Machine Learning Finance: How AI & Machine Learning Have Impacted Businesses Around The World

The article was written by David Shabotinsky, a Financial Analyst at I Know First, and enrolled at an undergraduate Finance program at the Interdisciplinary Center, Herzliya.

Machine Learning Finance

Without big data analytics, companies are blind and deaf, wandering out onto the web like deer on a freeway.” – Geoffrey Moore 

Today, through advancements in technology, both consumers and businesses are experiencing massive changes in not only our daily routines, but the way we think about statistics and data that the average person produces.

Summary:

  • How Artificial Intelligence has become ‘Game Changer’ For Businesses
  • Businesses around the world, from every industry are beginning to integrate AI
  • The financial sector has been deeply impacted by this major shift in technology
  • Japanese innovation is helping AI reach emerging markets
  • The costs of not adapting to change and looking forward
  • I Know First has a state of the art algorithm bringing real competitive advantages

Background: 

The hot topic on everyone’s mind today revolves around Artificial intelligence (AI) and how it is shaping our world today. Many are worried about a terminator like future where AI using machine learning will end human existence. While this scenario is more likely to be found in a sci-fi movie than in reality, machine learning and AI are taking tremendous strides today to change the world we live in and how businesses interact with consumers in the new information era that we live in today.

Machine Learning Finance

Machine Learning: An Advanced Form of AI

Machine Learning, a specific category within AI, was first developed in the late 1990s as a result of digitalization and further technological developments. In essence is a wide range of algorithms that are able to learn from data, on their own, and apply to certain tasked based on programmed rules. Today, most firms have begun to utilize this type of technology and use cheap computing power mixed with big data to extract valuable information for the companies. This development in the era of digitalization has gone as far as to even effect incumbents from traditional industries, such as General Electric, a company that is still a part of the Dow Jones Industrial Average after 119 years. They are now generating hundreds of millions of dollars in revenue, using data that they collect from deep-sea oil wells or jet engines to optimize performance, anticipate breakdowns, and streamline maintenance.

This all the more so true if one takes a look at the M&A activity by the technology industry, which is increasingly acquiring more and more AI based startups. Google, Baidu, and Amazon, all tech giants, have been purchasing new start-ups that use machine learning and AI to better garner more consumers towards their respective platform/product. As explained, “AI is about to transform industry after industry,” said said Andrew Ng, a deep-learning pioneer who is chief scientist at Baidu Inc. and an associate professor at Stanford University.

AI Integrated Into The Modern Businesses World

The specific category in AI most businesses utilize is based on Cognitive computing, which self-learning algorithms simulate human thought process through data mining. This allows businesses to better manage consumer relations and customer service, for example during incoming calls from customers these algorithms can help manage customer responses, thus reducing cost and increasing efficiency in the workplace. This especially prevalent amongst retail, telecom and financial service industries, which take thousands of calls every day and need to be able to manage them properly.

The main effect is through the utilization of Big Data, which is simply large sets of data that is analyzed by computers to reveal patterns that are collected by organizations to use to better analyze the world we live in. It is famously broken down into “4 V’s”. Volume, referring to that the massive amount of pure data available, just think that in a world of 7 billion people, 6 billion have cell phones. Variety is the different types of data that are available, ranging from structured to unstructured date in the world. Veracity referring to the whether the data collected is real and this requires further research to figure out what data we can trust and use. Lastly, velocity is the frequency of data that sent every second of the day, i.e. credit card swipes and/or text messages sent to your friends and family.

Machine Learning Finance

Source: World Newsmedia Network (http://newsbizblog.blogspot.co.il/2013/11/big-data-four-vs.html)

With regards to machine learning based off of AI, the average person’s life is greatly affected in a direct manner today. Although it helps with for example detecting credit card theft, a much larger effect is how businesses use it to better match their own product/service to the appropriate person, and able to capture their actual ‘wants’ in a product. Businesses collect big that people produce using technology, i.e. online sites, and that data is then utilized in Business Intelligence. This translate into different software applications of big data that is used in the businesses world. For example, Apple bought the then-startup Siri in 2010, which is now renowned by consumers as the voice behind the iPhone and helps users navigate the internet using AI. Although many people think that technological advancements are only enjoyed by wealthy businesses that can afford these changes, one of the main benefits in this technological advancement is that because of its cost efficiency nature it is able to be utilized by the majority of small businesses today, who can easily afford this adoption and save time and money as well. For example, firms today use SaaS, or software as a service, which helps businesses manage daily operations such as content, payroll, and other cloud services. This type of big data type software is a further utilization of AI and allows for small businesses owners to not only manage operational tasks faster and cheaper than humans but as well manage client relations. As a small businesses owner one of you biggest concerns is about managing your customers as a premium service you must offer is better relations and care than larger corporations. As a result, AI has enabled CRM (consumer relation management), a type of SaaS to help mainly small business owners to maintain their crucial relationships. They are now able to focus more on business development as opposed to remembering client meetings and smaller operational related tasks.

