Big Data Science in Finance

Conclusion

Big Data in Finance has enormous potential. The applications covered in this book alone deliver high-performance results, when executed with precision and care. The number of open problems, however, is vast, and many of these are listed in this book.

As this book illustrates, the applications of Big Data extend into all areas of Finance from trading to credit risk to back office management. Big Data technologies help companies break down traditional barriers between departments and organizations, by allowing them to agglomerate the data from various sources often without data standardization that traditionally was one of the biggest roadblocks to successful data sharing inside large organizations. As described in Chapter 7 of this book, for example, even missing data fields are no longer a barrier to extracting precise and meaningful inferences from all the available data. In fact, as this book illustrates, more data, not cleaner data, lead to higher-quality inferences. Higher amounts of data allow for the population properties to emerge. This contrasts with traditional econometrics, which relies on extracting data properties from pristine yet limited samples that often are not even representative of the entire population.

This book also illustrates techniques and results completely novel to Finance and in many cases original altogether. For example, the study of how noise and missing data impact the error in eigenvalue estimation, again in Chapter 7, is one of the ...