamental—questions will emerge, starting from the most basic one: how exactly should we define transparency when dealing with SWFs? Can there be one common set of rules that would equally apply to funds which are very different in nature? And what happens if the level of transparency of a particular fund is perfectly adequate for domestic purposes, but is found lacking by international standards? Can such a fund “square the circle” by opening up to international organizations and national regulators in recipient countries on a confidential basis, thus avoiding full public disclosure?

The SWF debate is happening at a time when regulators and market commentators are pondering very similar issues as they relate to other parts of the financial system, namely, private equity and hedge funds. There are indeed some interesting parallels. Consider, for example, the following quote from George Soros’s testimony to the US Congress back in 1994, in which he deliberates on the question of transparency and disclosure by hedge funds:

These two distinctions between supervision and regulation, and between information gathering and disclosure go to the heart of the matter not only in the hedge fund debate, but also with regard to SWF transparency. There are two reasons for this.

First, each SWF operates in its own country under a set of rules which are considered optimal by domestic standards. Whatever level of transparency and disclosure these funds maintain vis-à-vis local constituents is a matter of sovereign choice in each individual country. However, once these funds expand their investment activity overseas, they may face legitimate concerns from international organizations and national regulators in recipient countries. Arguably, all SWFs would be sympathetic and sensitive to such concerns and would be prepared to engage and work with their counterparts in recipient countries in a meaningful and productive way. It is when increased transparency is automatically equated to increased disclosure of information to the general public—above and beyond what may be considered optimal for domestic purposes—that these funds might find it politically difficult to accommodate such demands.

Second, increased transparency and public disclosure may carry with it a real economic cost. This is what George Soros alluded to when he referred to “unwarranted price movements”. Whatever one might think of hedge funds and their role in the financial system, the one sin of which nobody can accuse them is harbouring a political agenda or being motivated by non-economic considerations in their investments. In terms of their goals and objectives, hedge funds are the ultimate money-making machines: seeking maximum return with minimum risk, preferably within the shortest possible time horizon. If these institutions find it advantageous to operate with minimum levels of transparency and public disclosure, then surely they do it for no other reason than to maximize profits and protect their economic interests.

Of course, SWFs are different from hedge funds in many respects, and one should be careful not to draw too many direct parallels when it comes to formulating a policy response.Footnote 26 But the point still stands: if some SWFs appear reluctant to embrace full transparency and public disclosure, it may have less to do with their political agendas or any strategic considerations, and more to do with legitimate economic and financial concerns (e.g. protecting the fund from other market participants gleaning sensitive information about its investment strategies and trading positions, and profiting at the fund’s expense by front-running it).

Some people might argue that what is natural and appropriate for a private sector agent may not be the case for a public institution: if a fund is managed by the government in the interests of the general public, then surely full transparency is always and unequivocally a good thing. But consider the following counter-arguments.

First of all, SWFs are typically set up to provide better investment returns, broader diversification of national wealth, and more efficient management of public finances. In fact, it is one of the key tenets espoused by Western critics of SWFs, who advocate, in no uncertain terms, that any sovereign fund should be run on purely commercial terms. If that indeed is the case, then it would be perfectly natural and appropriate to apply to SWFs the same principles that govern the investment behaviour of private sector agents.

Second, even in the non-profit-orientated part of the official sector, transparency can sometimes be harmful or even potentially lethal. Consider the Northern Rock debacle of 2007, which produced the first run on a British bank in 140 years. While the lessons of this incident are still being digested by both the authorities and the general public, one particular point made by Mervyn King, the Governor of the Bank of England, in his statement to the House of Commons Treasury Select Committee is especially relevant to this discussion. Specifically, he said that he would have preferred to offer covert aid to the troubled bank. However, according to King, the EU Market Abuse Directive prevented the Bank from doing so, as it would have required public disclosure of any such covert support.Footnote 27 Instead, the authorities were forced to act overtly, which sparked worries about Northern Rock’s solvency and led customers to line up to withdraw deposits, thus precipitating the crisis and arguably increasing the eventual fiscal costs to the taxpayer.

In working with SWFs on what would later come to be known as the Santiago Principles, the team at the IMF recognized early on the complex and multifaceted nature of the term “transparency”. In a document called Sovereign Wealth Funds—A Work Agenda, published in February 2008, the IMF team suggested the following framework:

Interestingly, the term “transparency” does not appear once in the actual text of the twenty-four Generally Accepted Principles and Practices. Some principles do invoke explicitly the requirement for “public disclosure”, but, as the IMF team had anticipated, these principles typically refer to various aspects of the funds’ objectives, organizational structure, and institutional arrangements. With regard to investment portfolios, only the most general descriptions of overall investment policy, policy on exercising shareholder rights, and risk management are mentioned in the final document. Relevant statistical data on the funds, their assets under management, and details of investment performance are all duly noted, but not in the context of public disclosure. Instead, they are mentioned only with respect to timely, accurate, and regular reporting to sovereign owners of the funds.Footnote 29

As policy-makers and regulators tackle the issue of transparency, it is important to think through any potential unintended consequences that may result from their actions. For example, in a situation where fund stakeholders may not have yet reached the required levels of financial sophistication and understanding of markets, public disclosure of negative investment performance during difficult economic times may lead to unnecessary and unhelpful short-term political pressures that may overwhelm an otherwise appropriate and sensible long-term investment strategy. In a situation like this, the public good that results from an increase in transparency can be completely negated by the resulting suboptimal changes to portfolio composition and long-term investment strategy.

