Home » How big data has changed business for good

How big data has changed business for good

by admin

Artur Deliergiev, Algorithmic Quantitative Trading Manager at the company CMC Markets, looks at the changes automation is bringing to the trading platform and why he sees it as a step in the right direction

Over the last 20 years, automation has brought about a real evolution in determining the structure of the modern trading platform. Some commentators have expressed concern that this “changing of the guards” — seen as vocally established traders largely being pushed out by hordes of quantitative analysts building sophisticated models to manage risk — leaves liquidity providers exposed. Consequently, as monetary policy begins to normalize and we see an accompanying rebound in sustained volatility, questions have been raised as to whether rocket scientists will find themselves out of their depth, unable to outsmart counterparties and risk precipitating a market-wide systemic failure. . Artur Deliergiev explores what has changed – and why he believes it’s all for the greater good.

So the business floor has changed – and irreversibly?

Artur Deliergiev, CMC markets

Artur Deliergiev: Yes – it is. With one or two exceptions, the idea of ​​a dealer barking into multiple phones – or simply directly at his colleagues – while staring at a series of screens has been consigned to history. Looking back, the inefficiencies were staggering and, as with any human-based process, the risk of error was significant. However, it should be noted that this did not just happen overnight – it is the result of a culmination of aspects that have been manifesting for decades. But automation – along with advances in technology – means we are now able to accurately process significantly higher transaction volumes than would ever be possible in a voice-based world.

How is your team set up to ensure that you take advantage of all available data?

Artur Deliergiev: The key point is that historically, trading boards have been supported by analysts. The traders called the shots and the analysts’ response was to listen to the available data as best they could. This situation has now fundamentally changed because we now have quantum traders – traders who can code – and they are in turn supported by a larger team of quantum researchers. The fact that they share a common language means that both parties understand each other, work incredibly well together and can solve problems much faster than before.

We constantly work with a significant historical database at our disposal and current prices, as we receive more than 2 billion ticks per day and a book of outstanding risks. We use this information to continually improve the risk management strategies we put in place to deliver the best outcome for all stakeholders. Achieving this goal still requires some degree of traditional relationship management, but this now comes at the end of the process rather than at the beginning.

What happens when monetary policy normalizes and the market revives?

Artur Deliergiev: There are certainly people who believe that complex models designed by quants will struggle when volatility accelerates, but there are two key reasons why we can say with confidence that these ideas are misguided.

First, the underlying technology is so advanced that it allows us to have pricing relationships with many different parties—far more than individual sellers could ever maintain in real time.

We’ve spent years fine-tuning these models to ensure they can adapt to market changes.

Second, the automation aspect allows us to not only look at the underlying cash market, but also synthesize prices from futures and options in real time and then decide what kind of liquidity we are willing to make available.

This has already been tested with the impending dislocation in the gold market since March 2020, when a number of original liquidity providers were unable to set a price. A small group of tech-based non-bank liquidity providers – incl CMC Markets – had access to sources of prices which in return offered a reflection of the physical market – allowing prices to be maintained.

Comparison of threatened dislocation with equivalent removal performance CHF/euros Peg in 2015 shows once again how far this market has progressed in recent years.

With this extraordinarily long period of low volatility, how can you be sure that the models you’ve come up with can handle it?

Artur Deliergiev: We have a significant amount of confidence across the industry that the work that has been done so far will be able to respond to the unpredictable movements of the market. It’s all about deciding how to manage risk in our book, and ultimately automation doesn’t care if the move is 25 pip or 250 pip. There is no danger of the technology being spooked – it simply evaluates the situation and reacts according to preset parameters.

It’s also important to remember that we’ve spent years fine-tuning these models – CMC Markets have been creating item-by-item datasets for each instrument for the past decade. So by constantly monitoring our price performance and using it to expand our knowledge base, we have a really robust system.

We’ve also made several significant changes to our infrastructure that further reduce the latency of our traffic – notably by co-locating servers in financial centers and distributing pricing resources via the cloud. As the first brokerage to introduce live pricing to Amazon Web Services, we find that this is particularly attractive to new counterparties who value lower setup costs, ease of integration and the ability to bypass global data center requirements.

How much is retail order flow helping or hindering your ambitions?

Artur Deliergiev: I think it’s absolutely critical and gives us a significant advantage. For too long, the prevailing view has been that retail order flow is insignificant and, as such, unnecessarily distracting to those who process it. This couldn’t be further from the truth.

Not only does the internalization of flow at some of the original liquidity providers reduce market depth, but retail order flow provides us with additional liquidity at price points that the institutional market finds very attractive. It also means we interact with more market participants, which helps price formation – especially when we see abnormal market events, both at the top of the book and further down the ladder.

Finally, what advice would you give to a fund, manager or broker looking to build their quantitative analysis function?

Artur Deliergiev: The key point to remember is that it takes a significant amount of resources to successfully do this yourself. To get started, you’ll need basic data to work with, and — especially if it’s from the stock market — it’ll have a price tag attached to it. Then you need the capacity to interpret that data, build appropriate models and understand how to deploy it within your risk management or business strategies.

To this extent, we assist a number of institutional customers with these services, and our fundamentally different approach here means that counterparties who work with us always find that they benefit not only from our combination of huge volumes of data, but also expertise. Combine this with our consultative approach – rather than a one-size-fits-all approach – and the overall proposition is truly powerful.

Source Link

Related Posts