Good afternoon. When I was invited to speak at today’s conference, I was asked if I would like to talk about the systemic risk posed by investment funds to the financial system, and the regulatory changes that might be proposed to limit these risks.
These are of course very interesting and important areas of debate at the moment. And the FCA is actively engaged in the thinking. Both because we supervise the world’s second largest fund management market and a world leading financial centre in the UK; and because the FCA is a member of a number of UK and international bodies looking at these issues.
To be up front, however, I will say at this point that the regulatory work is not finished. We are still assessing the impact of new post-crisis market conditions; the risks these may pose to funds and fund managers; and vice versa.
It will only be once these issues have been considered that we will be able to think about what regulatory steps could be taken. I note that some possible steps may in turn have consequences we have to think through too. Therefore, I won’t seek to provide definitive answers today as to what regulatory changes might lie in store.
But with that said, I’d like to do three things in my remarks today:
- First, set out some background to this debate;
- Second, talk about market and fund liquidity, and work that has been done to understand today’s conditions, including by the FCA; and
- Third, give some views on the investor perspective on the issues.
It is this last point that I think has been missing from some of the debate to date, and worth pausing more on, to ensure markets work for investors.
Background to the debate
History
Starting at the beginning, what is this debate all about?
Well, after 2008 the world was focused on the systemic risks embodied in large global banking groups. And rightly so. The case barely needed demonstrating that such institutions, if they were to suffer large losses fail, could fail, and potentially bring wide and serious implications for other firms and financial markets. This was directly observed in 2007 and 2008.
And with that case made, the focus turned to potential solutions: additional capital; bail-inable securities; structural separation; more stringent stress-testing; greater disclosure; more effective resolution procedures, and more. All possible measures to address the ‘too big to fail’ problem in banks.
While that debate was still playing out, questions started to be asked about the systemic nature of other financial actors (so called non-Bank SIFIs and G-SIFIs). Insurers were next in line, particularly given the lessons of AIG’s losses.
Then market infrastructure took its turn – CCPs in particular. Clearly a critical part of the financial system, but a different business model that required different thinking from banks and insurers.
And in the last two years, the discussion has intensified around the role of asset managers, and others controlling private pools of capital for investment purposes, including funds and collective investment vehicles as separate legal entities. (The potential non-bank, non-insurer G-SIFIs). Our focus today.
A focus on asset managers
One reason asset management, and in particular the collective or mutual fund industry, has attracted attention is that it is a large and important part of global capital markets. The IMF estimates that the asset management industry intermediates assets amounting to US $76 trillion – 40% of global financial assets in 2015. And there has been a significant increase in investment in mutual funds and other open-ended fund vehicles in recent years. ICI Global reported in 2014 that the global mutual fund industry’s assets have grown more than sevenfold in the last two decades.
Fund management plays a vital economic function in being one link between those with money to invest (with the aim of achieving growth or future income) and the companies and governments who need capital for investment, expansion or funding their ongoing operations.
Of course, they are not the only and not even the largest source of “buyside” involvement in the market, where you could also note: direct institutional investment; pension funds; insurers; sovereign wealth funds; and others who also make up large parts of the market. As at the end of 2012, open-ended investment funds accounted for US$26 trillion of assets under management globally, or 11% of global assets.
A second reason for interest, as some will no doubt remember, is that there have been some high profile failings of investment funds, like Long Term Capital Management in 1998 and the recent closure of Third Avenue’s Focused Credit Fund. Both are interesting case studies.
A third reason might be the large and growing investment in collective investment schemes, including by retail participants – I’ll come on to this point towards the end of my remarks.
Funds or fund managers?
The Financial Stability Board (FSB) has been the main international forum coordinating the regulatory analysis and discussion of this subject, working closely with IOSCO. In 2015, the FSB and IOSCO consulted on a set of proposed methodologies for identifying non-bank non-insurers who should be considered systemically important financial institutions, including both investment funds and asset managers or investment advisers.
A typical structure might see the manager and fund company being separately owned and run, with a manager acting purely as an agent, executing investment mandates, and those controlling funds managing the fund operations. The proposed methodologies aimed to identify fund and managers whose distress or disorderly failure could cause significant disruption to the global financial system. While the FSB/IOSCO paper acknowledged that funds and asset managers could be seen as systemically important independent of each other, it also took into account the links that exist between the two. The proposed methodology for undertaking an assessment of which particular funds may be systemic included five different impact factors: size, interconnectedness, complexity, substitutability and cross jurisdictional activities.
