Actuarial life in the 1980s

After detailing stories from his foundation years as an actuary in the 1970s, Dermot Balson reflects on some more idiosyncrasies of the profession in Australia throughout the 1980s and some words of wisdom.

I moved to Melbourne in 1980 to work for the local office of a British life insurer. It was set up 25 years earlier by three expatriates who were still working there.

It was a small office with a couple of hundred staff. We were ankle-biters compared with the heavy hitters AMP and National Mutual. The business was arranged in a  traditional way, having actuarial, investment, property and other departments. It had an executive dining room (where I was allowed, being an actuary) with a three-course lunch. There was a new CEO from England and he had simple tastes when it came to meals, often asking for a bowl of hot chips on the side. I found those meals tedious, because the executives were all much older and boring, so when I could, I’d eat in the staff canteen.

Status and parking

During those days, most people believed that the company, and their jobs with it, would continue indefinitely. This typically translated into workplaces full of people craving status where promotions come infrequently. I remember when someone in my area was promoted to Chief Clerk, which entitled him to a larger desk. His workspace was too small for it, but no, he had to have it.

He somehow squeezed the desk in, and as I recall, he had to climb over it to get behind it. But he had his desk to show his status.

The funniest thing in my time there was parking. The office was on a main road out of the city so people could park their cars along it. At around 10am every day, workers had to move their cars to another part of the road or the parking inspector would ticket them. So just before 10am, a swarm of people would troop down the stairs to their cars, sit in them until 10am, then drive sedately round the block in procession, and park on the other part of the road. Overtaking was frowned on.


A couple of times a year, a senior executive from the UK would favour us with a visit and the red carpet would be rolled out. What puzzled me was that the targets he set for us were based on revenue, not profit. Admittedly, profit was hard to calculate, but surely it was important.

I had a direct experience of how little it mattered when we were trying to lure a major broker to sell our products and he was offered a sweet deal with additional volume commission. I calculated that we would be losing money on every policy we sold, so I caught the chief actuary in the corridor and told him. He literally backpedalled down the corridor not wanting to hear it. All that mattered to the UK was revenue, so that was his focus too.


Dermot’s Epson HX-20 notebook computer.

Technology in the early 1980s was still primitive, with a mainframe, a clunky mini-computer in the actuarial department, and desk calculators.

In 1983, we got hold of some small programmable Epson HX-20 notebook computers, which had 16k memory, 4×20 character LCD screen, a little printer and a micro-cassette (the hard disk of its day). You programmed it in ‘basic’.

I loved it and I used it for many projects. Ultimately, we programmed our life products and prepared a number of machines for our salespeople. I remember going to one of the first sessions where we had four machines set up. We made the mistake of letting the salespeople fiddle with them and by the time the training session started, three of the machines were inoperative!

As an aside, in the corner of the office was an enormous cylindrical slide rule (known as the actuarial tool, of course) that stood a couple of feet high and had been used for calculations not too long before.

My work

I started in the actuarial department which was a little dull.

But this was also the time when computers began to be powerful enough to unbundle life insurance. This meant endowment policies, which combined life insurance and savings, could become a savings product with separate life insurance. The client could choose any mix they preferred. When I arrived, the company had just issued its first product of this type and everyone was very excited.

One thing that hadn’t changed, however, was the commission which was 100% of the first year’s premium and 10% thereafter.

And that was before office setup costs. It meant that roughly speaking, your whole first two years’ premiums went in upfront expenses.

Commission was never disclosed to clients and its size would have horrified them. This became a real problem a couple of years later, when the company wanted to introduce an investment-linked product, i.e. the client could choose to invest in different unitised asset classes like equity and property. But how were we going to tell clients that this wouldn’t start until the third year because the first two years had gone in expenses? It was a real dilemma.

However, a competitor with the same problem provided a solution. The first two years’ premiums were labelled foundation units and were not invested in any asset class. Instead they were held to one side and refunded in full (without interest) at the end of the term. This gave the insurer the benefit of all earnings on that money and relied on clients not realising that the value of that money would be pretty minimal after 20-30 years.

Our company followed suit, but that didn’t provide enough to cover the agent commission. The solution was to pad the insurance premium to meet the shortfall, because people had no real idea what insurance ought to cost. Of course, this led agents to really push insurance at clients to increase their commission.

They also permitted agents to sell these products as long service leave savings programs for 10-year terms – still with 100% commission, meaning they were atrocious investments.

