Multi-firm review of liquidity risk management at wholesale trading firms

Good practice

• A culture of managing risks primarily to optimise longer term performance and resilience, and secondarily to comply with regulatory requirements.
• Dynamically, at the end of each day, identifying and quantifying their inherent forecasted intraday (T0) risk exposures, considering the effectiveness of their BAU and contingency mitigating actions and ensuring that their residual forecasted intraday and inter-day risk exposures remain within the risk tolerances defined by their risk appetite.
• A clearly defined qualitative and quantitative risk appetite, which:
  • clearly set limits and action triggers around how much risk the firm is willing to take
  • is used to set the required levels of resource (financial and non-financial), and
  • is not solely based on adequate financial resources but also on non-financial criteria.  For example, some firms had action triggers based on systems outages, reconciliation failures, the nature of contingent liabilities and the financial position of the overall group
• Ensuring their framework remains fit-for-purpose by frequently:
  • reviewing their framework with regular assurance reviews conducted by either risk oversight or internal audit, and
  • analysing internal and external risk events including near misses and incorporating the findings of their reviews into their framework
• Continuously improving frameworks and operational processes with a primary focus on reducing intraday liquidity risk. For example, some firms had pre-emptively introduced a range of measures into their standard BAU processes to help them withstand instantaneous liquidity shocks.
• A holistic approach to risk management with market, credit, operational and liquidity risk considered in a joined-up manner.

Poor Practice

• Many firms had well designed frameworks but struggled to embed them.
• In many instances, strong frameworks remained reliant on the ongoing sponsorship of one or more senior executives and were not a fundamental part of a strong risk management culture.
• Some firms had not assessed whether their CCP membership category continues to be appropriate for their business model, scale and level of financial resources. An example of this risk crystallised in September 2018 when Nasdaq Clearing AB placed a member of its Nordic market in default. The losses were large and significantly exceeded the margin placed by the defaulter and the CCP’s own capital. This depleted the commodities default fund with the consequence that other members of the market needed to cover the excess losses within two business days. In the event, this sufficed to absorb the loss resulting from the default. However, not alone did the default force a CCP to pass on losses to members. Members were also required to replenish the default fund – the critical nuance of this default is that this extra step surprised some. 

Good practice

• Conducting two forward looking liquidity stress assessments each day: one before banks and trading counterparties close; the other later in the evening.
• Well considered stress assessments typically involved the following steps:  
  • Estimating the next day’s T0 peak BAU and potential stressed liquidity requirements (specifically margin calls).
  • Ensuring liquidity stress models included a broad range of operational inputs and behavioural analysis covering clients, counterparties, and liquidity providers. These models typically estimate peak potential outflows from margin calls, settlement failures or increased settlement requirements, client facility drawdowns, withdrawal of unsegregated client money, pre-funding requirements and failure to rollover debt.
  • Notifying group companies of large potential next day use of liquidity facilities including when above agreed limits.
  • Minimising the risk that liquidity facilities could be partially or wholly unavailable by ensuring at the end of each day adequate levels of liquidity to meet the next day’s instantaneous BAU and stressed cash outflows is placed:
    • in operational bank accounts
    • with CCPs such as LME Clear and ICE Clear; central clearing houses (CCH) such as London Clearing House (LCH) and Euroclear, and clearing brokers  

      To be as prepared as possible to cope with foreseeable large liquidity outflows.

• Identifying large open positions, especially the short position legs of matched-principal trades, and ensuring inventory or stock borrows are in place so that these trades can be settled, avoiding the risk of being bought-in.
• Ensuring that in certain defined circumstances clients have prepositioned cash/inventory for next day settlement.

