The second day at the AICPA Conference for Credit Unions started with a robust economic update and moved on to credit union specific lessons with topics surrounding artificial intelligence and fraud prevention strategies. Below are some of the particularly interesting topics discussed by the panels of speakers today.
Not surprisingly, The Economic Outlook seminar started by covering the 2018 rate hikes. The Federal Reserve voted to raise rates three times this year the eighth such hike since the Fed first began raising rates in December 2015. Still, these rates are fairly low relative to historical rates although the Fed believes they are quickly approaching a neutral rate level.
ContentMiddleAd
Unemployment has hit its lowest rate since December 1969. At roughly 3.7%, unemployment has been at or below what the Fed considers full employment since March 2017. Still, wage growth has not been indicative of such a low unemployment level. This is largely due to the declining participation rate in the U.S. workforce. With people leaving the workforce all together, the unemployment rate and the supply of talent have decreased significantly.
The forum changed to introduce what seemed to be a theme of the day: fraud. The seminar Artificial Intelligence focused largely on fraud prevention. The address started with some eye-opening statistics:
- 40% of fraud is caught by mistakes and tips.
- 3% of fraud is caught through analytics programs.
- On average, it takes 16 months to detect fraud from happening.
The good news? Artificial intelligence is able to create and implement tools that are more effective in detecting fraud.
Artificial intelligence is being applied across company roles. This democratization of analytics puts analytical tools into the hands of everyone at a credit union, not just those who have data analytical training. But this creates conflict between finding artificial intelligence tools or paying data analytical professionals. The argument made in the seminar is that, today, software that can accurately provide data outputs can be deployed at a fraction of the cost of comparable personnel.
First introduced in Artificial Intelligence, a subsequent seminar, Anti-fraud Strategies for Credit Unions, isolated how fraud is committed at credit unions. The number one strategy used for fraud control at credit unions are internal controls. These are largely crafted by upper management and board members at credit unions and look to achieve profitability and performance targets while ensuring compliance and preventing the loss of resources.
The two types of internal controls discussed were preventative and detective. Having a robust system in place for both categories is important for an anti-fraud program. These are broken down largely in three major ways: human error, collusion, and management override. All of which must to be limited for an internal control system to work.
The two dangers discussed in the seminar were financial misstatements and embezzlement. The internal controls, if properly implemented, should theoretically address these. This is harder said than done, particularly for credit unions with larger branch footprints. An institution is as susceptible to fraud as its most vulnerable branch. Credit unions must assess all potential danger points institution-wide and create a program that will accurately and quickly remedy a troubling situation.
