For each customer, a specific risk profile can be created. This is created based on variables such as spending patterns, purchase behavior, public data sets etc.  When more data is used, the risk profile generated is much better thereby reducing credit default risks. To gain a better understanding about how an insurance company leveraged big data to its advantage, be sure to check out Insurethebox’s story.

Detection of fraud is made easier with big data. With careful analysis, if it is noticed that a particular customer deviates from their usual pattern which they have maintained for many years, there may be chance for potential fraud. Discovering these anomalies can be done by using outlier detection techniques. Take the case where a credit card is used in quick succession in different geographies, algorithms can easily detect this and alert the companies. Visa is the best example of this. They have incorporated big data systems which analyze 500 different variables of a transaction at once. With a fraud opportunity of potentially 2 billion USD, Visa had all the right reasons to do so.

These are some cases where big data can have a huge positive influence on the financial services industry.

Haven’t we all come across the phrase “It’s too good to be true!”? That may be the case when it comes to Big Data in finance too. Let’s see how this is possible.

We know that the data of financial institutions is extremely sensitive and highly confidential. Companies don’t allow employees to even access their personal e-mail accounts on premises. With the influx of big data and cloud storage, financial institutions are afraid of a data breach. With the recent Facebook and Snapchat leaks, there is resistance to the idea of big data. Watch the show Mr.Robot on Netflix and you will know exactly why financial institutions want to stick to traditional methods or data storage and sorting.

The volume, velocity and variety of data in the financial services industry is the highest. When there is an influx of such huge volumes of data, the room for error is almost close to nil! With that in mind financial institutions are apprehensive about the accuracy of Big Data. If they want a solution that is scalable, they will have to shell out a lot currency. With the already increasing expenses and overheads, spending tons of money on big data is a rather bitter pill to swallow. However, in the recent past, organizations in the financial services space are becoming open to adopting big data.

Whenever a new approach is adopted, the approach or process needs to be trusted whole heartedly. A financial services firm which usually follows traditional methods of data collation and inference will take a lot of time to accept the insights provided by a data scientist. If a data scientist, after examining the data provides actionable insights to a manager, the manager may feel threatened that his already set practices are in jeopardy. This in turn will bring about resistance to change within the subordinates.

Let me quote an example about a tennis player. There was a player who had a huge serve, a huge forehand and a huge backhand. The player liked to hit big and never compromised on shot speed. He in fact used to hit harder in times of desperation.  There was a coach who used to watch all his matches. He offered to coach him and the player accepted. The coach tweaked his game a little and asked him to play slower to reduce errors. Initially it wasn’t effective as the player felt lost. The player lost hope. Eventually he found that the coach was indeed right and he could last in the longer points for longer and could win them. He just needed to trust his coach and accept change. The player is the organization, the coach is the data scientist! And if you thought why I brought up the tennis reference, that’s because the player was me.