The importance of holding liquidity is well understood by large institutions. But how much is enough? In the first part of Aviva Investors' new article series on liquidity optimisation, Alastair Sewell investigates the key considerations for different investor types. Read this article to understand: the importance of sizing a liquidity portfolio correctly, and why institutions must consider “known” and “unknown” liquidity needs, the factors pension schemes, insurers, corporate treasuries and public institutions need to consider when sizing liquidity pools.
The amount of liquidity investors need to hold has come into greater focus in the wake of recent market events and regulatory changes. Regulators around the world are asking investors to hold more liquidity, or placing greater emphasis on existing rules.
For many investors the simple fact is they need to hold more liquidity. Acknowledging this is all well and good, but determining precisely how much liquidity is needed can be a complex undertaking (and holding higher levels of liquidity can bring its own challenges in the form of potentially reduced allocation to risk assets).
This article is reproduced with the kind permission of ICMIF Supporting Member Aviva Investors. For more information click here.
Author: Alastair Sewell,
Published August 2024
Taking a holistic view
Broadly speaking, investors tend to think about liquidity either as part of a broader strategic asset allocation, or as a specific and separate pool. For example, investors using derivative-based investment strategies will typically need to hold dedicated collateral pools against their derivative positions.
When thinking about liquidity from a strategic asset allocation perspective, getting the balance right counts; or, to put it another way, it is important to take a holistic view of liquidity. An investor with a significant weighting to illiquid assets may well need a higher liquidity allocation than an investor with a higher allocation to liquid markets portfolios. Liquidity needs to be valued appropriately as part of a broader strategic asset allocation.
This article looks at liquidity requirements from a bottom-up perspective. For many investors the practical reality of the overall strategic asset allocation framework may cap the amount of liquidity they are willing to hold, given broader portfolio objectives. There is a balance to be struck. When yields on liquidity instruments are high in absolute or relative terms, this balance is, naturally, easier to achieve than when yields are lower.
What is liquidity?
Focussing on market liquidity, we can understand liquidity both from an asset and market perspective; we must also consider timing. To put it another way, is asset X liquid in market Y at time Z? All three parameters can, and of course, do, vary, and can only be known after the fact. Until I have sold something, I do not know if I can sell it, and if I can, at what price. I will, of course, have a different level of confidence in my ability to do so depending on the asset, the market and prevailing conditions at the time.
Sizing liquidity portfolios
Sizing liquidity portfolios is simple in some ways, as regulatory requirements prescribe minimum levels in many jurisdictions. While this may simplify things in principle, it disregards the fact that investors’ liquidity needs will be highly varied. Regulatory parameters set the baseline; investor-specific requirements set the reality.
We typically discuss the following three questions with our clients to identify how much liquidity they (really) need:
- How often do I expect to experience liquidity calls of differing levels of materiality?
- What is my worst-case liquidity need?
- For how long might I experience liquidity stress?
Alas, these questions are more easily posed than answered. A structured approach is required to yield an optimally sized liquidity portfolio.
Figure 1: Known versus unknown liquidity needs
Source: Aviva Investors, April 2024.
Start with the past
The starting point in determining the appropriate size for a liquidity portfolio will almost always be the “knowns”. Investors know the cash demands they have experienced in the past; they may have expected cashflows they need to fund periodically in the future. For example, a pension fund will have a good idea of the disbursements it needs to make to pensioners, while corporates will estimate dividends they will fund from liquid assets. In combination, these indicators can provide a good indication of baseline liquidity needs.
However, as we all know, the past is not a guide to the future. A good example of how this is relevant to sizing liquidity portfolios comes from the insurance industry: consider the fact there is significant variance in weather-related claims in any given year.
Establishing a liquidity portfolio based on prior averages may therefore materially understate the actual liquidity demand (assuming liquidity needed to fund claim payouts in this case) in any given year. Furthermore, the trend is increasing (see Figure 2). Therefore, a liquidity portfolio sized on the past will not be adequate for the future, unless the trend reverses.
