Chartering an actuary ... or maybe not

Last month I overplayed my baking abilities by stating that I have sat for a pastry chef course (with exams and all). People in my office were imagining, and some demanding, eclairs on a weekly basis - having a chef in the building is the ultimate office perk!

Then again I never brought anything I baked to work, the course was an introductory one and you shouldn't really trust me with managing a large bakery (more about the similarities of baking to actuarial science here).

Size, scale and complexity

Yet it is all about size, scale and complexity right? I cannot produce a perfect Gateau St Honoré from scratch but I can dish out a mean sticky toffee pudding. I can't organise baking for a catering establishment having 100 people at any go, but I can probably feed the office.

But am I a chef? By any standard, no.

And this leads me to the next question - what is an actuary?

Being an actuary, and the hurdles to reach that, may be clear in the UK and US (as well as other jurisdictions that follow these exams) but it is not the same hurdle throughout. The International Actuarial Association sets the minimum syllabus required for actuarial studies worldwide - these tend to be equivalent to an associate actuary for UK/USA equivalents.

Right now, the Institute and Faculty of Actuaries (IFoA) is holding an election to allow a change in nomenclature - that we officially call ourselves Chartered Actuary. The IFoA is dedicating most of its blog to it.

Let me put it out there - I agree with the change and am publishing this article today for two reasons:

• voting finishes today so publishing it now evades the unlikely scenario of influencing anyone's vote, and
• this article is essentially me writing my developed thoughts on this while on holiday (returned today). I carried the Actuary magazine for reading while on flight and saw the hard-sell by the IFoA on the last two editions of the magazine. This article has a lot of 'I ' and 'me' - the aim of this article is not to influence but ramble.

Will the real actuary please stand up?

Despite coding algorithms is fun (yes I mean that), I stop many times during the day to offer a tea round around our Argus Malta offices. Once a month, usually from a new joiner, I get the question "What do you do?".

Thing is, actuaries are not well known - the majority of claims handlers never met an actuary nor understand what we do. There are people who believe me when I reply with a straight face "ahh I do fancy graphs".

It is clear in Bermuda what an actuary does - the island has the highest actuary-per-capita rate in the world (and you wonder why it never became a party island) but it is not clear in many other jurisdictions. I have been introduced to actuaries with no actuarial qualification - maybe an exam or two. It would also be na?ve to not mention that there is an element of examination arbitrage going on.

I believe the chartered title to be explicitly mentioned provides a clear international benchmark. There are counter-arguments that the fellowship title would lose its appeal. That is a potential risk but it should still remain the gold status in Australia, Bermuda, India, UK and other locations.

It would also help in actuaries working in other fields - many of us are way too happy to be confined in insurance pricing, reserving or capital modelling. We are well paid for it and the risk of using newer techniques, especially for reserving or capital are too great.

What will the future hold?

?I fear the chartered actuary title will not pass. It has caused a lot of hot debate, especially in UK. [EDIT 17 December: About two thirds voted in favour)

There are bigger problems facing the actuarial profession than the title. After all those years of training - we follow blindly methodologies that may be better suited for use the until the 1990s. For example the triangulation methods (Bornhuetter-Ferguson and so on) are beautifully created for the data limitations of their time. However they are inept and mathematically not up to scratch to today's data. Actuarial reserving work in Property and Casualty (or non-life for the UK based folk) has been based on running the same model but adding boot-strapping or other techniques. This is like taking a bow and arrow and adding a laser pointer - why not just use a crossbow?

Yet auditors and regulators want to see these methods - larger and smaller consultants also want to use just these methods. It's a circular feedback loop.

Frank Redington's statement that "An actuary who is just an actuary isn't an actuary." couldn't hold truer these days.

Solvency II was a great reason to rethink P&C reserving. Yet most companies, if not all, adopted an IFRS4 approach to setting reserves and this is translated to SII. Similarly IFRS 17 could provide a great excuse to rethink reserving in a back-to-basics methods, and by that I am not referring to anything as fancy as Machine Learning (although useful).

Mathematically, I believe this is the incorrect order to do it.

But I bet that IFRS4 thinking is still there - at a huge cost to develop software and techniques to transform according to IFRS 17 needs. Triangulation methods were not developed to derive these disclosures. The P&C techniques use judgement to set up the main parameters - be it the loss ratio, average cost per claim, or other item. This is likely to be on the prudent end due to the sensitivity of the results to the parameter. Yet both IFRS17 and SII should be on a best-estimate method. Life actuaries are at a bit of an advantage on this as they have been taking the concept of cashflows since inception of some techniques.

My fear is that the actuary of the future will not be providing useful input but remain the gatekeeper of reserves essentially using a method developed in 1972. I have a hunch that Bornhuetter and Ferguson would be honoured and shocked in equal measures that most firms still use their methods fifty years after inception.

And on that note, this ramble is over - the World Cup semi-final started - Messi just scored a penalty and I promised myself to watch it. If there are four things to take from this incoherent article, these are below:

1. As actuaries, we need to evade group think. If a method could be run on the computer shown on the screen - then maybe there is a better method now.
2. The need for a clearer actuarial classification that would help the profession to provide actuarial and math expertise in a wider context is clear - the change of title is a potential approach (however the IFoA's communication team may have faltered to understand their older members' fears. The risk of diminishing the value of the current certification is real and the counter-arguments to the new title were not presented by the IFoA),
3. Techniques we currently use were developed with a different problem to solve and in a different environment (some examples of 1970s vs 2020s: a management margin not a risk margin, a prudent estimate not a best estimate, data received annually, no computing power). If the model can be run on the PC shown in the picture above - then it's probably obsolete.
4. More importantly, I am not a chef.