Poor Practice

• Stress testing at frequencies, such as monthly or less frequently, that did not reflect the instantaneous nature of liquidity stresses.
• Stress testing that did not reflect the timing and magnitude of actual and potential liquidity outflows the firm has been exposed to in recent years. For example, during the nickel crisis firms:
  • underestimated the number and magnitude of margin calls (significant liquidity outflows) they had to meet promptly from CCPs for cleared derivative contracts across the day and into the next day, and
  • overestimated the timing and magnitude of margin (liquidity inflows) received, if at all, from their clients for ‘matched’ uncleared OTC derivative contracts, at some point during the next day
• After the crisis, firms assumed the LME’s daily price-limits, implemented after the spike in Nickel prices, would prevent similar extreme price movements happening across several consecutive days (which they were not designed to do) rather than during a single day as happened in early March 2022 (which they were designed for).
• Failing to calibrate value-at-risk (VAR) models, used to calculate their potential margin call liquidity requirements, to accurately forecast the potential timings and gross magnitude of each individual material cash outflow.
• Failing to cross-reference lookback-period based calculations, used to quantify peak stressed liquidity requirements, with the firm’s current positions to estimate peak exposures based on current business levels.
• Failing to include a realistic range of idiosyncratic stress scenarios and assumptions that would increase their liquidity requirements, reduce liquid asset levels or curtail access to sources of liquidity.
• Failing to capture implicit credit exposures and liquidity usage arising from:
  • the time given to clients to meet OTC margin calls, and
  • naked short settlement positions and the consequences for the firm if it was bought-in, ie, the client(s) failed to deliver the inventory
• Not considering maturity/roll-over risk impacting the effectiveness of medium-term notes, structured notes and similar programs. Despite the lessons of the global financial crisis, firms continue to rely on less stable funding sources to varying degrees and underestimate their exposure to:
  • roll-over concentration risks
  • auto-maturity concentration risks, and
  • risks of negative signalling by not being prepared to adequately support buy-back programs during periods of stress
• Failing to consider and model the behaviours of clients specifically in an idiosyncratic liquidity stress. For example, clients withdrawing assets and cash held over-and-above what is required under title transfer collateral arrangements (TTCA) and reducing risk exposures which in turn further increases levels above TTCA requirements. Clients carrying out enhanced checks and reconciliations which increases delays on when payments are made, including margin payments.
• Failing to adequately consider the behaviour of liquidity providers and not anticipating that:
  • parent companies or other group companies may not only severely curtail or withdraw liquidity facilities but also sweep liquidity back to group during stress events. Withdrawal of liquidity facilities and liquidity sweeps are common recovery and wind-down actions in group recovery and wind-down plans
  • third-party liquidity providers may severely curtail or withdraw liquidity facilities, either because they are in stress themselves or because they have reconsidered their risk appetite
  • for operational reasons, such as service outages at correspondent banks, access to liquidity facilities may be severely curtailed or totally unavailable

Recent financial stability board (FSB) reports have highlighted that the functioning and resilience of the non-bank sell-side ecosystem depends on the availability of liquidity and its effective intermediation under stressed market conditions. Certain activities and types of entities (so-called ‘key amplifiers’) are more likely to contribute to aggregate liquidity imbalances. Shocks can also be amplified due to the size and structural characteristics of such entities and their behaviour in stress. On the liquidity demand side, shocks manifest through unexpectedly large margin and collateral calls for derivatives and securities financing trades. This can give rise to asset fire sales by market participants seeking liquidity to cover these calls and the transmission of stress to other parts of the financial system and the real economy.

Finally, we observed that in their risk horizon scanning, firms were not considering emerging risks. For example, risks arising from technological innovations such as messaging applications and applications allowing clients to access their accounts remotely, thereby amplifying the impact of rapid behavioural responses such as withdrawal of money and assets, including surplus money and assets held under TTCA. This highlights the need for ongoing monitoring of all types of funding providers, including clients (especially those with credit balances), lenders, clearers, liquidity providers and providers of banking services so that firms may act promptly to make alternative contingency arrangements, where necessary.  

