Ciao!

Benvenuto alla terza edizione di "Report Review" la newsletter uffciale dello Starting Finance Club Statale.

Il report di questa settimana è in inglese ed analizza il ruolo dell’intelligenza artificiale nel mondo della finanza

The role of AI in financial modeling and forecasting

In today’s modern business landscape, organizations heavily depend on insights extracted from data to guide their decision-making processes. As such, the demand for effective Artificial Intelligence (AI) tools has skyrocketed. Among the plethora of options available, Power BI, Python, and R stand out as popular choices, each offering unique strengths tailored to different needs.

What is needed for finance and fintech? What is the best for financial modeling, running simulations, data science and developing AI trading algorithms? Let’s talk about it! 

Artificial intelligence is one of the hottest topics of current times because it has disrupted most industries in recent years, and the financial services sector is no exception. With the advent of fintech, which has a particular emphasis on AI, the sector has experienced a revolution in some of its core practices. Probably the most affected area is asset management, which is expected to suffer the largest number of job cuts in the near future. From automated analysis of recent earnings-report data to analyzing relationships between stocks and market indicators – and a whole lot more – AI as an asset management tool is emerging as a powerful way for investment firms to meet market performance goals and land more clients.

Machine learning (ML), a subset of artificial intelligence, isn’t just for programming self-driving cars or sorting cat pictures. It’s entering the investment management space and it is emerging. From Siri and Alexa to Amazon and IBM’s Watson, computer programs driven by artificial intelligence draw on massive amounts of data to solve previously intractable problems. Machine learning gives computers the additional ability to learn without being explicitly programmed. This type of AI enables computers to change and learn when exposed to new data.

Machine learning and Investments

In finance, the marriage between technology and investment strategy has birthed a new progeny: Machine Learning investments. ML interprets and learns from data at its core and it is able to analyze market trends, predict stock movements, and offer investment insights with unprecedented precision. ML algorithms uncover patterns and correlations invisible to the human eye, offering a deeper, more nuanced understanding of the market dynamics. It automates complex analytical tasks and adapts and improves its predictions over time, leading to more efficient and effective financial management. From personalized investment advice through robo-advisors to real-time fraud detection, ML is reshaping the finance industry, offering unprecedented precision and insights in market analysis, portfolio management, and algorithmic trading.

Machine learning can transform the way investment strategies are administered by all types of managers. Even the most fundamental, non-quantitative managers will be generating ideas from data that originally was sourced and synthesized via ML. Data isn't just the king; it's the kingdom itself. Neither the technology nor the math behind artificial intelligence and machine learning is new. What’s new is the vast amount of data that’s available now. The quality and quantity of data fed into ML systems directly influence their effectiveness.

Let’s see some practical applications of ML investments:

1. Predictive Analytics for Stock Prices: these systems analyze historical data, market trends, and economic indicators to forecast future stock prices. This predictive power enables investors to make more informed decisions, often with a higher probability of success;

2. Portfolio Optimization: by analyzing historical performance data and market conditions, ML algorithms can suggest the ideal combination of assets to maximize returns and minimize risk;

3. Risk Management Strategies: machine learning for asset management employs sophisticated models to identify potential risks, helping investors mitigate them proactively. This includes market, credit, and operational risks, among others;

4. Attitude analysis in Trading: by examining social media, news, and financial reports, algorithms can gauge market sentiment, providing a unique perspective on potential market movements;

5. Machine Learning for Cryptocurrency Investments: algorithms can analyze patterns in cryptocurrency markets, often influenced by different factors than traditional markets, offering valuable insights for crypto investors.

Forecasting

One of the most important skills for any financial analyst is the ability to forecast future outcomes based on historical data, trends, and assumptions. Forecasting can help you make better decisions, plan ahead, and evaluate different scenarios. However, forecasting is not a simple task, and it requires the use of appropriate tools and techniques to ensure accuracy, reliability, and efficiency. However, when choosing among forecasting tools, there are a few things to keep in mind:

1. Data size and complexity: Excel is a widely used tool for basic forecasting, but it has some limitations in terms of data size, speed, and functionality. On the other hand, Python and R, are powerful programming languages that can handle large and complex data, but they require more coding skills and time to set up and run.  

2. Visualization and presentation: Power BI is a tool that specializes in data visualization and presentation, and it can connect to various data sources and create interactive and dynamic dashboards. However, it may not have all the features and functions that you need for advanced forecasting. On the other hand, Python and R have many libraries and packages that can help you create customized and sophisticated visualizations, but they may not be as user-friendly and intuitive as Power BI.

3. Cost and accessibility: some tools are free and open source, while others require a license or subscription fee. Some tools are cloud-based and can be accessed from any device and location, while others need to be installed and updated on your computer. Excel is a relatively affordable and accessible tool that most people have access to, but it may not be compatible with some data formats or platforms. Python and R are free and open-source tools that can run on any operating system, but they may require more technical skills and resources to install and maintain.

After this brief yet critical overview let’s dive in the most useful and used forecasting tools:

• Excel: is a spreadsheet application that can perform basic calculations, formulas, and functions, as well as create charts and tables. It can be used to create simple forecasts through the built-in Forecast Sheet feature, which uses the exponential smoothing algorithm to generate a forecast based on a series of historical data. You can also use Excel to create more complex forecasts using the Data Analysis Toolpak. However, it has some drawbacks, such as limited data size, slow performance, and lack of advanced features and functions.