I left the insurance industry soon after.

Sales and marketing

When I heard a sales and marketing department was being set up, I did everything I could to get a job there and eventually I was successful. The manager was a former branch manager whose sole purpose was to help agents sell more products. Thye used the most hackneyed slogans, like “people don’t plan to fail, they fail to plan”.

There was a test they gave aspiring new agents to see if they were suitable. Called the ‘Crows and Canaries Test’, it asked questions like “Would you like a big house?”. Apparently, greed was good. Very good. I tried the test and came out as a solid crow. The test was probably valid because salespeople were easily motivated by money and status. The branch managers used to have a dinner with the biggest seller each month.

This dinner was coveted as were the gold stars put on the wall for new sales.

The funniest was when they had a sales drive with a range of prizes for different sales levels. It would be sent to the agent’s home, so his wife would see it first and (hopefully) set her heart on something to motivate her husband.

Long service leave model

Most of my time as an actuary, I virtually lived in spreadsheets. But spreadsheets aren’t always necessary or even the best approach. Simple insights can sometimes cut through the complexities.

This rang true for me working with an actuary who was producing an asset liability model for a client’s long service leave program. He had 1000 simulated sets of investment and salary growth figures and he was planning to simulate the comings and goings of the 4,000 employees.

For the last part of this model, he intended to use a program built for superannuation valuations which was going to be very slow, taking hundreds of hours.

It turned out that because long service leave operates over short periods (in that case 10 years) all that was needed for the investment simulation of different asset mixes was the shape of the liability at each of the 10 years. And the law of large numbers meant that employee movements would largely average out and not be very volatile. This was confirmed by previous work showing the liability profile as very stable over several years.

We were able to use this profile, discard the need to model employee movements, and apply the simulated investment rates This now took a few seconds rather than hundreds of hours.

Executive super

Back in the 1980s, we could give advice on most anything. Back then I would help client executives plan their superannuation contributions tax effectively. I wrote a program to measure the effectiveness of each year’s contribution through to retirement and noticed it kept giving the same pattern of results – closer to retirement was better.

Then I realised how simple it was: the client would get an upfront tax benefit of (say) 40% on his super contribution in any year.

Spread over the term to retirement, it increased the normal return by approximately 40% / years to retirement. So, 20 years from retirement, a contribution gave an effective return of 40%/20 = 2% on top of the normal return through to retirement. But if you contributed five years from retirement, it gave 40%/8 = 8% on top.

Now, using only pen and paper, I could show the client why it was advantageous to put most money in toward the end, when the tax benefit was spread across just a few years (and there was less risk of legislative change as the term got shorter). And I could still charge a good fee for doing it despite the short and simple consultation, because the value was substantial.

The main insight there was to look at each year’s contribution on its own. Rather than the usual approach of modelling a flow of contributions of $X pa, increasing at Y%, for the term to retirement.

First responder disability

We were valuing a new disability scheme for first responders which paid a large lump sum on permanent disability. As part of the scheme rules, a cap had been set on total cost. If it was exceeded, benefits would be scaled down.

In the first couple of years, claims were much lower, but built steadily until they began approaching the cap. The client asked us when it was likely to stop increasing because if it went over the cap and benefits were reduced, all hell would break loose among their employees.

Their question was impossible to answer because disability is complex. You can’t just pull standard rates off a shelf. First responders, with their risk exposure and the high level of psychological stress claims, also made this a unique situation.

It looked like no model was going to help. How were we to answer the question? The answer we settled on came from the realisation that:

  • stress was a major component of most claims

  • anyone who had worked on the front line for several years could probably list enough stressful events to qualify for disability benefits, meaning a high percentage of staff could claim if they chose

  • the organisations were light on backroom staff, so it wasn’t easy to find alternative jobs for front line people who couldn’t cope any longer.

This flipped the question completely. Instead of asking how much higher the claim rate could go, we asked ourselves why it wasn’t much higher, if this many people already qualified?

We assumed it was because people loved their jobs despite the stress. Keeping stress levels low meant keeping employees happy in their work, i.e., working conditions, pay, morale and so on.

The answer lay (at least partly) in the control of the managers asking the question.

The poorly designed benefits also needed to be changed from lump sums to an income, with regular fitness tests, and the cost cap removed.

That was the answer we gave. Thankfully they did change the benefits.

CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.