By Mohit Gupta, Senior Product Specialist, Cassini Systems
‘We now have just over 600 million tied up in collateral to satisfy our Initial Margin requirements’ was one of the statements of a COO in a meeting at a global asset manager conference which Cassini attended.
He continued by asking how, and who, determined and managed the costs of this outstanding collateral but most importantly wondered how the amount could be minimized and controlled to lower the increasing impact of the company’s performance. Optimization and monitoring were on the list of potential solutions for this asset manager – but where to start?
Collateral management has traditionally been a back-office, post-trade tool. However, with collateral costs causing a drag on real returns, firms have started to bring the collateral function into the front office and look at optimizing collateral postings.
The Global Financial Crisis (GFC) of 2008 led to a waft of sweeping regulations across the financial markets including the requirement of derivative transactions to be collateralized to reduce counterparty credit risk. Starting with posting collateral for variation margin (VM), the regulation has now also heavily extended to posting initial margin (IM). The advent of the Uncleared Margin Rules (UMR) has made the requirement even more stringent by making it mandatory even for bilateral derivatives to be
compliant for collateral postings.
With an increasing amount of margin requirements, the limited supply of high-quality liquid assets (HQLA) gets stretched and further constrained in times of stress. This poses a unique challenge for firms to not only make their systems compliant with the regulations, but also adept to tackling the challenge of collateral requirements. However, there is a light at the end of tunnel as this is an opportunity to have systems in place which can help streamline the workflow and optimize the requirements.
Margin Optimization – First step to achieving Collateral Optimization
To solve the problem of collateral optimization, the first step is understanding how much collateral is required to be posted. This requirement stems from the fact that derivatives require margin to be posted (VM and IM). Optimizing this margin requirement lowers the amount of collateral required and hence reduces the cost of collateral. This optimization can be done both pre-trade and post-trade.
Pre-Trade Margin Optimization
To show the importance of pre-trade optimization, we simulated a Relative-Value Hedge Fund clients’ historical cleared book from the start of year – with an empty book – to the end of year. During this period, the client would regularly trade cleared swap trades in EUR, USD and GBP, and clear them at broker on a currency basis to make it operationally easier rather than the cheapest broker.
Clearly, one gets offset between the trades of the same currency but loses out on offsets among currencies, which being a RV fund is one of the most common trading strategies.
As one can see, the savings are lower to start with, but grow significantly over the course of year and the average margin requirement over the course of the year is roughly 150m (million) lower, which means 150m lower collateral to be funded. Even though above is an example for cleared trades, one can easily use it for any derivative, cleared or bilateral.
Post-Trade Margin Optimization
Like pre-trade optimization, firms can engage in post-trade optimization which can help reduce margin requirements generally done on a weekly or monthly frequency depending on the firm.
- Risk Rebalancing Among Dealers/Clearing Brokers: Balancing positions across dealers and clearing brokers, helps in achieving maximum offset and reducing margin requirements along with reducing concentration add-ons due to build-up of large positions.
- Futures Cross Margining: Some exchanges offer cross margining the futures against available OTC positions and this can lead to tremendous savings especially if they offset with each other.
- Strategic Clearing and Backloading: Bilateral and cleared trades have different liquidity and different margins, and only new trades after the UMR phase in dates are in scope for bilateral margin. This offers some firms the opportunity to choose between trading certain trades bilateral or cleared depending on cost benefit analysis and/or back load legacy trades into either UMR scope or cleared portfolio to create offsets and reduce margin.
Once the margin requirement has been understood and optimized, the next step is optimizing the collateral as this will directly impact the cost and hence the returns. Collateral optimization is a multi-dimensional problem requiring understanding of the following:
- Collateral Eligibility: Depending on the agreements with counter parties, the type of collateral eligible with one might not be eligible with another. Further, rules regarding collateral substitution, transformation and re-hypothecation differ from counter party to counter party and can greatly influence the choice of collateral postings and hence the resulting costs.
- Collateral Availability: With the understanding of eligibility, the next step is to look at the inventory of available collateral, possibly incoming collateral (if re-hypothecation available) and collateral posted already to understand if a transformation can help in optimizing holistically.
- Collateral Haircut: Collateral haircut determines how much extra value of collateral will need to be posted over and above its value. This is due to the quality of the collateral and the possible loss in the value when trying to settle trades using collateral in case of default, stress etc.
- Collateral Cost: The most important component in the optimization exercise is the cost of collateral. The cost of collateral reflects the cost of funding the collateral (either in primary or secondary market) if the collateral must be funded. This can also work as the opportunity cost of the collateral in cases where funds are a flush with collateral.
- Transaction Costs: These are costs which the custodians/fund admins charge for safe-keeping the collateral. There will also be transaction costs involved with moving of collateral and this will further add constraints to the optimization problem.
Given the above dimensions, a simple and traditionally used ‘waterfall’ model for collateral allocation works sub-optimally and hence needs use of proper optimization framework. Using Linear Programming and Mixed Integer Model is one framework to solve this complex problem. Introduction of more constraints, like the amount of cash one can post to a custodian, further makes the optimization challenging.
Just like margin optimization, collateral optimization can be done both at Pre-trade and Post-trade levels.
- Pre-trade: The optimization would help not only to choose the broker/dealer with least margin increment but also look at the available pool of assets and check if enough collateral is available. Further, given different collateral inventories and possible different funding costs, it helps to optimize collateral at pre-trade level.
- Post-trade: This holistic optimization would look at all available inventory in order to move collateral across if need be.
To add context to the savings, let’s say a firm with 3bn(billion) USD AUM (assets under management) has an average margin requirement of 600m USD throughout the year. Using optimization on collateral, of one was to save 50bps on funding, it would translate to 3m USD to the bottomline which is 10bps to the returns.
As above, given the challenges and potential savings involved, more and more firms are looking at these optimization strategies holistically and from the point of view of arming their front-office to reduce the drag and ultimately, improve returns.