Following the consultation, discussions led to the FSB and IOSCO jointly agreeing to put on hold the entity-based assessment methodology work until further analysis on potential market-wide activities-based solutions that would be applied in the first instance could be identified.
And this is where the discussion now stands.
Potential sources of risk
So, let me move on to spell out in more detail the potential vulnerabilities which the FSB and IOSCO work has identified in its assessment to date, which might lead to systemic risks among these firms.
Well, the FSB and IOSCO have focused their attention on five potential structural vulnerabilities of funds:
- The possible mismatch between the liquidity of fund investments on the one hand and the terms and conditions for the redemption of fund units on the other;
- The high levels of leverage within (certain) investment funds;
- The operational risk challenges in transferring investment mandates from a fund manager in a stressed condition to another manager;
- The securities lending activities of asset managers and funds; and
- Finally, the potential vulnerabilities of pension funds and sovereign wealth funds.
If markets are stressed, or idiosyncratic firm conditions materialise, then each of these vulnerabilities could lead to losses for investors in the fund with potential impact on others firms or the market.
In terms of the leverage risk, banks who provide funding or act as counterpart in derivative transactions are arguably more protected than in the past from asset management clients and counterparties. Improving risk management, sufficiency of bank capital and other prudential measures, as well as central counterparty clearing all mitigate the risks of bank exposures. Further, a large number of regulated funds are not permitted to utilise leverage or have limits placed on the amount of leverage they can obtain. We monitor those who can leverage themselves, such as hedge funds, previously through the FCA’s annual hedge fund survey[2] and more recently through AIFMD regulatory reporting.
But, in addition to risks for individual players, there are systemic concerns which arise from the impact of any losses on counterparties, and from the impact on broader asset prices if redemptions were to force a rapid sell off of positions and falling market values. This market impact is certainly worthy of further consideration.
If some intermediaries have to rapidly liquidate a substantial amount of illiquid assets that they can no longer fund, it is certainly possible that there could be declines in asset prices. This may in turn lead others to sell out of positions, or trigger margin calls for transactions that are linked to the value of the assets. And then other market participants could add to the price deterioration by also selling assets, in self-reinforcing cycles. With correlated trades, market inter-linkages and multi-asset investment strategies, such selling pressure may not be contained to a single asset class. Of course, if markets become wholly one sided, there could be issues with discontinuity of prices, as seen in the darkest days of the financial crisis.
You can see why each of the five vulnerabilities I’ve listed are indeed, in theory, vulnerabilities. And it is possible to imagine the types of impact from them crystallising.
But it is really important for us to understand how likely it is that each scenario might arise, and if it did, how big the impact would actually be. We also need to reflect on what already is in place to mitigate the probability of each risk crystallising, and to cushion the impact if it were to. It has precisely been the aim of both UK domestic and international authorities to develop and deepen that understanding.
What work is being done to consider these risks?
The FSB has been working since early 2015, firstly to explore the financial stability risks associated with stressed market liquidity and asset management activities, and then to assess the potential structural vulnerabilities associated with asset management activities.
In September, the FSB signalled some of the risk vulnerabilities they’ve been looking at, as I’ve just noted. As an interim step, the FSB has encouraged the appropriate use of stress testing by funds to assess their ability individually and collectively to meet redemptions under difficult market liquidity conditions.
The FSB is now working alongside IOSCO to conduct further analysis and, as necessary, will develop policy recommendations in the first half of this year. The FCA is playing its part in these discussions, which are seeking to both understand the risks and propose proportionate and effective solutions as necessary.
Additionally, on the IOSCO side, it is worth noting that IOSCO has already previously published some authoritative principles of Liquidity Risk Management for Collective Investment Schemes, which are relevant here. And IOSCO is also looking into questions around liquidity in corporate bond markets, and has flagged this as an important potential risk in its annual risk report.
And in the UK, the Financial Policy Committee (FPC), a committee of the Bank of England which the FCA participates in, and which looks at potential systemic risk issues for the UK, began work last year considering similar points. The Committee considered three channels through which the activities of open-ended investment funds might exacerbate large-scale asset sales and lead to market disruption: first-mover advantage, where the incentive is to redeem ahead of fellow investors when market conditions take a down-turn; pro-cyclical behaviour by end-investors and fund managers, particularly where poor fund performance is reported; and thirdly, the use of leverage, given its potential to create larger losses, and greater volumes of selling to meet redemptions. The Bank of England’s December Financial Stability Report notes that after surveying certain funds, the risk around first-mover advantage appeared minimal due to swing pricing and other processes to ensure that remaining investors are not unduly disadvantaged. The other channels remain ones of potential concern, and the FPC supported the Bank’s intention to incorporate the activity of investment funds into system-wide stress testing to assess the resilience of markets to large-scale fund redemptions.
