In a year characterized by geopolitical risk, economic uncertainty and increased regulatory intensity, the findings of our 2017 Asia-Pacific Fraud Survey are a cause for concern: over half of close to 1,700 respondents across sectors believed that ethical standards have not improved in their local business operations.
In the insurance sector, fraud risk continues to be clear and present: reports have estimated that an average 5% of total insurance premiums globally is lost due to fraud. This implies that a whopping S$275 billion could have been lost to fraud in the S$5.5 trillion-strong insurance sector in 2016.
Insurance regulators worldwide are stepping up on vigilance and controls to prevent fraud in the sector. In Singapore, insurers need to follow the guidelines and framework set out by the Monetary Authority of Singapore, as released in the Reporting of Suspicious Activities and Incidents of Fraud 2013, and Guideline on Insurance Fraud Risk 2012.
Given heightened regulatory pressures, it is in the interest of insurers to nip the issue of fraud in the bud to protect their reputation and profitability.
Nature of insurance fraud
Generally, there are two types of insurance fraud. Hard insurance fraud refers to cases where someone or a group of people plans for or creates a scenario that eventually involves financial loss to insurance company.
These include policyholder falsifying death in life insurance, collusion between doctor and policyholder to falsify medical condition and bills in health insurance, staged car accidents in motor insurance, faked loss of personal belongings with genuine police report in travel insurance, and fictitious policy sold by insurance agents.
Soft insurance fraud – the more common of the two – refers to opportunistic fraud where the policyholder or agent, on behalf of policyholder, deliberately exaggerates or withholds information in order to secure a policy or a lower payable premium.
Some examples include misrepresentation of damage suffered in fire insurance and motor insurance claims, hiding health condition or overseas medical history in life and health insurance applications, and theft of premiums by insurance agents.
Clearly, insurance agents are a source of fraud perpetration, in the presence of greed and opportunity.
Agents are generally motivated by high commissions from new products and clients. Singapore insurance agents, on average, can earn commissions ranging from 50% on the first year’s premium to 5% after a few years for life and health insurance policies. For accident insurance products, agents can earn a commission of 20% to 30% in the first few years.
There are various innovative ways that fraud can be committed by an agent, which makes fraud detection challenging.
For one, agents may abuse the trust of policyholders and create fictitious policies where the premium is paid to the agent, who siphons it instead of passing it on to the insurer. The policyholder may be unaware of the fake policy as many would try to avoid filing a claim in the early years for fear of hike in premiums or leaving a weak insurance record.
Agents may also create churn as a tactic to earn more commission. They do so by influencing the policyholder to agree to terminate an existing policy for a “better policy” without educating the potential loss of switching to a new policy.
In other cases, agents may forge the policyholders’ signatures to cash out the investment or cash value generated along with the policy; or steal the premiums that are supposed to be credited into policyholder’s account.
Another fraudulent tactic is lapping, where the agent intentionally credits the payment by one policyholder to another policyholder’s account so as to buy time for the latter who could be under financial difficulty.
Closing gaps with forensic data analytics
As agents explore unethical ways to improve earnings, companies at the same time are not catching up on their fraud detection capabilities. Most companies today are using simple tools such as spreadsheets, macros and SQL and manually analyzing data sets including date of applications, claim timing and history, policyholder’s information and amendments, and policy termination timing and trends. This can be time-consuming and subject to human error.
Leading companies are deploying forensic data analytic (FDA) tools to improve the effectiveness of their anti-fraud programs. FDA tools can be tailored accordingly to the needs of each insurer based on the vast amount of structured and unstructured data available in their internal systems.
The fraud detection and prevention capability is further enhanced by adding customized predictive analytics parameters that include text mining and agent’s behavioral patterns relevant to their risk profiles. Such analyses offer a powerful approach to identify anomalies, taking into consideration the multidimensional attributes of data, enabling insurers to shift their focus to high-risk areas and agents.
With increased regulatory scrutiny comes the opportunity for insurers to consider FDA in their arsenal of anti-fraud tools. Based on our experience, a few large multinational insurers have adopted FDA tools and seeing the benefits. For example, our proprietary FDA tool “Know Your Agents” has helped a multinational insurer to identify over S$1.5 million of fraud in a single operating country. Smaller and local players continue to explore the capabilities of such technology.
Notwithstanding the above, the FDA tool is not a magic wand. It works well only if there is effective data management systems in place, adequate and trained resources to leverage the tool fully, as well as a consistent and clear anti-fraud framework that underpins visible action whenever fraud is identified.
In other words, a holistic anti-fraud program is needed and technology enablement will be a critical differentiator.
This article was first published in the September issue of Asia Insurance Review magazine.