InteDelta’s Nick Newport and SunGard’s Ted Allen offer answers to some audience questions from a recent DerivSource collateral optimization webinar. Read on for answers to questions on the impact increased collateralisation with have on pricing policies, how to measure the effectiveness of optimization activities and much more.
Q: What does optimization mean to buy side and sell-side participants? It seems optimization can mean different things to different people.
It’s certainly true that – across the industry – the term collateral optimization has frequently been used to refer to a diverse range of activities in the collateral management space. There does now, however, seem to be a common usage of the term emerging to refer to the optimization of inventory for collateral purposes? Even within this more narrow definition though there is still room for a number of different understandings of the concept. This primarily involves using similar optimization techniques for slightly different business goals.
For example, a sell-side broker dealer may have a large existing inventory of proprietary assets, which, if identified in a timely manner, could be used to satisfy collateral requirements. This is likely to be a more cost effective means of meeting these needs than borrowing cash or securities in the market. And even from amongst the existing inventory of assets, there will be some assets, which are more ‘optimal’ to place than others, depending upon what else these assets could potentially be used for (for example, were the broker dealer to lend them, repo them out etc).
The optimal use of inventory assets to cover collateral calls in this way is a common use of the term collateral optimization for both buy-side and sell-side firms.
The related idea of collateral transformation has also gained significant currency in recent times and this is of particular importance to buy-side firms. For example, a buy-side firm may trade cleared derivatives but also be sitting on a significant inventory of non-central counterparty (CCP) clearing eligible assets. Collateral transformation refers to the process whereby a sell-side firm will transform these non-eligible assets into eligible assets to enable the buy-side firm to meet its CCP margin requirements without having to raise funds in the market. This transformation is akin to a securities lending or repo trade, and the buy-side firm will aim to be able to raise as large a value of eligible assets as possible using its inventory (or conversely, use as little of its inventory as possible to raise the necessary eligible assets). This is where the optimization process comes in.
The same optimization techniques can also be used to minimize funding costs through re-use of collateral and other inventory assets held, to minimize risk-weighted asset (RWA) and for various other purposes. The same process is at the heart of each of these though, and the differences primarily come down to how the various optimization constraints are parameterized to meet the desired business goals.
– Nick Newport
Q: Is the issue “collateral transformation” rather than optimization and if so does this become a credit issue as in effect you are moving the counterparty credit risk from the CCP’s (for cleared trades) back to the dealers or the providers of the enhancement?
Collateral transformation can certainly be seen as part of the optimization question. Transformation is key either as upgrades to cover collateral requirements where there are no eligible assets available, but also in the context of transforming existing assets to other types of assets that can be better used taking into account eligibility and haircut rules to make the optimum use of any asset. Collateral transformation of course is achieved through securities lending and borrowing activity. This does create some credit risk, but as these are inherently collateralised trades, the additional risk should not generally be material.
– Ted Allen
Q: How do you measure the effectiveness of optimization activities (i.e. techniques such back testing, to assess teams/processes/practices)? What are the challenges in measuring it & who is doing it?
There are many distinct aspects to collateral optimization, some with tangible measurable effects and some resulting in process improvements and controls, which are more problematic to measure. However, we at SunGard are working with a number of our clients to measure the potential savings that could be made by implementing optimization algorithms in the allocation of assets against collateral requirements. This is not a difficult activity once you have access to an optimization platform with this kind of capability to understand collateral requirements, the current allocations and the available portfolio and its eligibility and cost of use. The results will show you your current total daily cost of collateral and your theoretical daily cost if you were to optimize.
– Ted Allen
Q: I understand that enterprise collateral management is the next big thing. How can an organization adapt to this trend?
The changes required to implement an enterprise collateral management process, in the way that most institutions are now viewing this idea, are manifold. The key difference to how collateral management was traditionally viewed is that an enterprise collateral management process is not just an operational process of calling, booking and settling collateral, it is much wider than this. It is an integrated process covering multiple stakeholders across – to name only the main ones – the front office, risk and operations. This brings with it many challenges – some organizational, some process, and some technology.
