At the Actuaries Institute Financial Services Forum, in May, I’m running a Panel discussion entitled Is MOS Accounting Responsible for the Woes of the Retail Life Industry? This post gives some background to the question and why it seemed a good topic for a panel discussion.
What are the woes of the Retail Life Industry?
I posted earlier this year about the issues affecting the Australian life industry as a whole. Annualised profit as a percentage of premium (after tax) has dropped from just over 15% in early 2009 to 2.0% in the year to December 2013. The whole Australian life insurance business (excluding reinsurance that went off shore) made a profit of $249m in 2013, a margin of 2.0% on a revenue of $12.7 billion income. That’s unlikely to be an adequate return on capital.
Retail insurance made more than 100% of the profits noted above; making a profit of $684m in 2013, a profit margin of 8% on $8.3 billion of income. Compared to the Wholesale (group) side, that is quite good, but it is a massive drop in profitability from a few years before. In the year to March 2009 (the first 12 months for which APRA statistics are available) $4.5 billion of income earned retail life insurers a profit of $857m, at a profit margin of 19%.
So why are the returns so bad now?
There are a few main drivers of profitability in retail insurance:
- Commission and other acquisition expense levels – most retail life insurance has high acquisition costs – either high up front commission rates (some over 100% of the first year’s premium) or other marketing costs, such as advertising. Over the life of the policy, the higher the up front costs, the lower the profits. And that leads to our next driver:
- Retention – the longer a contract stays paying premiums, the more chance the company has of recovering those high acquisition costs, and then making profits. Life insurance contracts with attached advice generally only break even over 5-7 years of the life of the contract.
- Claim rates – Around half the premium of a retail life insurance contract goes into paying claims, over the life of the policy, which means that higher claims will lower profitability; in particular
- Duration of claims – the longer an income claim stays disabled, the higher the total payments made to claimants.
Most of these issues are affecting the industry; some parts of it more than others, depending on the way in which insurance contracts were sold, and in some cases, exactly what types of claim can be made on the policy.
So far, all of this seems straightforward. The complexity comes when considering the long term nature of the contract. Because the cost of selling the contract is so high, a life insurer can only be profitable if contracts stay around for a while. That’s why retention is such a key driver. It is also why life insurance accounting is very complex; to try and capture the long term nature of the profitability in a balanced way.
MoS Accounting (Australian GAAP)
Back in 1995, Australia introduced its own method of accounting for these contracts – called Margin on Services (MoS). There is no international accounting standard for insurance contracts, and most methods in use at that time were unsatisfactory for a variety of reasons.
Very briefly, the intent of MoS accounting is to spread the profits of the contract over the expected life of a contract. Providing it is expected to make a profit, the company sets up an asset of the total up-front costs (often called DAC – deferred acquisition costs) at the point of sale. This means that there is no profit or loss at the point of sale. Over the life of the contract, the asset is gradually written off using a complex formula intended to spread the profits as a proportion of the “profit carrier” – generally premiums or claims.
So providing the assumptions made at the outset about the contract (retention, claims, expenses etc) are correct, the profit will emerge smoothly over the life of the contract.
Of course, as every actuary knows, the only certainty is that the assumptions will be incorrect. This presentation from NZ is the only analysis I have seen of experience gains and losses (where the outcome for a particular year was higher or lower than assumption).
So when experience suggests that assumptions should be changed, the pattern of future profits is changed. There is no change to past profits. So if the actuary forms the view that retention is worse than originally thought, or that claims are going to be higher in future, future profits will be reduced. There will be no impact on this year’s P&L (except the impact that came from the claims or retention being worse than expected this year). The only exception to this is if there are no profits left; if the profit is expected to be loss making, this must be recognised immediately.
Issues with MoS accounting
Any issues arising with MoS accounting are the interaction of the reported profits with the management and shareholder response. Ultimately the purpose of accounting for profits and losses is to help understand the health of a business. With an insurance business, there is enough complexity that no single measure will be perfect.
That said, what kind of issues emerge with MoS Accounting?
