resources Article

Bringing the Past
to Life

Understanding the past can tell us something about what might happen in the future, but in complex adaptive systems, such as financial markets, just understanding the past might not be enough. Nevertheless, for a long time, that’s the best that we could do. It gave us a murky but workable vision of the future.

These days we can do better. Regression-based models for predicting stock market behavior are flawed and no longer deliver the best results. In an environment that is more challenging than ever before, banks must turn to the latest technology to produce accurate simulations of future market behavior.

Why historical data belongs in the past

Models that rely on historical data struggle to predict events that haven’t happened before. Events such as flash crashes and unprecedented market patterns are rare but that’s little consolation if you miss one thanks to a flaw in your model. Or worse, if your model reacts to one by driving you to react in a way that contributes to the disorder.

Often it is not clear that a data set has a blind spot to a period of high volatility until it’s too late. Not only that, but historical data can also be expensive to acquire and clean to the point where it is useful.

In the real world, every trade has an effect on the market, which changes the behavior of other traders. Historical data cannot account for that because it is a static recreation of the past.

Growing complexity means pressure to change

When it comes to attracting order flow, sell-side execution desks face increasing competition and an environment in which it is getting harder to differentiate. MiFID II regulations in Europe have made things harder by preventing the bundling of research and establishing more stringent execution regulations.

The challenge for brokers is navigating the changing regulatory landscape while demonstrating to their clients consistent, reliable performance in a range of market conditions.

Added complexity comes from the regulatory responsibility to prevent algorithms from disrupting the market – and to have a plan in place to detect and contain those that might. There’s an element of self-preservation in this too: an unprecedented market event could turn into an existential threat if the firm is reliant on an algorithm that misinterprets it or does not account for it at all.

To quote more competitive trade pricing and manage regulatory pressures, banks need to turn to simulation, in particular agent-based modeling.

Advanced simulation brings the future to life

Agent-based modeling is superior to relying on historical data because it simulates the behavior of individual traders and other agents in the market, thus allowing firms to model the systemic effects of particular events or trends. How does the activity of high-frequency traders affect pricing for a particular group of stocks? What happens if an exchange alters its execution policies? How will a firm’s own infrastructure respond to stress scenarios?

Being able to run the same simulation thousands of times, with and without an order means that banks can compare the two scenarios and reach an assessment of the likely effect of a trade. The simulation can be run over and over again with different execution strategies, in order to determine the one that is most effective.

Whether banks want to simulate small actions, such as the effect of an individual trade, or something big, such as a change in regulations that affects trading activity, agent-based modeling allows them to do it and have a greater confidence that their prediction of the future is an accurate one.

Previously, banks had to rely on models built from historical data because the computer power needed for a reliable simulation was too great. However, the price of computers has decreased and the processing power that most companies can now deploy has increased significantly, which makes advanced simulation feasible and reliable for the first time.

The result is a far more accurate estimate for the cost of a trade and the ability to mitigate against unintentional herding behavior and demonstrate to regulators that you are not contributing to disorderly markets.

Understanding the past will only get you so far. The brokers that thrive will be the ones that can bring to the future to life.

Chloe Hibbert