I think the key to success here is to clearly understand the target organizational and process model prior to jumping in to wide ranging technology solution design. This doesn’t have to be as difficult as it sounds, but at least having a clear idea of where the organization is trying to get to can be a fundamental determinant of the final success. This will start with the collateral management governance framework, take into account the various touch points between departments in the process and then lead into the overall process model.
Once this initial organizational and process definition has taken place, definition of the target technology solution can begin with clear goals.
Most financial institutions will find that many existing technology components can be leveraged to meet the needs of the enterprise collateral management architecture. Other components – in particular the inventory aggregator and the optimization component – will need to be built or bought. There are vendors in the market, which have specifically developed such components, and some institutions have gone down the buy route and others the build route, depending upon the specific organizational preferences of the given institution.
– Nick Newport
Q: How can stress testing affects collateral and capitalisation transformation?
A real benefit to a collateral optimization solution is the ability to model what would happen to your collateral allocation under stressed market conditions. You need to be able to manage the risk across multiple dimensions, for example combinations of rating changes, FX rates changes and changes in collateral requirements together may have a significant impact not just on the amount of collateral you have to post, but on the cost of that collateral as assets you currently use may become ineligible across different agreements. Collateral optimization tools have to have the ability to model these complex scenarios.
– Ted Allen
Q: Most of the credit support annexes (CSAs) some work with are based on a grid. Some are cash and securities combined. How do you categorize these in order to decide what would be the appropriate collateral. Do you switch some to cash or just use all securities or keep status quo?
A collateral optimization algorithm will understand the full set of eligible assets and the haircuts that are specific to each collateral call, be it from a credit support annex (CSA), Global Master Repurchase Agreement GMRA, Global Master Securities Lending Agreement (GMSLA). Taking into account the total set of these requirements it will apply linear programming methods to solving the problem of what combinations of allocation of assets to calls gives the lowest overall cost. It’s not as simple as allocating cheapest collateral first and continuing sequentially, that approach will not minimise your costs. What is needed is true optimization that takes the holistic view. We at SunGard are working with our clients implementing just these solutions.
– Ted Allen
Q: Will the increased levels of collateral introduced by Dodd-Frank Act (DFA), European Market Infrastructure Regulation (EMIR), etc have an impact on pricing policies applied (for instance, the higher level of collateral impacting a risk based valuation approach)?
In many ways, all of the current trends around collateral management and the optimization of collateral are ultimately centered on pricing.
As regularly commented, the wave of current regulation – DFA, EMIR, Basel III – will result in a significant increase in the need for financial institutions to source and fund large stockpiles of high quality assets (cash, government securities and other similar quality assets) for both collateral and capital purposes. The largest single impact this will have will be on the cost of entering into derivative trades (whether these be cleared or non-cleared). Put simply, the cost of funding these assets is an additional cost of the transaction and will need to be priced into deals accordingly.
However, this is where the situation starts to get more complex. In order to price the cost of collateral into a deal pre-trade, the dealer needs to have estimation or expectation as to how much collateral may need to be placed through the lifetime of the trade to support the deal. This will be a function of the expected evolution of the mark-to-market (MtM) on the trade over the life of the deal. Moreover, these calculations need to be undertaken at portfolio level, per collateral agreement, as this is the level at which the collateral requirement is managed. Effectively, this is very similar to a credit valuation adjustment (CVA) calculation, except modeling the other side of the curve – i.e. the expectation that the MtM may move in the counterparty’s favor.
Having calculated, the expected collateral requirement, it is then necessary to calculate the expected cost of funding the collateral to meet this requirement. This will be a function of what collateral is eligible under the agreement, the ability of the firm to source the necessary collateral etc. This is where collateral optimization comes into the pricing picture, the more optimally an organization is able to fund its collateral requirements, the better pricing it will be able to provide on trading derivatives.
This is very much an evolving topic and most institutions are still in the process of reviewing processes and systems to effectively integrate the impact of funding collateral into the pre-trade pricing process.
– Nick Newport