As a result of all these technological developments, more and more small businesses are able to open their doors as the costs to run businesses have immensely dropped over the last decade. This has allowed many emerging markets to fuel businesses development as well provide middle-class families the capabilities to run cost-efficient and consumer friendly operations.

Global Development of AI In Asia

This new digitalization era is not only limited to America national firms, but it is being advanced throughout the globe. Across the pacific, Japanese companies are developing AI-based technology that can further enhance the way people live in their own daily lives. For example, in Tokyo, Kabuku Inc. has developed AI technology that allows outsourcing of 3D printing to be more operationally efficient, by increasing production speed using an open system that utilizes AI technology. Toyota Motor Corp. has even partnered with the company to allow consumers to personalize color and design choice on exterior parts of an ultra-compact electric vehicle. The main benefit that Toyota sees in the startup is the manufacturing flexibility offered through a high outsourcing speed.

In addition to allowing personalization in car manufacturing, exMedio Inc., a startup in Japan has developed new medical technology, utilizing AI, to help doctors and patients recognize diseases and treat them quicker with reliable medical expertise. They developed an application service that allows doctors and patients to send photos of infected areas, and then receive real medical advice in return. The development of this technology has immense potential to provide necessary medical attention to emerging markets, specifically in Asian countries where there are shortages of doctors. This incredible work shows how countries besides for well-known first world countries, i.e. the U.S. are able to directly benefit from the advancement of AI and machine learning. Whereas during other technological advancements, such as during the Industrial Revolution, emerging markets were far from being the first to benefit from the radical changes.

This Japanese innovation is not limited to the private sector, as Japan’s government as well recognizes the power of AI-based technology, they are developing technology for the pharmaceutical industry in Japan to be able to develop better drugs. This would essentially utilize data mining throughout all of Japan’s recourses regarding conditions and various diseases. The health ministry hopes to increase Japan’s competitive advantage with the global health market, as they are currently behind many peer countries with R&D into new effect drugs. However, they are not alone in this endeavor, as two startups in the U.S. have begun and achieved great feats. Berg, a startup in Massachusetts, has found a new cancer drug uses its own AI software to analyze over 14 trillion data points of cells and cancer tissues to achieve their results. Additionally, Atomwise a startup in California, is as well working on find new drug treatments, and has found already two existing drugs that can fight Ebola. These amazing discoveries in the medical world have all been achieved using Artificial Intelligence.

Besides for the medical world, Japanese startups are as well developing technology that uses AI to help students on more effective studying methods in school. Recruit Marketing Partners Co. in Tokyo, is developing technology that can analyze the behavior of students and then be able to pinpoint exactly what the mistakes were. This can then be used to determine what steps to take to improve their educational performance skills.

Machine Learning Finance

AI & Big Data Applied to the Financial Industry

The most valuable commodity I know of is information.” – Gordon Gekko, Wall Street

Within the Financial Service industry, AI has been immensely influential in helping expand profit margins and attract more consumers. Not only are banks and credit card companies able to more effectively detect financial theft, banks and other financial service companies now offer online financial wealth management advisors, offering consumers low fee-based models to manage retirement plans and other portfolios. Fintech companies as well range from their usage of AI as they are able to offer lending solutions to clients seeking mortgages and other loans opportunities, and others help manage budgets for the average consumer. For example, Sarah, a millennial can budget her newly found money at her new job, as well as finance her own loans, and manage her investment portfolio, all on her smartphone. Those who remember, just a few decades ago, in the late 80s, this was only a reality in sci-fi movies, but now their children can do this with ease.

