Sales Due Diligence, Big Data And Bad Apples

• When you check a stockbroker’s background, his co-workers’ records may tell as much about his potential for future misconduct as his own history.
• Does this have any relevance to business transactions… What’s behind the sales numbers of the company you just bought?
• Social network analytics identifies hidden networks and traces unknown links among individuals and entities in 3-D form.

FINRA’s BrokerCheck

The Financial Industry Regulatory Authority (FINRA) is the securities industry’s self-regulatory organization charged by the SEC with regulating stockbrokers (“registered representatives”) and the brokerage firms with which they’re affiliated.

FINRA maintains a public database, BrokerCheck [1], of investor complaints, and disciplinary and employment histories on 1.2 million current and former brokers.

From 2005 through 2015, 7 to 8% of brokers had “financial misconduct” records, including customer complaints, arbitration awards and settlements. Brokers with a rap sheet are five times as likely to engage in new misconduct as is a financial advisor with a clean record. [2]

But how is an investor to gauge the likelihood that a broker among the 92+ % with no record will engage in misconduct in the future?

According to three recent studies, the risk that a broker will cross the line for the first time is significantly increased if he works with co-workers who have misconduct records. [3] And in fact, one of the studies found that investors would be as well informed to know the average misconduct history of a broker’s co-workers as they would be knowing the broker’s own misconduct record. [4], [5], [6]

Tying sales misconduct to brokerage firm mergers, a fourth study done in 2015 for the New York Federal Reserve Bank, [7] found that after a merger, a financial advisor was 38% more likely to commit fraud if she were placed in a new office that included individuals with disciplinary records (relative to an advisor from the same firm who went to work in branch with no history of fraud). [8] It has also been suggested that misconduct risk in the banking industry may spread among firms via workers’ career and social networks. [9]

Is Misconduct Generally Contagious?

But does the bad apple syndrome extend beyond the financial services sector, which might concern, for example, a private middle-market manufacturer seeking to acquire a distributor?

Talent management vendor Cornerstone OnDemand’s recent study, drawing on a dataset of 63,000 employees (with customer contact) and 250,000 observations, found “extremely strong evidence” that toxic workplace behavior (violence, drug or alcohol abuse, sexual harassment, falsification of documents, fraud) is contagious, and that “employees are many times more likely to engage in toxic behavior if they’re exposed to other toxic employees”. [10]

Whether these and similar studies are seen as simply confirming the obvious or providing new insight, the challenge of identifying and dealing with problematic job candidates and employees, and their influence remains.

The due diligence focus here is on sales positions because of the:

• Leeway, slack or look-the-other way posture, depending on your point of view, often accorded top salespeople and sales teams who generate large revenue streams and commissions; and the
• Prevalence of unrealistic, short-term sales goals in lower level positions tied to bonuses and job retention

Predicting Misconduct from Behavior Patterns

When a business vets applicants for sales positions, new hires or monitors its sales operations there is no BrokerCheck analog, and workplace study findings are not tools.

However, advances in a field of fraud analytics known as social network analysis (SNA) hold promise.

SNA draws on a range of sources to identify and visualize patterns in data. It analyzes raw, unstructured information (employee financial transactions, communication records, credit scores, social media) about individuals and entities (“nodes”) and the relationships (“links”) among them. The results are presented as a visual network in the form of a multi-dimensional graph.

Depending on the particular SNA implementation, the results may be explored at multiple levels, filtered, additional data input and what-if scenarios created.

SNA business uses include identifying :

• links with known bad actors
• anomalous links between people, events, locations
• fraud networks, ring leaders, participants’ roles
• individual abnormal behavior, behavior changes compared to peer group

SNA may offer cost savings relative to expert rule-based systems (requiring specialized programming, continuous updating), statistical data mining techniques and machine learning approaches now widely used to combat fraud.

To date, SNA has gained most acceptance within the financial services, telecommunications and public organizations sectors. How quickly additional industry sectors adopt SNA as an additional anti-fraud tool will depend on how it’s significant data format, storage, processing time and automation challenges are met, and how data privacy requirements are addressed. [11]