Good practice

• CFPs with clearly defined quantitative action triggers that if breached provided Board delegated authority to a named individual to take action from a Board-approved list of contingency actions.
• Clearly identified contingency actions, appropriate to each firm’s unique circumstances, that could be taken promptly:
  • For example, actions to increase liquidity levels such as:
    • drawing-down from liquidity facilities
    • temporarily increasing the size of liquidity facilities
    • selling unencumbered inventory
    • leasing out precious metals
    • using sale and repurchase (repo) arrangements to preclude outright sales crystalising losses, and/or rehypothecate collateral held 
      Being mindful to avoid negative signalling.
  • For example, actions to conserve liquidity such as:
    • reducing trading limits
    • reducing size limits for transactions in single-name securities/commodities
    • having clients pre-fund large and/or illiquid transactions and having clients pre-position inventory (securities and/or commodities) to cover sales or provide proof that executable stock borrow arrangements are in place
    • using name give-up and exchange give-up of unsettled transactions
    • having clients pre-fund potential variation margin requirements arising from larger derivative positions or refusing credit facilities to clients to cover larger back-to-back margin calls
    • reducing the amount of credit provided to clients. For example, to ensure this action can be taken promptly firms:
      • periodically updated the list of impacted clients and the amounts by which credit facilities would be reduced
      • considered and modelled anticipated client behaviours in range of stress scenarios
      • considered the likely consequences of taking this action
      • included in their contingency plans agreed client communications strategies and, where appropriate, coordinated communications strategies across the wider group.
• Frequently test the operability of contingency actions during business-as-usual conditions and avoid the risks of negative signalling that arise if actions are only implemented during a stress situation. For example, frequent random drawdowns from committed facilities and increased utilisation of uncommitted facilities.   
• Considered in their WDPs the outputs of reverse stress testing (RST), ie stresses that would lead to a point of non-viability and result in the WDP being invoked.
• Ensuring WDPs are operable in a stressed environment and included contingencies for activities that potentially require longer execution times. For example:
  • notice periods required to terminate contractual arrangements
  • time to provide notice to clients so assets can be migrated off the platform
  • the eventuality that some clients may not be immediately contactable
  • resolution of issues that would arise regarding disputed client money and client asset positions, delayed and/or disputed settlements, and aged open reconciliation issues
• WDPs which include projected profit and loss, balance sheet and cashflow statements at appropriate levels of granularity. For example, at the beginning of the process when multiple critical cash outflows and inflows are expected to happen within similar timeframes, the cashflow analysis was presented by day/week detailing the timing of all material cash inflows and outflows.  

Poor Practice

• Failing to consider or put in place CFPs that can be invoked and executed in a reasonably timely manner in a period of stress.
• Not practicing, not being able to implement theoretical steps.
• CFPs without clearly defined quantitative action triggers, especially non-financial action triggers. For example, during the cyber-attack on ION’s cleared derivatives platform in 2023, many firms continued to trade at normal volumes even though their middle and back office were no longer operable. Consequently, trade volumes materially exceeded their capacity to adequately reconcile their positions.  
• Failing to consider the full range of actions available. Even though other actions were available, several firms included only one contingency action, usually to draw-down from a single liquidity facility, or request funding from their parent. Firms that have not considered the full range of contingency actions available to them would be at greater risk of failure in the event of a sudden liquidity stress. 

Good practice

• Teams with appropriate levels of expertise:
  • constantly updated their expertise and knowledge
  • constantly upgraded their frameworks to meet the ongoing realities of their businesses
  • put processes in place that act in tandem with the firm’s operational processes to help the firm effectively manage market and idiosyncratic liquidity risks and cope with periods of extreme market volatility and idiosyncratic shocks such as severe operational risk events
• Ensuring a holistic approach to liquidity risk management. For example, having a clear understanding of the timings and magnitude of BAU and stressed cash inflows and outflows arising from operational processes, especially outsourced operational processes.

Poor Practice

• Teams responsible for liquidity risk management lacking expertise in:
  • the liquidity risks inherent in the firm’s business model, leading to these risks being understated or absent from the liquidity risk stress testing model, CFP action triggers and contingency actions
  • the operational processes of the firm and the activities of the operations teams.  Where this happens, frameworks are not designed to support the activities and processes of the operations team. For example, some firm’s liquidity risk managers could not promptly identify when operations, processes or controls had failed and therefore could not promptly alert the treasury function of the need to take mitigating actions
• Weak understanding of outsourcing arrangements especially where separate unregulated subsidiaries provide operations and related support services. From a practical perspective failing to ensure that the liquidity and funding risks arising from these outsourced activities are identified and proactively managed with appropriate oversight either at a group level or at the solo regulated subsidiary level. And, not including assessments of these risks in either the group ICARA or the ICARAs of the regulated subsidiaries.
• Applying a tick-the-box approach to liquidity risk management. Some firms’ frameworks were narrowly designed around meeting regulatory requirements rather than being designed to be ‘fit-for-purpose’ in line with the practical principles of sound liquidity risk management.