8

Vision Six – Virtual Trading and Banking

The future uses for Spatial Computing in the financial industry are vast. There is very little Spatial Computing currently being utilized in that field, however. The main reasons for this include the general conservatism of the industry and the number of regulations that would need to be addressed relating to customer interfaces and trading. We believe that the great benefits that Spatial Computing could bring to the financial industry will enable the hearty adoption of its technologies in the future.

Since almost all of the views for the use of Spatial Computing in the financial industry included here are forward-facing, this chapter will be much shorter than the other six chapters we have written that address industry verticals. The functional areas we touch upon here include 3D data visualization, virtual trading, ATM security and facial payment machines, and virtual branch functionality and customer services. Let's start off by discussing 3D data visualization.

3D Data Visualization

One area where a start has been made is in data visualization. A company that has focused on using Virtual Reality as a medium to visually present data is a start-up called Virtualitics, based in Pasadena, California. Over the last year or so, it has added a demo using Augmented Reality as well. Many of the management team and employees went to Caltech; their understanding of data science is extremely good.

A major focus for the company has been to cater to financial clients that need to have complex data represented in a more visual way in order to simplify finding the possible solutions they are looking for. However, the software that Virtualitics has built is not specially built to handle certain problems or industry verticals―it is very open-ended.

Another company that had a 3D data visualization product, albeit at a very early stage, was a company called Looker. In 2019, Google announced it was acquiring Looker for $2.6 billion.

There are skeptics we have spoken with who are very enamored with the uses of Spatial Computing, but they do not see the usefulness of an open-ended 3D data visualization product. Since there are no companies that specialize in 3D data visualization products just for the financial industry, a big opportunity exists in this area. Virtual trading, that is, trading that would happen with a person using a VR or AR headset, cannot happen without these kinds of specialized 3D data visualizations. Any company, whether it be a bank, a hedge fund, an over-the-counter options and futures trading house, or any other financially-oriented company that wishes to do virtual trading, would need to have programs that have 3D data visualization capabilities integrated into them.

Currently, many kinds of trading that are not automated are accomplished only after a trader has looked at several sources of information and come to a decision. Most of the time, these traders have to view all the different kinds of information across many screens as simultaneously as possible. Due to this complexity, the rate of human error can be quite high. With the right kinds of 3D data visualization software capability, the numerous screens could be reduced to just one 3D representation, which would enable a trader to make a faster, more informed decision about a particular trade.

Photo credit: Bloomberg. A typical hedge fund trader's desk.

At this time, traders tend to use 19-inch monitors, in multiples of 4, 8, or 12. What is interesting is that there are still innovations being developed for traditional computer monitors. During CES 2020, Dell came out with a new 43-inch monitor aimed at traders that need to track multiple tickers or charts simultaneously.

However, with 3D data visualization software, multiple charts (literally, an almost limitless number of charts) could be viewed and manipulated in both Virtual and Augmented Reality. These could be placed side by side, tiled from top to bottom, tiled backward and forward in space, and in combinations of these at different angles in space.

Photo credit: Dell. Dell's new 43-inch monitor geared toward traders.

In addition to virtual trading, 3D data visualizations would also be useful in portfolio management since the intersection of many different kinds of information is often needed to make a decision there as well. Since much of what could be applied to portfolio management could be used for virtual trading, with even more specialized visualization being used there, we will focus on virtual trading in our next section.

Virtual Trading

There has been much talk among tech and financial insiders about how Spatial Computing, especially Augmented Reality, could be applied to virtual trading, but very few tests so far have been done yet with actual traders. Both Bloomberg and Reuters experimented early on with initial mock-ups for trading desk software using Google Glass and then, more recently, using a HoloLens headset. We've heard that a mid-2018 test that Reuters ran using a HoloLens on 30 or so traders was not successful, because they found the headset to be too clumsy and not wearable enough.

The HoloLens 2 could very well start to solve this issue, as well as future iterations and other headsets―such as those that Apple has leaked―that will be coming out in 2022 and 2023.

In 2016, the consultancy 8ninths (now a part of the Valence Group) was engaged by Citibank to provide a proof-of-concept mock-up for a trader's workstation that utilizes AR and a HoloLens headset. This proof-of-concept became very popular, with many believing that an actual software app had been developed, but it had not been.

Visual aspects of 8ninths' proof-of-concept mock-up include 3D bubbles of individual trading vehicles (these could be stocks, options, and so on) of a given market. The idea is that when a trader clicks on a bubble, they will be able to access its up-to-date trading information. 3D tabs appear to be able to be tiled in space and could provide all kinds of other information, including news. There is also a space for 3D data visualizations.

