OVERVIEW

In an earlier post, "What Type of Artificial Intelligence Do You Need?", two paradigms of artificial intelligence are described; deterministic and statistical.  Deterministic is similar to speaking with an expert.  In this blog post, we will consider how Deterministic Artificial Intelligence can be applied to Client Suitability, which is a well-known regulatory challenge in financial services.

Deterministic A.I. can help financial institutions process high volumes of Client Suitability analysis.  By automating large amounts of decision-making and data analysis, financial institutions can limit their regulatory risk and increase efficiency.  In addition, Deterministic A.I. can increase uniformity in decision-making by limiting personal biases and errors.

BACKGROUND:  WHAT IS CLIENT SUITABILITY?

Client Suitability is an analytical process used to determine if a financial product is suitable for a client.  It applies to both the retail market (i.e.: an individual who wants to buy an insurance policy) and to the institutional market (i.e.: a hedge fund that wants to trade in a sophisticated derivative product).

The rules that determine the Client Suitability process depend on the type of client, the type of product, the regulatory body that oversees a particular market, and the internal controls of the financial institution.

WHY DO WE NEED CLIENT SUITABILITY?: PROVE FAIRNESS AND LIMIT RISK

Client Suitability is needed to prove fairness.  Under federal laws and regulations, a financial services institution must show that it deals fairly with its clients by fully understanding each client's investment profile before selling a product to that client.  The process is also a tool to limit risk for the institution.  In order to protect itself from potential litigation and financial loss, a financial institution must show that the client is sophisticated enough to understand the product, tolerate risk, and absorb any potential loss.  To prove fairness, financial services auditors and regulators often use the word "reasonable", such as "reasonable basis to believe” and a"reasonable belief" is based on "reasonable diligence".  Reasonable Diligence is usually based on quantitative and structured data; however, the actual decision can be subject to the disposition of the financial advisor or the trader that day.  Decision-making can also be affected by personal biases or errors -- we are human, after all.  In other words, the quality of traditional Client Suitability analysis can vary -- a lot.  By using Deterministic A.I. for Client Suitability Analysis, that unstable quality can be eliminated.

DETERMINISTIC A.I. AND CLIENT SUITABILITY IN ACTION: ENGAGING AND SCALING UP EXPERTS

To give a sense of what goes into a Client Suitability analysis, here is a hypothetical enterprise wide client suitability system.  The system would be wholly software-based and no separate business consulting tasks are needed or desired.

This platform would access and summarize quantitative data, such as a client's trading/banking history, credit worthiness, and answers to questionnaires.  The platform would implement a collaborative workflow that presents that data to a variety of experts.  The experts can then make a collective decision.  For example, the relationship manager is the expert on the client, the sales-trader is the expert on the product, and the compliance manager is an expert on the regulation that needs to be followed.  Together, the relationship manager, the sales-trader, and the compliance manager use their collective expertise to jointly decide if the financial product at issue is suitable for a client.

Client Suitability analysis can be difficult when a financial institution has large numbers of sales teams, traders, financial advisors, and relationship managers that support a variety of products for many types of clients.  Individual expertise needs to be scaled up to handle volume efficiently.  This ‘scaling up’ and volume problem is where Deterministic A.I. can help.

DETERMINISTIC ARTIFICIAL INTELLIGENCE IN USE

Deterministic A.I. allows an organization to scale up its best experts by recording and/or codifying their decision making processes.  Deterministic A.I. aggregates only the data that is needed and applies the best decision making process for the relevant scenario through machine-to-machine and machine-to-person dialogue.

Here is an example of the information needed to sufficiently evaluate client suitability:

• Age (in the case of individuals) or years in business (in the case of institutions);
• Liquidity needs;
• Risk tolerance;
• Other investments that the client has;
• Investment objectives, experience, and time horizon;
• Financial situation and needs;
• Tax status;
• Market regulations.

Some of this is quantitative but some is subjective, requiring a dialogue with the client through questionnaires and internal experts that know the client, the product, and the regulations.

The result is an auditable, optimized, and adaptable decision making process.  More importantly, by using a Deterministic A.I. platform, the organization now has a productive group of people that can focus on other, more important tasks, such as improved financial strategies and higher quality relationships over a larger volume of clients and products.

Link here and here for this and other related posts.