Swiss Re defines frontier markets in its publication: Sigma 02/16 Insuring the Frontier Markets (May 2016) and those in particular with promising long term growth potential. Within ASEAN, Cambodia, Laos, Myanmar and Vietnam are mentioned as being frontier markets with promising potential.

In terms of challenges in a frontier market, this can be split between technical challenges and more general issues.

Technical issues:

• Mortality tables – we need to make an assumption for mortality in the frontier market but unfortunately there is little to no data and certainly no published mortality table. Take Cambodia as an example. We know the life expectancy of Cambodia and we can compare it to neighboring countries with a similar culture, lifestyle or economic conditions (and has a mortality table!). The life expectancy of Cambodia now is perhaps that of Thailand around the year 1986 or Malaysia around 1980. Looking at the culture of Cambodia perhaps Thailand is the best proxy so we can look at an insured mortality table for Thailand around 1986. Obviously, this is not an exact science and not ideal, but this is the best we can do.
• Reinsurance rates – we need reinsurance coverage for any insurance company, but in a frontier market this can be a challenge. In the Maldives for instance, when we were developing a new product for an insurer there, it was extremely difficult to get even one reinsurer to provide a quotation. This is completely the opposite from more established markets where reinsurers fight to gain market share.
• Reserving – would need to follow regulations, but regulations can be very simplistic and even vague at times. At a minimum we would want reserves at a best estimate plus some padding, which for life insurance means we need a discount rate. Unlike developing markets there wouldn’t be a yield curve and even if we wanted to use the actual investment return experience of the company in many cases there are very few relevant investments for the company such as bonds.
• Investment performance – beyond the effect on the discount rate, in more developed markets investments are used to match with liabilities to back the products being sold. In frontier markets this is simply impossible, which forces us to decide between selling products that are needed in the market but for which we cannot find matching assets or sell products where we can find matching assets even though this limits their popularity.
• Other assumptions such as lapsation – in a developing market we struggle to get full credibility for our assumptions and tend to use credibility theory to balance ‘industry expected’ assumptions with the more volatile company experience. In a frontier market there is simply too much volatility, making us go to more macroeconomic considerations to determine a reasonable assumption.
• Data – for pricing of products such as motor insurance we need data, which even if such data exists the quality can be quite questionable, making product pricing challenging.

More general issues:

• A less developed court system with respect to financial cases creates more volatility with respect to third party liability business. The payout for this business is a combination of the present value of the loss of income due to the accident plus medical costs plus punitive damages. In more established markets such as the UK there are tables for such benefits given to the judges but in frontier markets this can be much more volatile.
• A severe lack of insurance professionals exists in frontier markets, such as actuaries, underwriters, claims adjusters and intermediaries. When a new insurer enters the market there will be the usual pinching of staff and thus inflation in salaries, causing further challenges to the insurers.
• There tends to be a very low level of awareness of insurance in frontier markets. In one market we noticed virtually no sales of mortgage reducing term products (death and TPD coverage during the period of the underlying bank loan, to protect the bank if the borrower passes away). Upon closer inspection yearly personal accident coverage has been used to cover the underlying loan as it is cheaper (even though natural death is not covered and it is just a yearly coverage).
• Coupled with the lack of awareness of insurance there tends to be a lack of consumer protections to ensure the insured understands what they are buying.

Applying actuarial science in a frontier market requires us to get down to the basics of what we do in order to determine both assumptions and methodologies, yet at the same time requires us to be creative in finding solutions to the needs of insurers in this market. Difficult but extremely exciting! As an actuary, I believe that it is in these markets where we can really add the most value.