fbpixel
IFT Notes for Level I CFA® Program

R57 Fintech in Investment Management

Part 2


 

6.  Selected Applications of Fintech to Investment Management

So far, we have discussed Fintech in general, now we will look at selected applications of Fintech to investment management. There are four broad areas that we will consider:

6.1 Text Analytics and Natural Language Processing

Text analytics refers to the use of computer programs to derive meaning from large, unstructured text- or voice-based data. For example, text analytics can be used to gauge the consumer sentiment about a new product by analyzing what is being said about the product on blogs, forums, YouTube, etc. Based on this analysis, we can determine if the sentiment is very positive, positive, neutral, or negative.

Natural language processing is an application of text analytics whereby computers analyze and interpret human language. For example, NLP analysis can be used for communications from policy makers such as the US Federal Reserve. Officials at these institutions may send subtle messages through their choice of words and inferred tone. NLP analysis can provide insights into these subtle messages. Such processing is possible because of access to Big Data and processing power.

6.2 Robo-Advisory Services

This refers to providing investment solutions through online platforms. This replaces a human advisor with an online platform. Robo-advice typically starts with an investor questionnaire, which may include questions about income, spending, age, goals, investment horizon, etc. Based on the responses to these questions, the robo-adviser software uses algorithmic rules and historical market data to come up with recommendations.  The types of solutions offered through robo-advisory services include:

  • Automated asset allocation
  • Rebalancing
  • Tax strategies
  • Trade execution

Robo-advisers typically have low fees and low account minimums. This has increased the penetrating power of these services in reaching mass market segments, and people with relatively low wealth can now afford these services.

Robo-advisers cover both active and passive investment styles, but passive styles tend to be more common. They are usually more conservative in nature.

There are two major types of robo-advisory services

  • Fully automated digital wealth managers: As the term implies, there is absolutely no human involved in this model. These services offer low-cost investing solutions and usually recommend an investment portfolio composed of ETFs.
  • Adviser-assisted digital wealth managers: In addition to the online system, an investor also has access to a human advisor over the phone. The advisor can assist by giving a more customized advice based on the financial situation of the investor.

We need to recognize that robo-advice has its limits. There might be times, when an investor needs to speak to a person, especially in times of economic crises. Also, in instances where investors have specific needs or want to invest in alternative investments, robo-advice is not useful. However, despite these limitations robo-advisory services are becoming increasingly popular.

6.3 Risk Analysis

Stress testing and risk assessment involves a vast amount of risk data. This data can be in different forms – for example, structured or unstructured, quantitative or qualitative, etc. Also, there is an increased interest in monitoring risk in real time.

Instructor’s tip: These characteristics correspond to the three V’s of big data – volume, variety, and velocity. Hence this data can be considered Big Data.

Big Data and ML techniques can provide insight into changing market conditions. This can allow us to predict adverse market conditions and adverse tends.

Machine learning techniques can also be used to assess data quality. Faulty data, errors, outliers, etc., can be identified and removed from the analysis.

Big Data and ML techniques are also used in scenario analysis. Scenario analysis helps in evaluating the risk of a portfolio. For example, we can evaluate what would happen to our portfolio if the 2007 financial crisis scenario were to repeat itself. A common term used here is ‘what-if’ analysis. Here we evaluate what would happen to our portfolio under different market conditions.

These techniques have become increasingly popular because of our ability to deal with big data and the advanced analytical techniques that have been developed over the last few years.

6.4 Algorithmic Trading

Algorithmic trading refers to computerized trading based on pre-specified rules and guidelines. It can help us decide when, where, and how to trade. For example, after analyzing lots of past data an algorithmic program might tell you that trading during a certain time of day, on a particular exchange using limit orders is the most cost effective.

Algorithmic trading also allows us to take large orders and slice them into smaller pieces. These smaller pieces can be executed using the most appropriate exchanges and trading venues.

The benefits of algorithmic trading include:

  • Speed of execution: Since trading is done by computer programs based on predefined rules, the speed of execution is much faster.
  • Anonymity: Since large orders can be broken into smaller pieces and traded through different exchanges anonymity can be achieved, which may be important for some investors.
  • Lower transaction costs: As discussed above, by identifying the most cost-effective way to trade, algorithmic trading helps lower transaction costs. Also, because large orders are broken down into several smaller orders, the market impact (which is a significant component of the transaction costs) of the order is reduced.

High-frequency trading (HFT) is one form of algorithmic trading. Here orders or trades are automatically placed when certain conditions are met. Time is a crucial factor for such trades. Therefore, HFT takes place on ultra-high-speed, low-latency networks.


Portfolio Management Fintech in Investment Management Part 2