- Deferral of acquisition expenses – as outlined above, the costs incurred in acquiring new business are deferred, and gradually taken as expenses over the life of the contract. This means that if you use the profit line as the main way in which to measure the performance of your business, these costs will show up many years after they are incurred, and well after any action can be taken.
- Smoothing of assumption changes over the future – if the experience of the portfolio is gradually worsening, that worsening experience is spread over the future life of that portfolio. It doesn’t emerge immediately, until suddenly, it does, if there is loss recognition.
- Spreading of claims cost over future years – this is a special case of the smoothing of assumption changes. For many companies, a change in the assumptions about the duration of income claims (even existing ones) will be spread over the future life of the portfolio.
- No recognition of cost of capital – There is risk involved in insurance. Companies hold capital to guard against the risk of not having adequate assets to pay claims when they occur. Profit accounting does not allow for the cost of holding this capital (which is not an issue unique to insurance).
Does profit reporting drive behaviour?
One of the key questions in all of this is how much profit reporting drives behaviour. Some of the underlying issues in a portfolio that might be obscured by this method of accounting could include:
- Acquisition expenses that are uneconomic for the business being sold
- Overly optimistic assumptions at the point of sale
- Reductions in expected value of business (from changes in views of retention or claims behaviour), even to the point of not making an adequate return on capital, that don’t lead to actual losses
- Profitability of business being highly sensitive to different choices of assumptions
Some of these issues may emerge in a portfolio regardless of how much information managers and shareholders have, and how they report their profits. But if the profit reporting is all they have, will it change how they manage the portfolio?
Your views?
Please feel free to comment below. As is the case with any industry going through a rough patch, there is no single cause, rather a raft of them. But with changes in accounting policy likely to occur in this industry in the next few years, it would be good to tease out issues with our current approach.
Hi Jennifer,
Perhaps it’s merely surprising that the business model (multiple year’s commission upfront for an annual contract with guaranteed renewal at the then extant prices) was stable for so long rather than wondering about recent ‘low’ profitability.
Does the advisor have a duty to the insurance company or to the customer? I suspect a few of us examine market prices on eg motor insurance each year so why not life cover – in a market where long-term insurance prices are not available.
Increase prices each year for age increases may be palatable but increases prices each year because the insurer can’t behave as an insurer and price insurance risk is stretching the credulity of the industry.
I’ve only seen an early draft of the proposed international accounting standards but I think the proposal is to disallow EV of the typical Australian products because of minimal risk transfer.
Maybe you are right, the issue is the business model, rather than the accounting for it. For what it is worth (this may be a follow up post) the main difference with the new accounting standards is that one year’s accounting will have to take place at a time (ie no deferral of expenses beyond one year, and a reckoning of claims reserves in the same time period).
I’m sure that some parts of MoS accounting has exacerbated the issues within the Australian life insurance industry, but I feel that there are many other reasons that are more important drivers.
The largest is a long term insurance cycle with most of the learnings of the last major downturn (around 2000) being forgotten. Since the strengthening that occurred after that period, the product designs and underwriting have slowly been relaxed without being fully understood in totality.
There are a number of aspects that may assist in managing this going forward:
• Increased communicating of the profitability of new business. While this is available under MoS, it is less critical to current performance and hence can be downplayed.
• Improved focus on medical, legal and societal changes that are ongoing and understanding the impacts of these on assumptions etc. This also should become a key part of the FCR, or at least the detailed assumption documents that sit behind the assumptions.
Jennifer, I think you are right on the money with the issues you have identified here.
Your point on smoothing assumption changes over future periods strikes a particular chord with me. It always bemuses me when I see the amount of focus life companies put on the actuarial assumptions for products with no profits left over which to spread adverse changes to assumptions.
i.e. only when the changes hit the P&L in the current year does management really focus on what can be done to reverse the decline. I’d speculate that this focus may have occured earlier if the bad experience not beeen “hidden” by the MoS assumption smoothing mechanism.
I think it highlights a real issue with the accounting framwork when a product needs to go into loss recognition before management switches focus from clever ways to time the release of profit to actually making some!
I do think that the whole industry focuses more on assumptions that affect profits than those in profit making lines of business that don’t affect profits (auditors and analysts tend to have the same focus as management here) but that, to me, focuses the attention on the accounting standards that push us in that direction.
