Case Studies @ Bellwether

Case Study 1 - PMS Industry

A portfolio management team facing technology issues in terms of latency and number of scripts they could manage to trade. The data feed was consumed in excel and computation were run on them to trade the symbols manually which required lots of effort to monitor and opportunities would disappear before the guy could even trade them. There was also limitation on the number of scripts that could be traded.

The team were able to take maximum 20-30 trades that too in selected scripts which didn’t given them good diversification in terms of portfolio for days where the prices go against them. Further, they went actively able to manage the open orders and executed trades to managing the risk. There was a need to migrate to not only better "technology" but also a “better "solution".

Our Expert Value Addition

Our expert on the first day on the project not only understood the entire end to end system but came up with a strategy to improve the overall system. The solution our expert designed and implemented was,

  1. In memory storage of real time feeds for faster computation and view all possible opportunities in the market.
  2. Multi Thread architecture to process various functions together which include
    • a. Identifying opportunities
    • b. Filtering opportunities and placing order
    • c. Managing open orders and executed orders
    • d. Risk Management
    • e. Real Time Monitoring and Reports
    • f. System Logs
  3. End to end solution with no manual intervention
Technology Used
  • 1. Core Java
  • 2. MySql Database
Results
  • 1. Was able to capture more than 98% opportunities which was less than 5% before
  • 2. Previous system had lot of latency which eroded a lot of alpha till then time trade was executed while the new system was able to capture most of alpha thus increasing per trade profit.
  • 3. New system produced and managed more than 10000 orders. Previously it was less than 50, This helped to capture most of the opportunities and get the best execution price.
  • 4. Better risk management as all the open and executed order were managed to book profit, exit on loss and cancel and modify the open order as per current opportunities. Before no scope to get the execution at best price as hitting market orders was the only options and executed orders were not managed in a efficient manner.
  • 5. More than 1000% increase in the overall profit with the new system in place.
  • 6. More robust than the previous system and not a single glitch or issue since it went in production in more than 2 years.
Case Study Epilogue

This case is a success story from the previous assignment done by our expert. The objective of this case was to demonstrate our capabilities. Our experts have consistently delivered such order of magnitude performance improvements. Typically these are "turn-around" stories for the organizations.

Case Study 2 - Hedge Fund

A Hedge fund team was facing technology issues in terms of latency and getting the desired execution price for a statistical Arbitrage strategy (Pair Trading) they were running . The data feed was consumed in excel and trade were taken manually. In pair trading we need to execute both the stocks in a pair together with a specified slippage to capture the alpha generated by the diversions in the stock execution was very keen to this strategy.

The team was not able to get the price generated by the algo since by time they would execute or modify the orders the prices would no longer be favour. Due to which they were not able to capture the entire alpha with manual trading. Further, they went actively able to manage the open orders and executed trades to managing the risk. There was a need to migrate to not only better "technology" but also a “better "solution".

Our Expert Value Addition

After understanding the complete flow of the project we found many areas where the current implementation can be improved to not only minimize risk but also improve profits. The solution our expert designed and implemented was,

  1. In memory storage of real time feeds for faster computation and view all possible opportunities in the market.
  2. Execution algorithm to minimize the slippage.
  3. Multi Thread architecture to process various functions together which include
    • a. Managing open orders and executed orders
    • b. Risk Management
    • c. Real Time Monitoring and Reports
    • d. System Logs
  4. End to end solution with no manual intervention
Technology Used
  • 1. C++
  • 2. SQL Server Database
Results
  • 1. Was able to trade more number of pairs with lower slippage.
  • 2. Previous didn’t had a execution algorithm to get the best execution price which we implemented does improving the final pnl.
  • 3. New system was able to trade more than 80 pairs to possibility to trade lots more without any changes architecture
  • 4. Better risk management as all the open and executed order were managed to book profit, and if only one leg of the pair was executed then the other was squared off to avoid any directional exposure
  • 5. Increase in the overall profit with minimized risk
  • 6. More robust than the previous system and manual intervention required
Case Study Epilogue

This case is a success story from the previous assignment done by our expert. The objective of this case was to demonstrate our capabilities. Our experts have consistently delivered such order of magnitude performance improvements. Typically these are "turn-around" stories for the organizations.