Lindsey will also be speaking at…
By Lindsey Burik, Head of Electronic Trading Americas/Europe, Managing Director, Mizuho Americas
Lindsey, can you please tell the Risk Insights readers a little bit about yourself and what your current professional focus is?
I spent the bulk of my career in international equities and execution services, though I originally thought I would pursue a career in intelligence or law. After graduating from Boston University,I landed at the brokerage arm of Bloomberg, Tradebook. From there I helped drive electronic trading for Credit Suisse, Goldman Sachs, and now Mizuho.
In my current role as Managing Director and Head of Electronic Trading for the Americas and Europe, my focus is on the development, marketing, and sales of Mizuho’s electronic trading product, known as MET. I was tasked with building Mizuho’s algo platform from scratch. Now, with a state-of-the-art platform, we’re able to present unique liquidity access and trade execution for a wide range of clients around the globe.
What, for you, are the benefits of attending a conference like the ‘X-Tech 2019 Convention’? What can attendees expect to learn from your session?
Mizuho is in an important growth phase and very focused on how technology can accelerate that growth. We have chosen to focus on the equity markets’ microstructure to educate people about how liquidity changes in the market, how it affects trading objectives, and who is participating in individual stocks. We use various AI tools and technologies to create real-time views into “what is happening” in the market.
Can you please provide an overview of what you expect the key takeaways to be from your ‘Crazy rich analytics’ session at X-Tech 2019?
On average, 15% of all US liquidity is “inaccessible” with some stocks as high as 50%. That means you can’t actually trade with as much as half of the volume in a particular name that you see available on the tape. And as you chase it, your trading costs go up and up.
We do the analysis to fine tune algorithms that find real liquidity and tailor portfolios with stocks that have the right trading characteristics for their portfolio. So while many people have an electronic trading platform, it has to be customized properly. Think of a smartphone. Many people have one, but the apps and settings on individual phones are not the same. Algo trading should mirror that concept – the features personalize the platform.
In your opinion, how can the market work effectively with AI? What are the challenges / benefits?
‘Man with Machine’ is the right approach to AI models. At Mizuho, we use Deep Learning AI to create signals for our Asia Algos in order to try and improve their performance. We use 5400+ data feeds and process trillions of data points overnight to try and predict the real-time ups and downs of a stock.
Our biggest challenge is processing power. We own one of the largest server farms on the sell side, yet we still need more to expand to the US. We are working on this now.
With 85% of all equities ending up in an algo, what does that mean for the market?
When you consider that there are 13 stock exchanges and about 40 off-exchange trading facilities in the US alone, I’m surprised the number isn’t even higher. But the rise of algorithms has occurred gradually over the last 15 years in response to the proliferation of trading venues and the need to comply with SEC guidelines for best execution.
The same way you use kayak.com to find the best-priced airline ticket, you use an algo to find the best-priced execution. So what about the 15% that is not algo-traded? Staying with the analogy, if you have a complicated, multi-city flight to buy, you might choose to call the airline directly. Similarly, if you have a large, complicated trade to make, you go to a “high-touch” trading venue where institutional traders work to find liquidity that is not publically posted.
Looking ahead, how do you see metrics and electronic trading advancing within the market?
We are looking to leverage more and more technology to utilize real time, accurate indications of stock direction to fine tune our algorithms. I foresee a more “proactive” algorithm that utilizes machine learning to better understand trader behavior. Currently, we are using real time analytics to show the trader what is happening with their order and, within minutes, we can react and adjust the next order’s behavior. I see this self-correcting process eventually being automated.
Lindsey will also be speaking at…
Sign up for your free account:
Don’t miss out on…