What is not evident in the mock-up is how a trader would be able to execute a trade using AR software and an AR headset. It also appears to us that there could be consolidation among the kinds of information that an individual 3D object could offer. For example, clicking on a 3D bubble could present both trading information and access to news, as well as other relevant information, such as access to a list of a financial vehicle's competitors that could be clicked on from there.

What does work very nicely is the idea that the size of a 3D bubble, as well as the color of it, could indicate different discrete features, such as the relative market size and industry vertical, respectively. Other novel ways could be integrated into the user interface so that many other types of features could be easily visually evident.

Photo credit: 8ninths and the Valence Group. This is a proof-of-concept workstation mock-up of what a trader could potentially access using AR and a HoloLens headset; developed for Citibank.

Besides wearing an AR headset, another possibility (though a little further in the future, by three-to-four years) is wearing smart contact lenses made by Mojo Vision, called Mojo Lens. Mojo Vision, a Saratoga, California start-up with dozens of PhDs and engineers, came out of stealth mode in January 2020. Even though the Mojo Lens is initially being marketed for healthcare purposes, many other kinds of use cases are very evident. The company calls what you are able to do with their lenses "invisible computing." Mojo Lens would be able to make trading even more efficient―no need put on and take off or swivel up a headset, and eye-tracking and voice capabilities would make information gathering and trade fulfillment faster and more intuitive.

Using an AR headset or Mojo Lens and specialized software, all kinds of trading support could be accomplished, which includes equities, options, futures, REITs, commodities, and any other tradable financial instrument. Where the most benefit would be gotten, though, are those instruments that require the analysis of many different streams of information in a fast manner before making a decision to trade. These include options and futures, both at times when decisions have to be made as to whether or not to purchase the instrument and also as to whether or not to "roll forward," that is, to extend the expiration or maturity of an option or futures contract by closing the initial contract and opening a new longer-term contract for the same underlying asset at the then-current market price.

Photo credit: Mojo Vision. Mojo Vision's Mojo Lens feature roadmap for the 2020s.

The amount of information needed for these decisions would normally necessitate the use of many monitors. The stress of having to consult these monitors and then have to make very quick decisions as to what to do has worn down many traders. Before reaching trader burnout, traders can make very costly errors. Using an AR headset or Mojo Lens with software that supplies 3D data visualizations with plotted relevant points based on criteria input previously could cut down on burnout and errors. News and information could be pulled up alongside the 3D charts, with elements there added as criteria for the charts themselves. Actual trade order execution could also be achieved.

Some delay in using AR headsets and/or Mojo Lens and specialized software for information and decision-making, as well as trade execution, will come from financial regulatory authorities. Getting their "OK" will take time, because of the potential risks involved. If information from Reuters, Bloomberg, or any other company that provides financial quotes and news is not updated in exactly the same timely manner in the new software as it is in existing systems, then the new system will not be reliable. And if trade execution lags by even a millisecond compared to existing systems, there would be a major issue. The new software would have to match with existing systems when running tests; it would have to match at all times in every instance.

Having said this, there is no technical reason why systems using AR headsets and Mojo Lens would not be able to match existing systems' reliability. We believe that the future of trading is one that relies on Spatial Computing.

Management from the New York Stock Exchange, Nasdaq, and the hedge fund Bridgewater Associates expressed interest in our book but declined to comment. We know of at least one very large global financial services firm that has a great interest in Spatial Computing and has started developing technology for it, but we have been asked to keep it confidential.

Different aspects of Spatial Computing, including Computer Vision and associated AI, make the next section we are focusing on in this chapter possible.

ATM Security and Facial Payment Machines

It is now commonplace to unlock smartphones and computers using fingerprint and/or facial recognition. In some countries outside of the US, such as China and Spain, facial recognition has already been implemented for ATM security.

When we spoke with Brett King, the author and founder of the banking service company Moven, he became especially animated when talking to us about the developments happening in China. In China, facial recognition is even being used to pay for vending machine items. A TikTok video from August 2019 shows a woman doing just that. Chinese technology expert Matthew Brennan has reported that he found another TikTok video that shows that non-Chinese people are blocked from using that facial recognition payment system, presumably because they don't have IDs issued by mainland China.

Photo credit: TikTok. A still from a video uploaded to TikTok that shows a woman paying for a vending machine item using facial recognition.

In 2015, China Merchants Bank launched an ATM that uses facial recognition, eliminating the need for the use of cards. It was the first time that this had been made available for public use. The company claimed that the system would be able to distinguish between twins since several parts of a face were used to create the profile used for facial recognition comparison.

Wearing glasses or different makeup was tested and shown to not make a difference.

