Traditionally found in the dusty corridors of back-office systems, margin is increasingly moving out of the shadows into the spotlight. Firms are looking to understand more accurately how their margin is derived, how it might be optimized and what a theoretical position might cost in the future.
Ironically, while margin calculation is increasingly demanded by more market participants, clearinghouses are making it more difficult for market participants to reproduce their margin calculations independently, as they did in the past.
For the uninitiated, “margin” in the context of derivatives trading is perhaps a misnomer. Simply put, margin is the deposit of cash or collateral needed to guarantee a derivatives position. With any derivatives position, its value over time will fluctuate either positively or negatively based on the underlying data on which the contract was derived. To protect both buyer and seller of this contract as it moves through time, margin is collected to guarantee payment of this fluctuation – variation margin covers price movements observed and initial margin to cover potential future price movements.
While margins may be posted to a clearinghouse or a custodian, the need to effectively check, predict and validate this margin is an essential part of risk management for any firm holding derivatives positions.
There are three stages an organization moves through as it on-boards and becomes expert in dealing with margin:
The first stage is usually operationally driven and is about calculating the margin requirement, whether it be margin from a central counterparty or bilateral margin required for non-cleared derivatives. This may be to calculate amounts for sub or client accounts or in the case on non-cleared margin, to determine how much the institution itself has to post. All that is needed is the margin number under the appropriate methodology and very few questions are asked.
Next, to explain what makes up a margin number. This is very important in the case of bilateral margin on non-cleared derivatives as both sides must agree on an amount, but in the case of cleared derivatives even though margin is centrally defined by a clearinghouse, there can also be a surprising amount of variability. The type of account holding the position, the risk profile of a position holder at their clearing member, whether certain offsets between positions are taken and other variables can significantly alter the final margin.
The last and most sophisticated use of margin is to be able to forecast what that margin will be under different market environments or more importantly, with new positions. Most commonly, this is a “what-if” – where new trades are added to an existing margined portfolio and the impact of the new trade on the margin call is requested. This is equally useful in the cleared or non-cleared case, and is gaining importance all the time as banks are ever more conscious of the availability of liquid assets and the costs of funding that margin.
It’s important to note that this last forecasting requirement is driven by a different persona within the bank – front-office traders. Traders look less for penny-accuracy, placing importance on cleanness of user experience and calculation speed.
With this change in emphasis of requirements around margin, there needs to be a re-thinking of the engines behind margin – from a precise but unwieldy infrastructure component to a fast and agile service.
No longer are margin algorithms “open-source” and shared with the community at large. Some are partially shared, others only accessible via clearinghouse APIs or deployable code. What this has created is an extremely complex, varied and constantly changing environment. For market participants trading across global markets, this means accessing margin is becoming more difficult and more expensive relative to the past.