Hey Jennifer,
I think the onus should be on the Actuary to communicate MoS messaging in an appropriate way. It is actually quite a neat profit reporting tool whose benefits should not be so easily discounted.
It is a reporting tool that has allowed profits to be reported in the same manner as other companies in different industries (being extremely important if the life company was listed):
DAC – companies are able to choose between capitalising their costs or expensing them.
Firms that capitalize costs and depreciate them over time can show smoother patterns of reported income. The regulations say that as long as you can reliably show future benefits you can capitalise costs and depreciate (recogonise) it over its expected life. This is awfully similar to DAC and it doesn’t make sense to not be able to do this as long as other companies around Australia can.
Capital – No other company would reflect its cost of capital in its profit reporting.
Smoothing of Assumption changes allow you to recognise the experience as it emerges.
If a mining company thought costs of its operations were going to increase over the next 3 years, it wouldn’t recognise that loss immediately however the business would signal changes would need to be made.
My personal opinion is that Actuaries are not engaging with the business as much as they should. There is too much emphasis on setting assumptions and reporting the profits than engaging with the business in how to change those assumptions and what those profits mean.
This also could be due to the separation of valuation and pricing actuarial teams.
Overly optimistic assumptions could be a bi-product of APRA deciding to move the Appointed Actuary to a compliance role rather than an independent audit. I think if the business were engaged to seek approval from the AA rather than advise, the conversation would have a different tone.
You mentioned that the purpose of accounting for profits and losses is to signal the health of the business. As APRA has taken a conservative approach to insurance profit reporting in the form of loss recognition (which other industries do not have) the onus is on the Actuaries to change any preconceived notions of profit and losses.
Some signals that show experience losses should be interpreted in the same regard as actual losses even if there are future profits to smooth the result.
If the frame of reference of a good and bad result was changed the company wouldn’t be surprised when it hit loss recognition
My view is that Actuaries need to properly engage the business on how to interpret the results as well as how to change them going forward.
Hi David, thanks for the comment – I agree actuaries need to engage the business – my biggest question is whether the current accounting standards make that harder to do.
To begin with, I’m a bit confused why DAC is being brought into the picture. In MoS, DAC is implicit and not explicit like in US GAAP. The implicit deferrable ratio is 100% which means the acquisition commission and expense is deferred in full.
At inception the PL is 0 because the PM = – BEL. Right after t=0 the PM remains stable but the BEL becomes “more negative” as upfront expenses are no longer present after t=0. You can have a negative liability or a positive asset…… correct me if I’m wrong
———————
To answer the open ended question in the above blog, in the big end of town (Aviva, Sun Life, Prudential, Manulife, New York Life etc) nobody really uses IFRS results to assess the performance of a life insurer. When accounting numbers are used they are limited to IFRS Operating Profit.
The days of using IFRS results are now over and the investment world is now more sophisticated. Terms like EV excess growth, EV non operating profit, surplus cash generation, value of new business, economic capital at risk etc are the new fad.
Actuaries in these overseas markets have evolved with numerous “economic profit/loss reporting” roles.
Further to this, with this evolution actuaries are more important and are moving up the corporate ladder (not down).
For example I have been through experiences where major hedge funds have queried Embedded Value reports and have desired a meeting with the EV actuary rather than anyone else.
————————-
Embedded value movement from one quarter to another gives one a fairly good idea of what’s going on due to the capitalisation of model, economic and operating assumption changes. The movement indicates the “health” of the organisation by breaking down the inforce profit transfer, new business strain, expense overrun, economic and operating variances. At a high level an investor can see what areas are being subsidised.
The Source of Earnings on IFRS is also an important report. This is because under US GAAP, the assumptions are locked in at inception. At most investor day briefings the SOE report is queried in detail.
Surplus cash generation is a number which many insurers (Prudential in particular*) quote. Investors are interested in knowing the monetization profile of the inforce business and understanding how much locked in profit will be released. Generation of surplus cash is important as it will fund the new business strain.
This then flows into value of new business. Profitable new business sold today will generate surplus cash in the future.
