Imagine you’re a stock broker on the floor of the New York Stock Exchange. Seconds may make the difference between your customer making a profit, or not. Then an unexpected real-world event occurs, and the network that connects the financial market with the billions of dollars in investments isn’t able to handle the traffic. In a matter of seconds, millions of dollars are lost.
While this may sound overly dramatic, in just the last few months we’ve seen this scenario play out several times, with the Brexit vote – which holds the record for the worst one-day market-drop of nearly $2 trillion (and total drop of over $3.1B), followed by the US Stock Futures and Forex collapse after the announcement of President-elect Trump.
To combat this, financial service firms are becoming increasingly serious about their networks, but it’s an environment with many challenges? Market trading networks must deliver data in real-time for traders to conduct transactions on a global scale, and this creates fundamental obstacles for network performance monitoring (NPM) systems including:
- The use of User Datagram Protocol (UDP) is what allows for real-time data distribution. However, this comes at the cost of losing network transaction redundancy.
- The need to handle massive swings in the amount of network traffic caused by real-world events that impact the market.
- SLAs between exchanges and their customers for dropped traffic that require proof of the source of missing data. If broken, fines, financial liabilities and losses will ensue.
- Stringent compliance regulations for both accuracy and transparency, since any errors impact both trading prices and spreads.
What is the solution? I’ll address that in my next FinServ blog post, but if you want a hint, it has to do with next-generation monitoring.