• Power BI: can connect to various data sources, such as databases, files, web pages, and APIs, and create interactive and dynamic dashboards and reports. Power BI can be used to create forecasts using the built-in Analytics panel, which allows you to add a trend line or a forecast line to your charts, and adjust the confidence interval and seasonality parameters. You can also use Power BI to create more advanced forecasts using the R or Python scripts, which allow you to use any of the libraries and packages available for these languages, such as forecast, prophet, or scikit-learn. However, Power BI has some limitations, which are dependency on internet connection, limited customization, and lack of integration with other tools and platforms. Looking at the users of Microsoft Power BI based on their industry, the top three industries with the most customers are Professional Services (16.8%), Manufacturing (12.8%), and Banking and Financial Services (8.8%).

• Python: is a general-purpose programming language that can perform various tasks, such as data analysis, machine learning, web development, and automation. Python can be used to create forecasts using the Pandas library, which provides data structures and operations for manipulating and analyzing data, such as Data Frame and Series. For example: prophet provides a fast and automated way to forecast time series data based on an additive model; or scikit-learn, which provides machine learning algorithms and tools, such as linear regression, decision trees, and neural networks. However, Python has some challenges, such as steep learning curve, complex syntax, and dependency on external libraries and packages. Python is used by the financial services industry by 8.1%, after Information Technology and services and Computer software by 35.7% and 24.2%, respectively. 

• R: is a statistical programming language, and can perform various tasks, such as data analysis, data visualization, and modeling.R can be used to create forecasts using the “ts” object, which represents a time series data, and the forecast package, which provides methods and tools for forecasting, such as auto.arima, ets, and tbats. You can also use R to create more elaborate forecasts using the various libraries and packages available for forecasting, such as prophet, which provides a fast and automated way to forecast time series data based on an additive model; or keras, which provides a high-level interface for building and training neural networks. However, R has some difficulties, such as inconsistent syntax, memory management, and compatibility issues with other tools and platforms.

The specific AI tools employed by firms in the finance analytics sector in their investing strategies may vary depending on the team, project requirements, and technological preferences at any given time, also as sensitive data is securely stored and access is restricted to authorized personnel only. Unfortunately, we are unable to provide access to this data externally due to strict confidentiality measures and regulatory requirements.

Top-tier investment management firms using AI

In this section are shown and discussed the top-tier investment management firms that are already deploying AI for asset management. The following table represents the Best Financial Advisors 2024 based on time magazine.

JPMorgan Chase & Co. (JPM): this investment management giant is leveraging AI to build a software platform, similar to OpenAI's widely used large language model ChatGPT, to choose investments for client portfolios that are custom-designed to meet their unique needs. The company filed to trademark the term "IndexGPT" for the system in May 2023, with the AI-fueled stock-selection service trained on a mammoth 100 trillion words' worth of investment themes like stock prices, earnings reports, and analyst reports and ratings.

Morgan Stanley (MS): the Wall Street staple Morgan Stanley is linking up with OpenAI to enable the firm's financial advisors to immediately access the firm's research library to delve into information on client portfolio strategies and harvest relevant content in seconds. 

Vanguard Group: Vanguard embraced AI as not only a dynamic asset management powerhouse, but also as an all-around investment company management vehicle. Vanguard is already using AI tech to act as a robo-advisor that uses the company's exchange-traded funds, or ETFs, to generate personalized retirement portfolios for clients.

Deutsche Bank AG (DB): Deutsche Bank announced a "multiyear innovation partnership" with Nvidia Corp. (NVDA) to embed artificial intelligence into Deutsche's financial services and generate multiple AI-based applications, including intelligent avatars, speech AI and financial fraud defense. The AI partnership is expected to speed up analytical risk-and-return analysis and allow portfolio managers and traders to run investment selection scenarios at an accelerated pace.

ING Group NV (ING): In 2017, the company rolled out Katana, its bond-trading market analysis system, which can run through hundreds of thousands of real-time trading scenarios. The company reported Katana improved trade analysis times by 90% and curbed trade desk operating costs by 25%. More recently, ING reaffirmed its commitment to AI with its June 2023 hiring of Bahadir Yilmaz as chief analytics officer. At the announcement of his new position, Bahadir noted AI should revolutionize financial services in general, and at ING in particular.  

Fidelity: Boston-based financial services behemoth Fidelity is investing heavily in technology, hiring over 700 new technology specialists in the first six months of 2023, around half of those new hires headed to Fidelity's Research Triangle Park in Durham, NC. In January 2022, Fidelity announced Saifr, its AI and machine-learning system that automates its compliance management operations, giving the company a big advantage in a highly regulated industry. The Fidelity AMP automated digital advice platform, which first launched in 2017, is leaning on AI and machine learning to analyze data and make investment recommendations for Fidelity clients.

Wealthfront: the online-only automated investment manager's engagement with artificial intelligence dates back to 2016, when Wealthfront unveiled its AI-based application programming interface. The API merged personal financial interfaces for finding hidden fees and evaluating household budgets with portfolio management advice modeled by machine learning. The dashboard also featured a futuristic component (for 2016, at least) that forecasted the users' net worth over decades if they followed a programmed investment strategy over a long period. Wealthfront has expanded its AI-based asset management services, adding more features to its personalized investment portfolio services platform, such as automatic portfolio rebalancing, tax-loss harvesting and the rollout of Path, the investment firm's AI-based holistic financial planning advisory analytics tool. Additionally, the firm's Self-Driving Money strategy completely automates a user's savings and investment plan, relieving its customers from having to make direct money management decisions like moving cash around or monitoring financial accounts.

Autore: Parviz Izadi - Caporedattore: Tommaso Topa