Deception of banks

Occasionally people discuss banks’ controversial actions on the Internet. However, credit-card fraudsters, pickpocketing individuals and groups of defrauders specializing in “carousel”, “phools”, “con tricks” are practically always unsighted in the information space

How do those credit ninjas work and how are banks technologically advanced to cope with them? Has the time for Big Data come? Has the era of Big Brother finally arrived?

The serious problem in the current baking sector is consumer lending fraud.

The majority of losses is concentrated on credit fraud, whereas rates of losses significantly increase. Bankers assume a 10-percent non-repayment risk, thereby increasing their retail-loan rates.

Fraud is a wide-ranged term for theft and fraud committed to obtain goods without paying. The simplest scheme is “carousel fraud”. Fraudsters choose unsuspecting people, train them to fill forms and answer tricky questions. Victims are sent under control of supervisors to stores and branch banks. The lucky ones whose credit card limit was approved become participants of the blue light special, rushing to get a bunch of loans in different retail stores until banks will not start getting information from credit bureaus.

Those people are usually called “phools” for being na?ve and those who organize all are named “phool phishers”. The “phools” risk their credit history pursuing 10-20% of “plunder” and then run into continual troubles with debt collectors. The “phool phishers” spend a lot of efforts creating a factory with assembly line recruiting next victims and coping employees out. The scheme is pricey, labor-consuming, short-lived, but works quite well with newly established banks that have not accumulated loss statistics and skimped on scoring systems.

Panhandlers on the streets are often trained by professional acrobats and psychologists, fraudsters frequently cooperate with former law-enforcement and bank’s officers. In order to increase the possibility of approving credit card limits, legends of application forms are verified by shills: assiduous participants of a group, managing a bunch of mobile phones and answering tricky questions of underwriters and antifraud using dodgy scripts. Professional slang is one of features that is perfectly mastered by such people.

Not waiting for direct questions about the idea of work and ways of picking goods they in advance implement patterns into a speech: jokes and bywords about professions and object of purchase. Pretend to be crane operator saying “under construction”, an electrician – “ground it”, a rigger – “cargo frozen”. They chatter about types of matrix in laptops, spatial resolution of a camera, discuss basses of musical instruments, creating an impression of the reasonable choice.

The serious underestimated problem is to locate the big number of telephone sets in a single zone of triangulation according to base stations of mobile carriers. Same negligence let banks block credit card on suspicion of fraud, which are located far away from owners’ mobile phones. Here we stand on the land of creativity, admissibility and prosperity.

Fraud detection system apply script frameworks, successfully detect defrauders criterion-related to previously revealed and investigated cases. Until banks do not accumulate enough statistics and comprehend it they will incur losses, whereas fraudsters will gain profit. In this reality increase in loan loss provision is obvious, since from year to year defrauders’ attacks are becoming more and more sophisticated; when banks need more and more time and resources to elicit new fraud frameworks.

Is it possible to prevent fraud in advance; before banks discover losses and infer their fraud background? It is possible when it comes time for analysis of the borrower’s social graph.

The majority of organized bank fraud can be tracked on the web – the considerable share is organized on the Internet. You can recruit “phools” sticking posts on pillars at bus stops; however, for defrauders it is easier and safer to look for partners using social networks. Generality and anonymity are guaranteed – this is exactly what fraudsters need. How to understand that a potential borrower is not a defrauder? In some cases it is enough to use Big Data to research his or her entourage in social networks; if among first or second circle there are any suspicious people, for instance promoting help to get a loan, then the possibility of fraud default increases to 30% and more.

Another popular fraud framework is false documents: from passports to personal income tax forms. A group of people who are capable of manufacturing and distributing documents required to obtain a loan using fake identification documents. Wherein there are more than one person, social networks appear, an analysis of which allow detect suspicious patterns in potential defrauder’s surroundings: participating in suspicious groups, contacts with distributors of fake identity documents.

The analysis of applicant’s social graph can point out cases adjacent to credit card fraud; however, significantly increasing the probability of default.

Taking out a loan for someone else: after being refused the person starts asking friends or family members to get or cosign loan. Not everyone can be sensible evaluating the consequences of such “help” and they agree to support them. Banks are interested to know the final beneficiary of the granted loan and close connection between borrower and the one who recently got denied allows assess risks more precisely and accurately check credit applications.

The lack of knowledge to plan income and expenses, financial illiteracy, participating in pyramid schemes, attempts to make money on fluctuations of currency rates or stocks or gambling are not pure fraud, these are symptoms of the high possibility of borrower’s default. In order to determine such “interests”, you need to analyze the borrower’s activity in social networks or his/her web search queries and surroundings’ interests in those themes. If similar topics are popular among the friends, then it means solid interest in unreasonable waste of money and high probability of non-payments in the future.

The analysis of social connections and person’s online activity is not just an indicator of the high fraud risk; it is a specification – what type of fraud can we expect in every single case. It allows banks carry out more effective and precise risks’ assessment.

The analysis of social graph lets banks minimize losses from the most dangerous type of fraud – internal fraud, which is extremely complicated to identify. Losses are the highest among all. However, like in other cases withdrawal of money requires an accomplice. Then start looking for the path between the employee and the client. If the employee and the borrower have intermediaries between then the probability of loan loss provision increases. Besides, the employee can be related to the borrower not directly, there could be connected in one or several degrees. This means that the employee did not organize fraud, he was a participant involved in professional fraud scheme.

Frauds will continue to occur here, until severe punishment such as cropping or cutting hands in Middle Ages is imposed. Or until we are inflamed with equitable anger, because we overpay twice for loans due to being indifferent to frauds. You can bury your head in the sand thinking it does not concern you. On the other hand, you can be more responsible for your real and online connections. Do not agree to be a cosigner for the person you would not lend money to. Do not lie over the personal telephone that you are from human resources or accounting department. Do not accept someone’s friend request and do deny people offering to you suspicious business ideas.

Author: Oleg Braginsky
Translation: Marina Alexandrova
Source: Banki.ru