The lesson contained herein for all proponents of more transparency and tighter regulation in financial markets is that sometimes it may be more efficient and beneficial to all parties involved to have less transparency and more flexibility. Therefore, in designing regulations and putting together and refining codes of best practice for SWFs, it is important to recognize that increased transparency may come at a cost, both to the funds themselves and to the financial system overall. At the very least, this calls for thinking long and hard about the exact definition and scope of the term “transparency” when applied to SWFs.

VI. Definitional challenges with non-commerciality and reciprocity

The other two typical concerns surrounding SWFs are their potential for non-commercially motivated behaviour and lack of reciprocity in foreign investment regimes. As with transparency, the effectiveness of any regulatory response will depend in large part on the proper understanding of definitional nuances and the appropriate use of both terms.

Non-commercial motivation—or “extra-financial” considerations—is part and parcel of the investment management landscape. This is nothing new; we have lived with many examples of such investment behaviour for decades. For instance, trading activity of central banks and monetary authorities intervening in the foreign exchange (FX) markets has nothing to do with profit expectations or narrowly defined financial risk management. In fact, many active currency managers in the private sector openly cite such interventions as one of the many sources of alpha available to a profit-maximizing agent in FX markets. Clearly, when central bankers trade in financial markets in such instances, they are making financial decisions based on non-commercial motivation related to their policy objectives.

In a similar vein, long-term institutional investors who practise “responsible investment” or apply ethical screening, which may or may not be motivated by religious beliefs, engage in a non-commercially motivated investment practice. Some of the largest public and private pension plans, notably in the US, have been known to divest from entire stock markets based on considerations of an openly political nature, such as the lack of a democratic government or deficiencies in labour laws.

How can one seek to limit, let alone outlaw, non-commercially motivated behaviour by SWFs without applying the same principle to other institutional investors? Surely, such actions would go against the key principle of the equitable and non-discriminatory treatment of all investors. But even more fundamentally, non-commercially motivated investments in the majority of cases can actually be good for financial markets, because the idiosyncrasies introduced by such behaviour increase the diversity and heterogeneity of markets.

All policy-makers and regulators will agree that any attribute which helps enhance and deepen the liquidity of financial markets should be considered highly desirable. Market liquidity, in turn, is a function of three attributes: size (market capitalization), activity (market turnover), and diversity (market agents with different risk/return preferences, investment horizons, liquidity needs, etc.). Any extra-financial motives that add to this diversity will actually strengthen the market, not weaken it.Footnote 30

One of the key issues here is the exact definition of “non-commerciality”—it may just be too broad in scope to be helpful in applying to SWF regulation (or self-regulation). Perhaps the best way forward would be to replace such expressions as “non-commercially motivated investment” or “extra-financial considerations” with a much narrower and more precise term. Perhaps narrowing it down to “political motivation” would be a good start, although this would not resolve the dilemma arising from the above-mentioned actions of US pension plans divesting from overseas markets for political reasons. Maybe an even narrower term like “geopolitical motivation” would be more appropriate, but it may still be a challenge to come up with a universally acceptable definition for such a nebulous notion. Whatever the case may be, one thing is certain: the legal community needs to address the definitional challenges surrounding SWFs as part and parcel of any serious debate about whether and how sovereign funds should be regulated.

Reciprocity is a similarly difficult concept to define precisely. Does it mean an asset swap of roughly equivalent strategic and market value? Is it a quid pro quo opening up of specific sectors of the economy based on bilateral agreements? Or does it mean signing up to a multilateral agreement, which sets out just the basic rules and principles for achieving a level regulatory playing field for all investors? In the public debate about reciprocity one can find references to all three different interpretations.

The positions of policy-makers and commentators in recipient countries are also starkly different. The US and the UK have taken a fairly relaxed approach to reciprocity. Even as they insist on resolving matters of transparency and non-commerciality, their position on reciprocity is that while it would help, it should not be made a condition to allow SWF investments in their economies. This view was clearly articulated in the article “Public Footprints in Private Markets” by Robert Kimmitt, then Deputy Secretary of the US Treasury: “It is in the United States’ interest to be open to market-driven investments—from both private and sovereign entities—even if other countries are not.”Footnote 31

But while such an open and relaxed position is firmly supported by sound economic arguments, politically it may be very difficult to sell to politicians and the general public. A special report on the EU policy response to sovereign investments published by the Centre for European Reform (CER) provides a very good background on the state of the debate in continental Europe:

The CER publication then proceeds to argue that such hard conditioning of SWF access on reciprocity can be very difficult to define, implement, and measure, that it may create more problems than it would resolve, and that a logical and internally consistent view on reciprocity needs to be formed within the EU before it can be applied meaningfully outside the Union. Unlike policy discussions around transparency and non-commerciality, there doesn’t seem to be much consensus among the recipient nations in their debate about reciprocity. However, in many respects the root cause of this disagreement might be similar to the policy risks associated with transparency and non-commerciality: the lack of a clear and universally accepted definition of what exactly is meant by these terms can lead to policy mistakes in all three cases.