Technology

Why Behavioral Finance Needs Great Tech To Thrive

Tom Burroughes, Group Editor, May 13, 2022

articleimage

The field has been winning wealth management fans and generating plenty of noise. To make its insights bear fruit, lots of data needs to be handled. That means technology – lots of it. We talk to a figure working in the space.

There’s a convergence between the field of behavioral finance and Artificial Intelligence, as wealth managers use technology to digest data that new investment ideas generate. 

As already reported here, behavioral finance is gaining wealth sector fans. The term applies to understanding how people mistake portfolio gains by pure skill rather than also accepting the role of chance, or treating losses more emotionally than they do with gains, following crowd behavior. These insights, which draw on views about how humans have evolved from pre-history, are used to explain events such as stock market booms and busts, share trading frenzies (such as over GameStop in the US more than a year ago), or the regular gyrations of bitcoin. The pandemic, Russia’s invasion of Ukraine and a spike in energy prices have given plenty of reasons for emotions to hold sway in markets. 

It is hoped that, armed with these insights, people will make fewer mistakes. And they can apply ideas to tech tools to provide “guard-rails” around their portfolios. That, at any rate, is the aim.

As far as Edinburgh-based Sonia Schulenburg of Level E Research is concerned, behavioral finance is an important field. But without the necessary tech tools, it is hard for investors to use the area’s insights wisely and achieve their goals.

“I think it [AI] is empowering people….it is going to automate processes….a lot of routine activities that we have to do every day, reporting, gathering data, looking for similarities and looking for solutions,” she told Family Wealth Report in a call. 

“From the Spiderman movie, we can say that `with great (high-processing) power comes great responsibility.’ I love this quote. Computers and good algorithm design is allowing us to do hyper-customizations that we could not dream of doing years ago. Clients have many different objectives and constraints and traditionally clients are placed in three to seven buckets depending on their age and risk preferences. This has to change. We have sufficiently developed AI to allow unique investment solutions driven by proper client optimizations,” she said. 

One of the main advantages is that such technology can save managers’ time and allow them to focus more on other client needs, Schulenburg continued. 

A big challenge for AI is how systems can offer explanations for the decisions taken. “Financial markets are very dynamic systems and you need technology that can explain decisions,” she said. 

Mainstream
As reported here recently, behavioral finance is now being used as a marketing and branding strategy. For example, behavioral finance is the “dominant differentiator” in the philosophy and marketing strategy of Fusion Family Wealth, a Long Island, New York-based wealth manager, according to CEO Jonathan Blau. The giant software vendor Orion issued a white paper Don’t Call it a Fad: Behavioral Finance is the Future of Fintech, along with an on-demand webinar on behavioral finance’s role “in the future of our industry” by Dr Daniel Crosby, the firm's chief behavioral officer.

Without the tech to gather the insights and put them together to work, however, the “cash value” of all these theories will not be achieved. 

Schulenburg took a step back to review where the subject came from. “2008 highlighted the limitations of classical economics and its associated models of both the economy and risk. Central banks knew their tools had failed. This failure gave rise to new ideas being developed in areas such as Complexity Economics, Agent-Based Modeling (ABM) and associated areas of Artificial Intelligence (AI) and Machine Learning (ML),” she said. “More specifically, within this realm, our data-driven approach uses AI/ML to identify trading strategies and Agent-Based Modeling to adapt them to market regimes. In other words, strategies that can exploit evolving trading patterns in security prices. We use Agent-Based Modeling to adapt those strategies to market regimes, where ABM considers the market as an ecosystem of continually evolving `agent’ investors.”

To illustrate, Schulenburg said she finds “core” investment strategies to exploit inefficiencies in short- and long-term market dynamics – such as periods of higher market volatility and changes in direction that create opportunities in the short term, where there are steady and trending periods with lower volatility – can be exploited by long-term approaches.

When it comes to trying to automate strategies to make money, investors might naturally worry that “circuit-breakers” or other in-built tools are needed to stop a strategy running amok. 

“Because of high levels of automation, we would expect that systematic strategies (including the mean reverting and trend following named above) would have breakers to avoid extreme market behavior and events such as fat finger trades,” she said. (By “fat finger” she means when a trader hits the wrong button on the keyboard, with subsequent mayhem.)

“Other types of breakers are designed to prevent getting carried away and keep trading constantly in big market moves. These are typically part of systematic execution strategies where algorithms are in place to control the trades. And, in addition, risk systems will typically take care of these types of extreme events,” she said. 

“For example, in extreme events (as market prices are dropping intra-daily or over a few days), a trend-following strategy will start selling a stock as the price drops and crosses certain metrics (this reinforces the drops), and in the case of a mean-reverting strategy designed to buy stocks at cheaper prices, the strategy starts buying cheap bringing some liquidity and stability to the system. However, if prices continue to fall sharply, the strategy has to stop buying at some point,” she said.

“Regarding fat finger trades, examples include a 'flash crash’ in May 2010, and more recently we experienced a flash crash in Europe where it was reported by Citi that one of their traders booked an erroneous order in the Nordics. This had a ripple effect in Europe. These errors are mainly originated by people and seem to happen every other decade,” she said.

Not a taboo any more
Schulenburg argued that acceptance of AI tools is growing and is not seen as a weird or sinister development any longer. 

“AI has had a significant impact over the last few years in the asset and wealth management sectors,” she said. As costs continue to rise and managers are under pressure to keep fees down, they need to look at AI as a way to help with running portfolios as efficiently as possible. “We are trying to help the end-investor….it is a noble goal and an innovative one.”

Register for FamilyWealthReport today

Gain access to regular and exclusive research on the global wealth management sector along with the opportunity to attend industry events such as exclusive invites to Breakfast Briefings and Summits in the major wealth management centres and industry leading awards programmes