*http://www.prudential.co.uk/investors/financial-highlights
Thanks for this comment, it is great to have an overseas perspective. We tend to talk about DAC in Australia when it is just one part of the policy liability (confusingly a negative part). It is not, strictly speaking a DAC (because it moves with assumptions about the future) but largely we have an asset on our balance sheet which is mostly a deferred acquisition cost.
For me the important part of your comment was the point you made that it isn’t sensible just to use accounting (IFRS or Aussie GAAP or anything else) to assess a life insurance book – EV is just as important, and quite a few other measures as well. Genuine cashflow is also important, and a measure that can be forgotten as people talk about “cash” earnings (which are actually reported earnings).
I think this is such a classic case of how organisations (and the market) measure management performance. The right weighting to the right performance measures will drive the behaviour. This is really hard to do when the onus is likely on the actuaries in these organisations to educate and to influence senior management on the differences in accounting profits under MoS versus EV value add, and to embed this into decisions – it is a journey years long if you are not already on it. If there was pressure to disclose EV as they do in some markets then this will help drive the right focus.
Perhaps in Australia we should have followed the Europeans down the EV disclosure route. Maybe that is something worth discussing at our panel discussion.
To give a more technical answer….
The twin impact of higher lapses and claims (ie operating assumptions) are being “masked” in MoS. Due to the implicit nature of the DAC under MoS, the DAC write-off on lapses is not explicitly mentioned. Rather the merged profit margins would become lower…… if you follow what I’m saying.
Under US GAAP, variances due to lapses are more explicit due to the DAC write-off being passed through P/L. This is because deferrable expenses are expected to be met by future premiums. Under US GAAP not all acquisition expenses are deferrable but under MoS the implicit nature of it results in all acquisition expenses at time 0 being deferred. Hence MoS is more aggressive.
If you select your profit carrier to be claims, then you can maintain a similar MoS profit in a higher claims environment even with a lower merged profit margin.
In order to overcome the current challenges businesses have to write more new business with a higher profit margin to bring back the merged profit margin to a “pre-crisis” level. This requires both the volume and the profit margin of new business to be greater.
A secondary problem is that higher lapses on the inforce block reduce the expected maintenance expense which can create significant expense overruns.
———————
At an EV level the twin impact of claims and lapses are not great. The capitalisation of assumption changes means that the EV impact can be several times higher than the profit downgrade.
For example from public information, for one insurer the profit downgrade was one number. However the embedded value movement analysis shows the capitalisation of the change in assumption is quite large.
Hence the MoS profit and the EV impacts are very different. There can be instances where one has a MoS profit but a large negative EV operating profit.
————————–
Changes in EV are important because it signals a change in the generation of cash and the overall monetization profile of the inforce block (You can picture the inforce block as being a layering of cohorts of business). Different cohort years would have a different monetization profile – some would be generating lots of cash profit and others would be having losses due to claims costs, reserve increase requirements etc.
If you think through it the bulk of the MoS profit is coming from the cohorts of business that are not generating lots of cash profit.
Rainmaker, I think you have reaffirmed Jennifer’s points. Having accounting standards that require that much complexity in communicating the impacts is surely going to confuse many actuaries, let alone executives.
If you look at the scorecard for a life company CEO in Australasia, what would you see? They are likely to be rewarded on IFRS profit measures, new business growth and customer satisfaction measures. EV measures are not common because they behave differently to IFRS / MoS and so it would confuse the messages. Similarly, aggregate lapse rate measures are difficult to use with a diversified mix of products.
The result is an incentive to increase initial commission. This has no material impact on profit for the year, helps with new business growth and keeps customers (i.e. brokers) happy. You can then leave any problems to the next CEO when the impact of lower planned margins sets in.
I don’t know what the answer is here. Communication maybe, but different accounting standards would help as results invariably come back to what is reported.
Why not consider moving to general insurance accounting, particularly for yearly renewable term business to only allow deferral of initial costs for a year. This would give executives the right incentives and reported profits would be more consistent with solvency ‘cash’ profits. Or, alternatively, force advisers to account for their commissions using MoS too and reduce the pressure on initial commissions.