Figure 2: Growth in global natural catastrophe insured losses (US$ billion, 2022 prices)
Source: Swiss Re Institute, 2023.1
Build scenarios and stress testing
Because history provides only a partial guide to potential liquidity challenges, investors must therefore build scenarios to account for the “unknowns”. Our clients will typically build a range of scenarios. These may be based on information such as:
- “Internal” scenarios: For example a mutual fund planning for the possibility its largest investors may simultaneously redeem.
- Documented liquidity events experienced by peers: For example, a bank assuming it incurred a proportionately similar deposit withdrawal to SVB, or the “wrong way risk” experienced by pooled liability-driven investment (LDI) schemes in the UK in 2022 as collateral portfolios fell in value in lockstep with the leveraged assets (gilts).
- Parallels drawn from other sectors: For example, an insurer building a lapse event scenario informed by the large-scale withdrawal of deposits at SVB.
When it comes to stress testing scenarios, this process broadly assumes one of two approaches. In the first, there is some level of liquidity need; if this need can be met, the test passes. The second, or reverse stress test, estimates how much liquidity could be provided before all liquidity is exhausted. The latter can then feed into a liquidity governance process in determining the appetite for a given level of liquidity risk. These stress tests will typically form part of any scenario generation exercise.
Bringing it all together
A combination of “knowns”, “unknowns” and stress tests will provide a range of liquidity needs, which can inform assessments of the size of liquidity pools required. From these, a judgement must ultimately be made to determine the right level, based on the various inputs.
Too much liquidity may have broader consequences, as equally could too little. As the Chinese proverb says: “just as water keeps the boat afloat, too much water will sink it”.
In striking the right balance investors should consider:
- The plausibility of the scenarios involved.
- The severity of the scenarios.
- The broader effect of holding a given level of liquidity. This is particularly acute when cash yields are low in absolute terms or relative to risk assets; or more prosaically, if the liquidity holding crowds out other activity or investment.
- Wider considerations, including other liquidity sources (recognising that such sources may not necessarily be immediately accessible).
Other liquidity sources might include intra-group transfers or, for pension funds, a cash injection from the sponsor. However, the governance processes associated with these changes may be relatively involved or slow and thus not suitable for an immediate liquidity event.
An investor holding a liquidity portfolio sized on worst-case historic or modelled outflows will have allocated more to it based on the severity of potential scenarios rather than their plausibility. Conversely, an investor using averages will have sized their portfolio for less severe, but more plausible outcomes.
For a practical example, consider the insured losses due to natural catastrophes indicated in Figure 2: a liquidity portfolio sized using averages would have been largely aligned to the size of losses in most years, although in several years (2005, 2011 and 2017) actual losses materially exceeded the trend level (and indeed, prior averages).
An important mitigating factor in insurance claims is of course the time between the claim being registered and then subsequently paid out. Time enables liquidity provision, and it is therefore another of the “wider considerations” to factor into the debate on sizing the liquidity pool.
The regulatory perspective
The amount of a liquidity an investor determines in the exercise detailed above must be considered relative to regulatory requirements (where applicable). Regulators vary in how prescriptive their rules are for liquidity. For open-ended funds, for example, overarching guidance is principles based, whereas as we have seen above, rules for MMFs are highly prescriptive.2 Regulation will also be tailored to the nature of the investors.
Where regulation does apply, this will set a base-line level of liquidity an investor must hold. Most investors will target a buffer above the regulatory minimum, informed by the requirements of their activities. Where there is no specified liquidity regulation, this does not of course elide the need for a prudent approach to liquidity management.
Investor type considerations
Pension schemes
Some pension schemes may need to hold relatively little liquidity. Think defined contribution (DC) schemes, or open (unlevered) defined benefit (DB) schemes. Indeed, the European Insurance and Occupational Pensions Authority calculated European occupational retirement schemes held just four per cent of their assets in cash and deposits and a further two per cent in MMFs, implying an overall allocation of around six per cent, or a collective liquidity pool of €122 billion.
