Quantitative analysis is largely a matter of knowing what data to look for and then knowing what to do with it. Analysts must know a number of different statistical techniques to manipulate raw data and turn it into something useful for individuals and companies. That means not only must analysts know the basics of statistical analysis, but they must also employ the cutting edge technologies to model and calibrate their data.
Before an analysis can even begin, analysts must prepare a data plan that will guide them through the process of analyzing a giving situation. Analysts must understand the context of the market or company they are evaluating, as well as the nuances of the given subject of analysis, both of which require research and creative thinking. If an analyst fails to understand the market environment or misses a crucial piece of data, they may produce flawed recommendations and cost a company or individual a great deal of money and even lose their own job. As a result, having a plan requires a great deal of attention to detail and assessing all possibilities in a given situation.
The second step to analysis is finding the data. Analysts can collect this data through personal observation or by amassing reports compiled by others, either the owner of a set of financial assets or disinterested third parties. If an analyst is preparing an investment plan, for instance, he or she must look at stock reports, risk assessments of available company stocks, costs of derivatives, individual portfolio data, and more. The data for each step of a single recommendation can be vast and overwhelming, and it is the analyst's job to process, quantify, and prioritize all of the available information before it can be useful.
Once all of the data is secured, organized, and quantified, it is time for the actual analysis to begin. The individual steps of quantitative analysis depend upon the data plan. Sometimes the information sought can be found with an easy analysis of descriptive statistics looking at means, medians, standard deviations and the like. Other times the analyst seeks more complex information such as correlations, probabilities, and skewness, looking respectively for associations between different data, frequency and likelihood of specific events, and outliers to larger bits of data.
For even more complex data, an analyst must deploy statistical and mathematical models to make sense of the information they have collected. Statistical models are formal ways of describing the relationships between data variables. Analysts apply these models to their data in an attempt to understand how one bit of data relates to the next. If the data fits a certain model, the analyst can draw certain conclusions about that data. Using modeling, analysts can also simulate what will happen if they recommend one course of action over another.
How the analyst interprets the results of their analysis and modeling determines what recommendations they make. If for instance the analyst found that based on their data a certain type of stock that had been declining was likely to increase over the next seven months, due to a cyclical nature of those stocks in the market, the analyst would recommend adding stocks of that type to a portfolio. Others may have been reluctant to invest in that same stock because of its current downward trajectory, but because the stock fit a model the analyst applied to it, the investor may enjoy the benefits of the analyst's interpretation of the data and subsequent prediction.
The successful analyst can employ many different techniques for collecting, analyzing, and interpreting financial data. Some rely more heavily on mathematical models and stochastic calculus to determine the exact right price value of an asset. Others prefer statistical modeling and educated guesses on the future of particular markets. Either way, quantitative data analysis requires an exact and efficient mind to turn raw data into successful financial action.