Blockhouse: Finding $4.2M in OTC Bond Transition Costs

Utsav joined Blockhouse in August 2024 as a Quant Analyst, stepping into a post-trade analysis project on the US OTC bond market that was already a week underway. The dataset in front of the team covered roughly 58,000 bond trades over three months, representing approximately $200 billion in notional value.

His first two weeks were spent entirely on understanding the problem. He mapped the data structure, the libraries in use, and the industry-specific KPIs — slippage, transition cost, market impact — before touching any analysis. He then spent the following two weeks collaborating across the ML team, the quant team, and the CEO, contributing to documentation and small library builds while building a shared view of what the data could answer.

Identifying the Right Levers

The core question was where an institutional investor was losing money in execution. OTC bond markets are broker-intermediated, meaning a buyer or seller of size experiences slippage and market impact from their own activity. The team narrowed the problem to three optimizable variables: trade size, market timing, and dealer selection.

Utsav and the team back-tested these levers week by week and day by day across the dataset. The data validated the direction: breaking large orders into blocks reduced market impact, timing trades to lower-liquidity windows reduced slippage, and routing to dealers with historically tighter spreads on specific instruments produced consistent savings.

Outcome

By the end of the three-month engagement, the analysis had identified $4.2 million in transition cost savings across the bond portfolio. The work covered TIP bonds, T-bills, and standard Treasury bonds in the US market.