Actuarial Control Cycle: Defining the problem (part 1 of 2)

Actuarial solutions tend to have a common theme - there is a series of cashflows and some uncertainty on magnitude, timing or even binary occurrence around these cashflows. The objective is usually funding, profitability or capital allocation. However once in a while a project may have different aims altogether. In these two articles, I discuss how the actuarial control cycle was applied to create a new scheme to benefit many that could not get on the property ladder in Malta.

The actuarial control cycle is a conceptual framework that can be used to essentially tackle problems. This is usually introduced in the latter part of the associate stage such as the Fundamentals of Actuarial Practice (FAP) for the Society of Actuaries (USA) or Actuarial Practice (CP1) for the Institute and Faculty of Actuaries (IFoA, UK). This framework has 5 essential elements which are discussed in application to this particular solution below.

The Actuarial Control Cycle: Defining the Problem and External Forces

The problem was that a number of individuals in Malta with good credit and income were unable to purchase property through a mortgage as they were not quoted life insurance. That seems to be pretty straightforward: some individuals are too risky to insure or their relative risk would be too high... I wish it was as simple as it looks! Some individuals did not seem to be at much higher risk categories - for example cancer patients who have been in remission for a number of years. Despite this, insurers did not feel comfortable in pricing a premium - and when insurers cannot price a premium (due to lack of data), they reasonably do not quote or else provide a high estimate. This is because an insurer's premium is based on a few key parts: the expected claim cost, profit, expenses, contingency and taxation. In one-off cases, the variability in potential claim (due to lack of data) cost leads to a much higher level of contingency.

The team at The Ministry of Social Accommodation, especially Stefan Cutajar, Matthew Zerafa and Chris Spiteri, had been researching the problem for a number of months. A working group was set-up (Wohoo I could put the chief hat on for this one) to discuss, or better yet define this problem. The group included many industry experts, including a strong participation from the both the insurance and banking industry.

The working group mostly focused on defining the problem and understanding the key processes of a mortgage was essential. In the interim parallel discussions were being held with many other stakeholders, including individual banks (including the Malta Development Bank) and NGOs. We had to redefine the problem countless times so for the sake of conciseness, I provide the below.

Banks in Malta are not legally required to require a life insurance cover as security for a mortgage. Yet this practice is applied by every bank. The rationale is simple: Maltese civil law makes repossession of property quite difficult and cumbersome, even after death. The life insurance cover provides a simple clean solution for the bank and safety for its depositors.

Some of the smaller banks in Malta are proud (and rightly so!) of their small number of repossessions required over decades. Sadly, this meant that individuals with good credit rating may still be unable to get a mortgage. This limitation was less apparent in other jurisdictions - I asked two ex-colleagues in the UK that would have not been able to obtain a life insurance cover in Malta. In the UK, they both had a mortgage - the only difference is that their banks can more easily repossess their property than here.

Despite my legal ignorance, I could say that the safety net against homelessness provided by law (making repossession difficult) was actually a stumbling block for this cohort of individuals (who are unable to find life insurance). Obviously no government would be willing to take out these guarantees that essentially limit the chance of homelessness.

The sceptic in me would also like to believe that banks were adding insurance costs as most banks in Malta are affiliated with a life insurer through shared ownership. However if this was the case, why would banks with no life insurer within their group ownership also request a life insurance cover?

In parallel with this project, I was communicating a lot with Iain Allan who is leading the path in the banking sector for Actuaries within the IFoA. The aim of those discussions were on introducing banking related examples for the Actuarial Mathematics (CM1) and Financial Engineering and Loss Reserving (CM2). Iain indirectly helped me understand the stresses face by banks when considering losses on loans.

In summary, essentially a legal risk (by banks not being able to repossess property post-death) was being transferred to the insurers as a mortality risk.

While I notice that there would still be some other risks that are maintained by the bank (such an unemployment risk), one couldn't simply create a banking / insurance solution. In essence we could not create an insurer just to cover this cohort.

We needed to further define the problem - who should we focus on? Since our scope was limited to home ownership, we came up with a definition:

• individuals who were not quoted life insurance (or it was prohibitively expensive),
• maximum cover of €250k - this being enough to cover 75% of property purchases in Malta, and
• individuals who are very likely to reach retirement age (age 65).

Moreover there was one essential element of interest: this scheme helps individuals reach their aspirations of home ownership - it is not intended to be a charity/give-away and that any solution needed to be sustainable. The last requirement (ability to reach age of 65) was more set-up to guarantee the sustainability of any solution. Any solution is bound to need revision and we can see how the latter point could be relaxed in future iterations.

The problem was not just defined at the initial stage but redefined continuously. For example initially the maximum cover was lower. Yet all the banks involved hinted that this may need to be higher.

Now that the problem is somewhat defined, in the next article - I discuss how a solution was developed and what further improvements may be required...The second article can be found here: Actuarial Control Cycle: Creating a non-actuarial solution (part 2 of 2)