Photo credit: China Merchants Bank. An advertisement for the China Merchants Bank that tells users they can now use their face to withdraw money.

Several banks in mainland China now, including the Agricultural Bank of China, China Construction Bank, and China Merchants Bank, have ATMs that use facial recognition as part of their security process.

Taiwanese bank E.Sun partnered with NEC to provide ATMs that use facial recognition; the ATMs were made available to the public in February 2019. In order to set up facial recognition cash withdrawals, customers must first insert their cards and then have a photo of their face taken with a camera installed within the ATM. They are then sent a one-time password to their mobile phone, which must be entered into the ATM within a minute, thus registering their facial image.

In Spain, in February 2019, CaixaBank started offering ATMs that use facial recognition. Trials conducted among customers before indicated that 70 percent would use it as an alternative to entering their personal identification code on a keyboard. Registration for the facial recognition system is conducted in a bank branch and takes a few minutes using tablets.

Photo credit: CaixaBank. A CaixaBank ATM that uses facial recognition instead of a card to access its system.

Outside of banking and ATMs, facial recognition could be used to pay for products. Again, in China, this has already been achieved with "facial payment machines."

Mobile payment in China is already the most advanced in the world, with customers making a purchase simply by posing in front of point-of-sale (POS) camera-equipped machines, after linking an image of their face to a digital payment system or bank account.

With devices in already over 300 cities, Alipay, the financial arm of e-commerce giant Alibaba, has been leading the facial payment machine business in China. Its facial recognition system comprises a screen that is as about as large as a mid-sized iPad, its AI software and Computer Vision components, and mobile payment software called Dragonfly. Alipay has earmarked $420 million over the next three years to roll out its technology.

Photo credit: Alipay. Pictured here are two Dragonfly facial recognition systems, which are currently actively used in China to make mobile payments.

Additionally, in August 2019, Tencent, which runs the WeChat app with 600 million users, announced its new facial payment system called WeChat Pay. Besides Alipay and Tencent, a growing number of start-ups in China are focusing on the facial payment machine business.

Facial recognition technologies can also be used to facilitate virtual banking, which is the topic of our next section.

Virtual Branch Functionality and Customer Service

Second Life, the virtual world, shut down all the virtual banks that had been built and operated there in 2008. Now, almost none of those banks were real financial institutions, but rather individuals who had deigned to create a bank out of thin air and put it in operation among the denizens who populated Second Life. However, Second Life virtual banks actually had deposits made with real money, and ultimately, they were shut down due to a lack of regulation.

Photo credit: Second Life. One of the virtual banks that operated in Second Life before they were all shut down.

Nowadays, talk of a virtual bank refers to a real financial institution that operates without any physical bank branches―basically, an all-digital bank. A good example of this is Simple, which calls itself a "neobank."

CaixaBank, in 2017, became the first bank in Spain to incorporate the iPhone X's Face ID technology so that customers could access their accounts solely using facial recognition on their mobile devices. Now several banks around the globe offer this capability.

With many of the major banks around the world closing down physical bank branches due to a lack of customer foot traffic, there has been some movement around the idea that all bank branch functions could be digitized, such as those associated with opening up a first bank account and mortgage paperwork.

Shaun Moore, cofounder and CEO of Trueface, a company working with top global banks on facial recognition uses, has commented that "We are seeing the financial services sector test face recognition as a part of multi-factor authentication for ATM withdrawals, mobile banking enrollment, and mobile account access and transactions. By implementing face recognition as the key step in multi-factor authentication, banks are able to mitigate their exposure to risk and fraud, saving themselves millions of dollars in the process."

Along with bank services being available post-facial recognition check are those available using voice technology, with USAA being the first to offer limited voice banking. Capital One offers banking by voice via Amazon's Alexa, on devices like the Amazon Echo and Amazon Dot. Capital One customers can check balances, review transactions, make payments, and more, using simple voice commands. Bank of America's "Erica" works similarly, but off of its own app, versus using Alexa-reliant apps.

What we envision is virtual banking that combines the use of facial recognition and voice commands and is accessed via an AR headset and also on Mojo Lens (though, features that would enable that to happen haven't been built yet, but are expected to be). This would also allow both virtual AI assistants and real bank customer service reps to be tapped by the customer on demand when needed.

Future Adoption

Even though the use of Spatial Computing technologies for virtual trading and banking is at such a nascent stage, we are confident that given the great benefits that Spatial Computing could bring, these technologies will be adopted on a large scale in the future.

Spatial computing has the potential to make trading and banking functions take less time and make more money―particular outcomes that are shared with other industry verticals, including manufacturing, retail, and healthcare.

In our next chapter, we turn to how Spatial Computing could be used for education and training across several industry verticals.