“107,000 fraudulent insurance claims worth £12 billion were uncovered by insurers in 2019” (UK)
“The total cost of insurance fraud is estimated to be more than $40 billion per year “(USA)
“Insurance frauds see an increase during the pandemic in 2020, says a survey. The survey goes on to enlighten, frauds cost a whopping Rs 45,000 crore every year to insurance companies” (India)
Knowing that insurance fraud does happen is one thing, but knowing the extent of the fraud is another thing. Insurance fraud is a global malaise. Insurers across the world are struggling with fraud. One would think that countries with stricter laws would have fewer incidences of insurance fraud. Unfortunately, as the above publications would indicate, this is not the case, as malicious dealings in insurance transactions have become endemic to most nations. The problem is worsened by the fact that fraud happens through a nexus, the producer, the user and the providers are in collusion while committing insurance fraud.
So how exactly does insurance fraud happen? Looking closely at fraud cases in India, the Middle East and the USA, it is apparent that the modus operandi of fraudsters in various countries is shockingly similar. The following are the most commonly found types of insurance fraud
Fraud By Users At Underwriting Stage:
- Hiding pre-existing medical conditions
- Resorting to impersonation while undergoing medicals.
- Taking insurance policies for the non-existing person
- Providing fabricated income evidence for insurance applications
Fraud By Users At Claims Stage:
- Furnishing false treatment records for claim settlement. This has reached gigantic proportions in the pandemic phase.
- Inflating bills
- Making false claims about accidents. This has also seen a rise during economic downturns in 2009, demonetisation and most recently during the pandemic
- Treatment was sought for the non-insured person on the insured person’s health care.
Fraud By Providers:
- Unnecessary tests and treatments are done for patients
- Over hospitalization, investigating and over treatment for minor illnesses
- Falsifying treatment papers and medical records
- Falsifying certifications and specializations of medical providers by using unrecognised qualifications
- Performing unnecessary surgeries
Unbundling surgeries or tests to bill for a larger amount separately.
Indian insurance companies have borne a loss of over Rs 30,000 crore in 2011 due to different kinds of frauds, the figure went up to Rs 45,000 crores in 2020. So, it’s adequately established that fraud is a critical issue and insurance companies lose a significant part of their revenue due to fraud. Now the bigger question is why should an average citizen be concerned about insurance fraud? According to the FBI (in the United States, the FBI is involved in curtailing insurance fraud), non-health insurance fraud costs an estimated $40 billion per year, which increases the premiums for the average U.S. family by between $400 and $700 annually. The rising losses due to fraud raise premiums for all further decelerating the already low penetration of insurance. It also reduces the trust of policy owners and insurers.
For decades, insurance company risk managers have tried sophisticated mechanisms to counter fraud, but have not been able to plug in the leak completely. Analysing trends of fraud, various modes of fraud, and collaborating with eco-system partners to mitigate fraud have been tried. However, fraudsters have always been one step ahead of risk managers. Put in other words, the changing nature of fraud makes fraud detection and prevention extremely challenging.
The only way to counter fraud is through Data. The insurance application process generates data at every step from lead identification to underwriting and policy issuance. With these data points, a real-time fraud framework that can predict instances of possible fraud can be built. These data points could come from structured and unstructured sources and any section of the journey including interactions with external data points, transactions, relationships, and demographics. The platform should have the capability to analyse and identify fraud before it happens and let the business user develop models in real-time to counter newer types of threats that emerge.
Let’s take an example. Data points like age, gender, occupation, income, mode of premium payment, frequency of premium payment, nominee relationship, geographic location, and channel of sourcing business are available in each & every application. Using these data points, multiple data sets can be created and by using the machine learning approach models can be developed. The more the data sets, the more the model will be near perfection. If one can add external data points, the Credit Score model will become even more mature. The expected result is clear risk categorisation like High Risk, Medium Risk, Low Risk etc. It is important to create this risk categorisation right at the beginning so that each application travels a different journey in terms of due diligence.
Besides risk categorisation, predictive modelling has also been viewed as a promising approach to fraud management. As more data becomes available, one can apply learnings from past fraudulent behaviour to predict its likelihood in the future and develop models to assess fraud risk. These models also use several variables, each weighted depending upon their impact on fraudulent behaviour risk. As individual cases come in, they are assigned a score based on historical industry claim experience. The final risk score is then calculated by applying the weighted values.
A newer concept of fraud risk management is Hindsight Underwriting, which essentially means monitory suspicious insurance policies even after issuance. Since in India, claims cannot be repudiated basis of misrepresentation or suppression of a material fact post completion of 3 policy anniversaries (Section 45), insurers track high-risk policies for at least 2 years post-issuance. A leading life insurance company in India has rendered policies with a cumulative sum assured of Rs 2,358 crores null and void post-issuance in 2021.
The problem of insurance fraud may be both gigantic and universal, but it’s not insurmountable. Clear anti-fraud organizational philosophy, stricter and clearer laws to deal with the fraudster, along with data and knowledge-backed algorithms, can create a new path for controlling losses from fraud while creating confidence in the mind of policy owners.
Blog Author – Rohit Boda, Managing Director, J.B.Boda Group