Free Chapter
The financial industry has managed to find varied uses for time series. Some of the uses we are going to cover are as follows:
Stock price prediction is based on the assumption that the efficient market hypothesis doesn't hold true. The efficient market hypothesis states that, at any point, the price of stock is already reflective of all information and rational expectations. Hence, no amount of insight generated from historical price trends or influencer variables will predict the stock price movement effectively. There is academic evidence to support the efficient market hypothesis, but also an acknowledgement that there are some individuals and institutions who have managed to beat the average returns of the stock market by using their judgment. However, it is worth noting that stock prices are highly reactive to news and events, and also seem to be driven by both rational and irrational expectations. Time series data does have a role to play in predicting stock prices but its application is changing. In predicting stock prices, the problem isn't the availability of data but rather about cancelling out the noise and finding the real reasons that a stock moves.
Time series is also helpful when making portfolio decisions. Unlike stock prices, which can change by the second, portfolio decisions are taken over a slightly longer time-frame, ranging from a day to years. Time series data can help us understand the elements of an investment portfolio, the expected returns in a number of years, and even the probable behavior of investors once they re-invest their money from a maturity fixed bond into available products in the market.
Basel norms were introduced by the Basel Committee on Bank Supervision (BCBS), which set out the minimum capital requirements that financial institutions need to hold to minimize credit risk. What started out as a voluntary framework that institutions were free to adopt is now a key requirement for some central banks. The Basel norms have been revised to safeguard against the growing risks that financial institutions face. Time series data is used to build various models related to the probability of default and various other metrics that help in assessing credit risk and deciding the level of capital that institutions need to hold to offset risk. The Prudential Regulatory Authority (PRA) in the UK and similar federal organizations regulate and monitor adherence to the Basel norms.
Any organization, industry sector, or government body needs to estimate demand for products or services. The estimation needs to be primarily done at the firm and industry level, as there might be many more models required within one firm to estimate demand. A finance team may use demand estimation models to assess its funding needs; an inventory management team may assess consumer demand and its current production levels to assess stock needs and plan production accordingly. At times, forecasting might happen at a macro level, for example involving the economy, market sizing, and so on. A company planning to diversify into a new sector will want to know what the current market demand is, how much is it expected to grow by, and what proportion of this market can be captured as a new entrant in what might be a crowded marketplace of established players. Demand planning helps with all of these scenarios.
Inflation is a measure that affects all aspects of our life, including earnings and spending power. It is produced by central banks, or some other nominated government institution, and is used as a benchmark to assess the health of the economy and set expectations on the level of future earnings to ensure sufficient returns on investment. Various levels of inflation can highlight different problems in the economy. Japan can be considered as a classic example of experiencing deflation, where a government tries to increase spending through various measures to get the inflation rate higher. Run-away inflation in Zimbabwe and Venezuela has caused havoc for its residents, on the other hand, while the Eurozone is struggling to get inflation to a meaningful growth rate. Whatever the rate of inflation, forecasting it using time series data is of the utmost importance.
Managing customer journeys and maintaining loyalty aren't the most obvious uses of time series data, but by assessing the history of customers' past product choices, transactional data, and their engagement with an organization, you can try to manage this customer journey. After all, in most instances, it is much cheaper to keep a customer loyal and onboard than to acquire a new customer.