To keep this in context, however, European pension schemes do also maintain significant government bond allocations, which should be liquid in most market circumstances (although the UK’s gilt-market dislocation in September 2022 provides a counterexample).3
On the other hand, closed DB schemes have materially different liquidity requirements. First, they are cashflow negative – they must have liquid assets available to service distributions to pensioners. Second, their funding position will be material to the way they allocate assets. Schemes in deficit will need to close funding gaps; funds in surplus may be able to achieve their objectives with a lower risk – and potentially more liquid – asset mix. Third, if using LDI, they must maintain liquid collateral buffers against potential margin calls. Here, rules become materially more prescriptive. For example, the UK pensions regulator stipulates that LDI schemes maintain an operational buffer of 100 basis points (bps) and a market stress buffer of 250bps, for a total of 350bps.4
Insurers
On the insurance side, regulatory requirements typically focus on stress testing. These will be the responsibility of the insurer in question to define, taking into consideration the profile of their businesses and a range of stress scenarios. In some cases, more specific tests may be defined, such as on lapse risk for life insurers.
Figure 3: Selected regulatory guidance on insurer liquidity stress scenarios
Source: Aviva Investors, February 1, 2024.
Corporate treasuries
Corporate treasuries will typically not be subject to liquidity regulation. They will, however, be subject to varying liquidity requirements depending on the nature of the business. In some cases, these will be predictable at a high level – for example, businesses with clear funding and disbursement profiles or distinct seasonal patterns. In others, the requirements may be predictable based on corporate actions – think dividend payments or bond coupons or maturities.
In other cases, calls on liquid assets may be sudden and not easily predicted. For example, hotel occupancy rates fell to zero extremely rapidly during the COVID-19 pandemic, leaving hotel operators with no cash inflows, while they still needed to pay out (minimised) maintenance costs on their properties. In other words, an unpredictable yet material liquidity event occurred. More commonly, a rapid decision on an M&A opportunity may demand rapid liquidity provision.
For corporate treasuries, therefore, the sizing of a liquidity portfolio must consider both the normal liquidity requirements of the business and the potential for major stress events.
Public institutions
Within public institutions, liquidity risk management is decidedly nascent. The World Bank found almost 40 per cent of respondents to its annual reserve manager survey do not measure or monitor liquidity risk.9 This reflects the fact these institutions will have materially different objectives and governance to other investor types, not to mention investment horizons extending well beyond those of most investors.
A major environmental incident or financial market development, for example, may trigger a sudden, significant requirement for mobilisation of reserve resources. Similar to corporate treasuries, liquidity regulation will be limited at most, although perhaps expressed in constituting requirements.
Conclusion
Sizing liquidity portfolios appropriately is ultimately a function of synthesising competing claims. On the one hand, regulatory requirements may set a floor on the amount of liquidity an investor must hold. On the other, the requirements of a given investor – both in the normal course of business and in “event” scenarios – will drive liquidity demand. And lastly, a top-down strategic asset allocation may force a specified liquidity allocation.
Integrating these competing requirements requires good governance. In our view, this is underpinned by the quality of the information available. A clear appraisal of the facts and robust scenario analyses can combine effectively with a risk appetite decision to guide an optimally sized liquidity allocation.
But the process shouldn’t stop there. As we will discuss in the next article in our liquidity optimisation series, constructing a liquidity portfolio entails a complex set of decisions. We believe there are significant opportunities for investors to optimise a liquidity pool of any size.
References
1. Chandan Banerjee, et al., “A perfect storm: Natural catastrophes and inflation in 2022”, Swiss Re Institute, March 23, 2023
2. “Revised policy recommendations to address structural vulnerabilities from liquidity mismatch in open-ended funds”, Financial Stability Board, December 20, 2023.
3. “In brief: How do IORPs invest?”, EIOPA, 2023.
4. “Using leveraged liability-driven investment”, The Pensions Regulator, April 24, 2023.
5. “Supervisory statement | SS5/19: Liquidity risk management for insurers”, Bank of England Prudential Regulation Authority, September 2019.
6. “Consultation paper: Proposed enhancements to the regulatory regime and fees for commercial insurers”, Bermuda Monetary Authority, July 28, 2023.
7. “Insurance Authority welcomes passage of the amendment bill for the implementation of the Risk-based Capital regime”, Insurance Authority, July 6, 2023.
8. “Letter to : Chief Executives of authorized insurers”, Insurance Authority, Ref: INS/TEC/6/20/6, August 6, 2018.
9. “International reserves remained resilient in a historically volatile year, says new World Bank report”, The World Bank, November 2, 2023.
