- Kingland Platform
*A portion of this article was originally published in the Journal of Securities, Operations & Custody by Henry Stewart Publications.
Imagine what would happen if Amazon couldn't authenticate and verify payments at the time of purchase. Instead, at the end of each day, Amazon had to figure out if they had enough products to match demand and also determine if their clients had the cash to make the purchase? In a simplified way, this is the challenge with settlement optimization of securities - understanding the amount of securities transactions that will clear the settlement process.
Settlement is the final leg in the life cycle of the securities transaction process. Confirming failed transactions can be expensive and time consuming for back-office operations. The month of March 2020 saw the amount of fails range from a low of $41.8 billion to a high of $110.1 billion. A fail is when a buyer is unable to deliver funds or a seller fails to deliver an asset by the settlement date. Most stocks and bonds settle within two business days (T+2) after the transaction date. This is of great importance to financial institutions as missed transactions can expose parties to credit risk and settlement risk.
Determining the optimal order in which transactions should be processed is a core problem of the settlement process. Firms need to understand the relationship of buyer and seller where if someone sells shares, someone else has bought those shares. The current legacy systems typically process these large quantities of transactions across many asset classes as one large volume of buy/sell information. This creates failures which lead to teams of individuals chasing down details on short sales, securities lending discrepancies, allocation issues and a multitude of other fail scenarios.
The simplified goal of settlement optimization is to increase the percentage of the day's trades that can be resolved. Can a firm take all of its transactions and apply an algorithm that will find the optimal set of transactions that will settle?
We've proposed the use of algorithms to improve the settlement process by identifying the optimal set of transactions that will settle. The whole point is to increase the percentage of the day's trades that can be resolved. We've estimated this technique could improve efficiencies as much as 80 percent by improving the amount and accuracy of securities settled.
Visit this link to learn more about our approach to optimizing the settlement process at the Journal of Securities, Optimization & Custody. The full article is available for JSOC subscribers.