So, 2016 will see continued exploration of fund and market liquidity issues by the FSB, IOSCO and the UK FPC. While that work continues, I will not try and speculate what the conclusions might be, nor what further steps may be needed to help protect against the key vulnerabilities and risks.
Fund liquidity
So that is the broad shape of the international regulatory work. Let me turn now to two pieces of work which the FCA has done to contribute to the analysis of the first of the vulnerabilities. Liquidity mismatches in investment funds.
At the beginning of March, we published the findings of some recent FCA supervisory engagement with a number of large investment management firms[3], where we looked to understand how they manage liquidity in their funds. This sought to inform our understanding but also highlight good practice and flag risks to the wider industry sector in the UK.
One concern is that funds are often marketed as offering, and indeed do offer, daily prices of the value of the units of the fund. So there is an expectation on the part of investors that they will be able to exit the fund immediately at that daily price. In practice, funds may have to sell assets to meet redemptions. And the question is whether those assets can be sold quickly to pay that redemption, and if they can, at what discount to previously traded prices.
In deep and liquid markets, of course, it is assumed that sales of assets can be made with no impact on market prices, including in relation to large blocks of assets. If markets aren’t deep and liquid, prices may fall as managers seek buyers and their intention becomes known to the market. Things will depend on both the characteristics of the assets and market conditions at the time, but often there could be some mismatch between fund unit liquidity and the liquidity of securities.
To address this risk, the EU rules which apply to funds in this area (UCITS, AIFMD, and so on), require good liquidity risk management across a variety of market conditions, so fund managers can deliver to investors what they have promised. They also allow fund managers under these regimes to utilise certain day to day and exceptional liquidity management tools.
As part of the recent work with investment management firms, FCA supervisors looked at adherence to these rules and found some areas of good practice in this space. Based on these findings, the FCA raised with the broader industry positive steps that can be taken including:
- Clear disclosure to investors upfront about liquidity risk;
- Subscription and redemption processes that are designed with the fund’s investment strategy in mind;
- Regular assessments being made of liquidity demands;
- Gathering price data from varied sources; and
- Stress testing of portfolios and possible redemptions in light of a range of market conditions.
There was also good practice in terms of redemption policies. For example, recognising the interests of all investors, not just those exiting, via readjustments of portfolios following large redemptions to ensure remaining investors don’t retain only illiquid assets.
We were pleased to find good practice in this space, and hope by publishing results, this may be helpful to other managers. The FCA will continue to look at these issues, because we feel they are really important for the fair treatment of investors and market integrity as a whole.
Market liquidity
Fund managers will no doubt have dealt with and considered redemption policies since the very first mutual funds were marketed in Amsterdam in the 18th century, but the genesis for the current discussions has been the change to market conditions in recent years.
Many people in the industry argue that changes to financial regulation, following the 2008 financial crisis, have reduced market liquidity. First among those reforms, it is claimed, are changes to prudential rules (liquidity coverage and leverage ratios) and capital requirements, which have increased the charge that falls on an institution holding assets like bonds on balance sheets. While clearly designed to ensure that firms hold more capital to meet liabilities if assets do not perform (a necessary requirement, given the lessons of the crisis), there is no doubt that all things being equal it will cost institutions more to hold the same assets today than it did 10 years ago in terms of capital.
In OTC markets, buyside firms have perhaps previously placed some reliance in the past on dealers holding large stocks of tradable assets, warehousing risk and making continuous markets for those wishing to buy and sell. Some have described this market-making role as being a shock-absorber against volatility in corporate bond markets.
Research shows that the amount of corporate bonds held on balance sheets has gone down. For example, FCA regulatory returns suggest UK dealers held around £400bn of corporate bonds on their trading books in mid-2008, but just £250bn at the end of 2014.
