Overbond integrates an AI-based margin optimization model

Overbond integrates an AI-based margin optimization model

Overbond, the leading API-based credit trade automation and execution service in global financial markets, has integrated an AI-based margin optimization feature into its existing automated trading system. As a result, traders can now train Overbond’s automated system to optimize their success rate based on desired desktop trading parameters. This increases the number of trade requests that traders can respond to without trader intervention.

The development of new financial products, the emergence of all-in-one electronic platforms, and the rise of non-broker-based liquidity providers using algorithmic and high-frequency trading have all reshaped fixed income trading. Now, bond traders face heightened volatility and an evaporation of liquidity amid rising rates, inflation and fears of recession. Electronic and automated trading has become the new standard to meet these credit desk profitability challenges.

Until now, automated workflows allowed traders to discover price and liquidity, but often required the intervention of the trader to ensure trades were within the desk’s desired trading margins. This need for human intervention has created a bottleneck in the workflow and prevented the full automation of many trades. Overbond has achieved a major breakthrough in fixed income trading by fully automating the margin optimization function within the Overbond automated trading system and data feeds through the API.

Overbond’s margin optimization model optimizes the distance to be covered based on the best executable price in the RFQ protocol under Overbond’s pricing model, COBI-Pricing LIVE. The margin model now incorporates variables that provide insight into market risk at the security, issuer and macro level and ensure that the automated margin is sensitive to intra-day risk movements. This data is collected from data providers such as TRACE and includes bond specific data such as coupon and amount outstanding, issuer specific data such as number of quotes and average price volatility for the issuer, and sector-specific data such as the volatility of the sector benchmark’s bid-ask spread.

Consideration of trade size is necessary to fully automate trading, as it is an important factor in pricing and determining desired margin. To solve this problem, the Overbond Margin Model estimates total market capacity per bond, which the model then uses to isolate margin sensitivity to trade size.

Market risk and capacity are calculated from data available to all market participants, but desks gain their competitive edge through the unique metrics they use to make their trades profitable. Any fully automated trading system must incorporate these parameters, which is why Overbond incorporates desk-specific execution records into its analytics, allowing automated trading to adhere to the execution style, bias and approach of desk margin.

“Achieving an optimal win rate is key to maximizing sell-side desk P&L and the integration of auto-margining capabilities into the Overbond trading system enables true end-to-end automated trading that does not require human intervention,” said Vuk Magdelinic, CEO. from Overbond.

Source: overbond

Similar Posts