Additionally, AI based machine learning has been able to recognize patterns that further help businesses, specifically in the finance industry identify trends in the marketplace. Today, many on Wall Street have come to accept that although the market is neither completely efficient, as explained by the Efficient Market Hypothesis, it is neither Chaotic at the same time. Rather, it is a mixture of both of which it is a complex system. To understand more about the construction of a complex system, click here. As a result, the financial industry, specifically money managers, need to be able to maintain their high alphas above the benchmarked passive investments, i.e. ETFs that track the S&P 500 Index. Currently, on average these active fund managers underperform the market, after fees, they have thus begun to turn to algorithmic trading in an attempt to achieve the expected high returns. There are essentially two forms of AI-Based algorithms. There is high-frequency trading (HFT), and self-learning algorithms that use machine learning. HFT has mixed opinions on its ethical usage, as well as known to be highly expensive. Machine learning algorithms, on the other hand, are able to create patterns using empirical evidence and big data and eliminate “noise” from the market to create market trends. “Noise”, a terrible sight for investors, refers to short-term (daily or intra-day) fears, worries, and negative fueled perception regarding the price of a security or general market atmosphere. By ignoring it once is able to identify trends in the market. The AI-based algorithms are able to adapt as a result of neural networks built to allow for deep learning, which then allows the algorithm to adapt accordingly. More on how specifically it goes about adjusting can be found here, in a past I Know First article.

I Know First Competitive Advantages

Today, many self-learning algorithms are constantly being developed, however, I Know First has a state of the art algorithm that has a clear distinction from the rest of the market at hand today. The algorithm, was developed by Dr. Roitman, who has a vast 35+ years pf experience in the field of AI and machine learning. Every day, the self-learning and self-adjusting algorithm produces market forecasts with trends of stocks, commodities, and indices over 6 different time horizons ranging between a few days and a year. The kind of analysis and pattern recognition that the system does each day could never be accomplished by a human in any amount of time. On August 25, 2016, an article was published detailing the competitive advantages that directly apply to investors.  Through a mixture of chaotic system and efficient patterns found in the market, the algorithm is able to successfully detect trends to enable investors to understand entry and exit points in investing.

Adapting To Change

In addition, I Know First’s research team has recently developed a short term trading strategy model, based off of daily stock selection using the self-learning algorithm. The strategy involves filtering stocks made up of the S&P 500 Index, of which is then broken down into five separate smaller strategies. The two main overall filters that are applied are based off the heat map, that algorithm produces for each stock it forecasts. A more detailed explanation can be found here. The two filters are the predictability indicators (algorithms confidence level) and the signal strength (forecasted volatility from spot price). Afterwards, there then five separate filters applied to make up smaller strategies, which have all achieved high alphas above the S&P 500 Index, which can be found below. The highest alpha was at 40.02%, relative to the S&P 500s return of 9.08%, from January 7th, 2016 to August 31st, 2016. More about this trading strategy can be found here.

Machine Learning Finance

Costs of Not Adapting To Change 

Besides for the ample benefits of incorporating AI-based technology into the workplace, like many technological adaptations, AI does come costs. However, the costs can be considered an inverse cost, as it is associated with not using and adapting to AI, today. Many if not most of the Companies that did not adapt to having an online presence during the information era, during the 90s, have ceased to exist. This adaption to the cultural and technological times, is how great firms continue to exist for decades, and those that are too stubborn and/or too hubris to change their ways, will eventually see their own downfall. Although one can only speculate about the exact future, empirical evidence does support this idea of adaptation through the different industry lifecycles and eras. Governments around the world, i.e. the U.S. are even replacing military technology to adapt to AI, and are spending billions of dollars achieve this. Companies that do not constantly innovate will be left behind by those do take the risk and not play it ‘safe’.

Looking Forward

Most of AI today is preprogrammed and thus the design is only narrow to that of its command limit. However, technology giants such as Alphabet are developing such systems, that utilize deep learning, or Google’s DeepMind and IBM’s Watson. These programs are designed to learn automatically, and thus their capabilities are go beyond the imagination. This new type of technological shift, can even lead to a new industrial revolution, further atomizing repetitive functions, as the past revolution had began in the 18th centaury.

Conclusion

The world that we live in today, has drastically changed from the ones our parents and grandparents grew up in. Today we live in an era of digitalization which has allowed businesses to grow much quicker and as technology as helped them become more efficient and reduce costs. Although these advancements especially in AI and computer optimization have and will cause job loss, i.e. reduction in bank tellers or customer service representatives. This change is similar to how factory workers have been replaced in recent years; and like any technological progression in history, it creates a new job market as well, that in the long term will benefit the consumers and average household. Barriers of entry of course still do exist for these emerging companies, as traditional managers historically lack the desire for ‘change’ in the workplace, due to their risk adverse nature. For example, while firms such as Blockbuster have gone bankrupt, as a result of online streaming, it gave rise to new businesses like Netflix. However, the overall benefit to the market by providing cheaper consumer and businesses friendly products is evidently shown in today’s world. Firms need to recognize these changes and adapt quickly to survive in the long-run.

 


premium