Of course, it is important in any analysis of where we are today not to look back at the pre-2008 period as a golden age that we need to get back to. That is definitely not the right benchmark. Secondly, it is not clear that current liquidity conditions have deteriorated to a point where serious structural issues have arisen (for example, excessive volatility, or prices not reflecting the fundamentals of supply and demand, or a discontinuation of markets). Nor, if there has been any deterioration, what role financial regulation has played. Potentially there are other drivers.
Arguably, certain issuances of corporate bonds have always had thin liquidity, given the nature of the asset and lack of standardisation. And of course, it goes without saying that certain bonds are issued in greater volumes and are more frequently traded than others, and that this can vary significantly between different jurisdictions. This hasn’t previously had the consequences that some claim such market conditions bring. The FSB in its consideration of corporate bond market liquidity noted the “need to recognise that a reduction in banks’ trading activities was an intended objective (of the post-crisis reforms), and that the unsustainable excess liquidity prior to the crisis should not be used as a baseline for comparison.”
As regulators, we should be forward-looking, not backward-looking, thinking about how markets will evolve.
None of this is to say that there are no risks around market liquidity. Far from it. And as regulators, we should be forward-looking, not backward-looking, thinking about how markets will evolve.
FCA economists’ research
Market liquidity is a difficult concept to measure. Many have noted the change to market structure over the past few years, and have perceived via certain proxies a change in market depth. These have been tinged with a sense of pessimism about the state of the market. But perception may or may not accurately gauge what the actual price-impact could be for parties seeking to sell blocks of assets into the market.
Independently, two FCA economists, Matteo Aquilina and Felix Suntheim, have been analysing transaction data available to the FCA to see what these data can tell us about market liquidity in the corporate bond market. They looked at transaction data available to the FCA on corporate bond transactions that took place between 2008 (just before the crisis) and 2014, and considered different potential measures of market liquidity.
I won’t go into full detail about the findings and results today – their paper is available on the FCA website[4], and I recommend this to you.
But to note, the authors paint a sunnier picture than most about liquidity conditions in this period, finding that the fall in bond inventories led to no corresponding drop in liquidity in the UK. This of course is a finding only of the authors themselves and not the FCA, but is itself an interesting evidence-based contribution to the debate. And it supports other authoritative findings too. In fact, they found aside from a relatively brief period immediately after the crisis, illiquidity has actually been decreasing to very low levels.
There are some possible explanations given in their paper, but equally, there are some caveats too. These include that the data does not cover the most recent period, nor the pre-crisis years. And they obviously only capture the liquidity conditions of trades that took place - some trades were perhaps considered and didn’t take place because of real or perceived issues with the liquidity of markets. The post-crisis period has also been one where we’ve seen low to medium levels of market stress against historic norms – greater stress may lead to different results.
As the authors wisely note, it may be some time before we have a definitive conclusion.
While this particular research does not show that corporate bond market liquidity is now less than in the past, it remains a plausible risk that regulators should be considering. If it were to materialise, it could make management of redemptions harder than in the past. If that’s the case, then the scenarios of self-reinforcing cycles of selling are at least plausible, although the likelihood is not clear.
But does this really matter for investors and what role do they play in the debate?
The investor angle – expectations, behaviour and disclosure
Investor behaviour
I will start with what we can assume about retail investor behaviour.
The increase in the participation in corporate bond funds has reportedly surged in recent years, alongside a tripling between 2000 and 2013 of the global corporate bond markets. This is clearly driven, in part, by record low central bank interest rates, pushing down savings account rates, and a corresponding ‘search for yield’ among those with investable sums.
Investing implies risk though, and we should presume that investors have some awareness of this, although it probably focuses around risk of loss.
But any reasonable analysis of investor behaviour would also tell us that average investors do not track fund portfolio performance day to day, and do not pull out of such investments at the first sign of loss. This seems to be the experience. They are, on the whole, not professional or day-traders.
Instead, they have longer term horizons, investing regular savings, pension contributions, wind-falls like inheritance, and wishing to get broad exposure to some investment risk to get a greater return for the future.
And as such, they are likely to ride out losses and are unlikely to redeem from funds in response to immediate market news, as professional investors might react in relation to direct exposures to similar assets. Or, even if some do, it would seem unlikely that all investors would run for the exit at the same time. This is certainly historic experience, although of course investors do not always behave rationally or as one would expect, and it is possible that future behaviour may differ from that today.
But you might also expect professional investors – pension funds and insurers for example, to act counter-cyclically, take the long term view, and equally not to ‘cut and run’ at the first signs of trouble. They have long term liabilities, which also benefit from patient investing over equally long periods. Further, such institutions’ investment intentions may see them buying, instead of selling, as assets become cheaper, cushioning falls in prices. More speculative players, such as hedge funds, may also come in with demand to buy as asset prices fall and assets such as bonds become more attractive, in risk-reward terms.
Perhaps we need to reconsider how much market liquidity really is necessary in order for markets to work well for investors.
These would seem to be strong mitigants against the risk of rapid redemptions and firesale prices I noted before. If this is the case, perhaps we need to reconsider how much market liquidity really is necessary in order for markets to work well for investors. As Chris Hitchen, CEO of the Railways Pension Trustee Company said recently, there could be an “irrational desire for liquidity”.
Disclosure
An important angle for a consumer-focused conduct regulator is that funds provide the level of service that is promised and that investors could reasonably expect. Regulators are not there to protect investors from all loss, and we don’t expect portfolios to be constituted at the outset in a way that guarantees their liquidity in all conceivable scenarios – neither of which would be desirable for anyone.
What we do look for is those who run funds to manage their risks well and clearly disclose the risks they run, the types of assets they invest in, their broad investment strategy and time horizons over which they expect to invest, and so on. These disclosures should be clear to the investor before they part with their cash. The Third Avenue fund is an interesting example of this, because its stated investment strategy was investing in very illiquid assets, with the potential for high-yields but the risk of sharp loss. This much seems to have been clearly disclosed to investors. Issues allegedly arose when they couldn't meet redemptions in the manner expected and the fund had to close. Although again there was no noticeable contagion effect.
It is incumbent on the fund and its manager to act in the best interests of all investors in executing the service, at the start of the service, on an ongoing basis and at the point of redemption. To reinforce this, the FCA also put out wider message last year[5] to remind investors of the need to understand their own demands for access to their money, alongside expectations of return, and the risks where corporate bond funds might not be able to deliver. We’ll be checking whether investors, in fact, are aware of, and understand, the potential liquidity risks as part of the Asset Management Competition Market Study[6] we launched in December and will be working on this year.
Outcomes
There is clearly some trade-off between protection from loss on the one hand, and investor returns on the other, that has to be recognised, considering the likelihood of events materialising. If funds take a very cautious approach to being able to meet redemptions (such as holding large amounts of cash), this may make it more difficult to deliver high potential risk-weighted returns.
Interestingly, recent figures suggest that bond funds have doubled the amount of cash they hold compared with five years ago. There is nothing wrong with this, of course, and holding extra cash serves a variety of purposes. But funds need to consider if they are getting the balance right between liquidity and return, as expected by their investors.
It must be remembered that there has been no recent example of a shock large enough to trigger the sort of systemic risk scenario we have been discussing. But, equally, the situations we are thinking about are not unforeseeable and future market conditions could look very different from today.
It would also seem fair to say that given the differences in business models, in investor expectations and in the appetite for losses, many of the solutions that have been rolled out for banks would not necessarily be suitable for asset managers. For example, imposing capital requirements on funds, while a key prudential tool in banking, would not be appropriate since funds are by definition loss-absorbing. Such rules would impose costs on investors without a material benefit.
Getting good investor outcomes is a key consideration for markets that work well.
So, as I have said more than once already, further analysis is needed. My point here is that getting good investor outcomes is a key consideration for markets that work well.
Conclusion
So, to conclude.
It is possible to imagine a set of scenarios in which a trigger event (for example a sudden, unexpected increase in central bank rates) or losses at one fund could cause knock on effects for other firms and broader falling asset prices. There could even be self-reinforcing factors that exacerbate this.
A key vulnerability being discussed is the ability to manage fund redemptions in an orderly way, particularly in the context of post-crisis market conditions, including liquidity. I have said that FCA evidence shows that funds have demonstrated some good practice at understanding and managing both the liquidity of their assets and expected demand for redemption. And the case is far from proven that market liquidity is now at low enough levels to be a dangerous exacerbating factor. Research from FCA colleagues is contributing to the debate on this point.
The picture of the systemic risks is still being drawn through the work of the FSB, IOSCO and others.
Banks, insurers, infrastructures, asset managers and others all interact, and can both exacerbate and mitigate risks and processes within the financial system, in many and complex ways. And don’t forget investors too. Ultimately, we mustn’t forget the purpose of these markets – to serve the needs of investors and it’s that message I’d like to leave with you today.
Thank you.