The Quest for Value-Based Insurance

One side of the Coin: Value in Underwriting

A lot of folks are talking about Value Based Healthcare. It’s the in-thing now. Two public health specialists (Drs Jeremy Lim and Jake Goh Jit Khong) joined in the IP debate recently (25 April 24) and wrote in the Straits Times recently that “Ultimately, every stakeholder wants ‘Value’”.

Let’s see if the IP insurers are delivering value. To do that, unfortunately, we have to look at the numbers. This hobbit is not particularly a whiz at numbers but let’s give this a shot. And if the figures given in this post are proven to be very wrong, this hobbit unreservedly apologises in advance.

In Jan 2021, The Singapore Actuarial Society (SAS) produced a paper called “Medishield Life 2020 Review: SAS Comments” based on data between 2016 to 2019. A very illuminative table was produced in that paper which was reproduced by the SMA in its Position Statement which was released in March 2021 : Troubled Integrated Shield Plans. Table 1 is a reproduction of that table in the Position Statement.

Table 1: Combined Profit and Loss of 7 Integrated Shield Plan Insurers 2016 to 2019

Year	Gross Premiums	Gross Claims	Management Expenses	Commission	Change in Reserves and Other Expenses	Underwriting Gain/(Loss)
2016	$1608m	$1190m	$106m	$131m	$279m	(98m)
2017	$1859m	$1390m	$126m	$160m	$329m	(146m)
2018	$1836m	$1399m	$140m	$182m	$153m	(38m)
2019	$2143m	$1617m	$166m	$197m	$206m	(43m)
2016-2019	$7447m	$5596m	$538m	$670m	$967m	(325m)
% of Gross Premiums		75%	7%	9%	13%	(4%)
% Increase from 2016-2019	33.3%	35.9%	56.6%	50.4%
CAGR (Compound Annual Growth Rate	10%	11%	16%	15%

The last two rows are additional observations by SMA. The rest of the table is exactly what was presented by SAS. We can trust that SAS are experts in this field and the figures they presented have not been contested by anyone as being erroneous even though there was some debate as to what conclusions one can draw from these figures.

One of the things that I need to say I am happy with the Monetary Authority of Singapore is that they require insurers to submit data in a form called Form A5. These submitted data is then published online. The link is given at the end of this post ^1,.

In any case, this hobbit and his gnome friends have dug deeper into such publicly available sources and mined more recent data.

Table 2: Combined Profit and Loss of 7 Integrated Shield Plan Insurers 2019 to 2022

Year	Gross Premiums	Gross Claims	Management Expenses	Commission	Change in Reserves and Other Expenses	Underwriting Gain/(Loss)
2019	$2143	$1617	$166m	$197m	$206m	(43m)
2020	$2351	$1625	$170m	$202m	$948m	(593m)
2021	$2451	$1807	$177m	$215m	$79m	172m
2022	$2571	$1889	$192m	$215m	$153m	123m
2019-2022	$9516	$6938	$705m	$829m	$1386m	(341m)
% of Gross Premiums		73%	7%	9%	15%	(4%)
% Increase from 2016-2019	20%	16.8%	15.7%	9.1%
CAGR	6.3%	5.3%	5.0%	3.0%

From Table 2, it would appear that things have become stabler from 2019 to 2022 compared to 2016 to 2019. Management Expenses and Commission costs are not increasing faster than gross premiums collected and gross claims paid out.

However, the claims ratio, which is the amount of money disbursed on behalf of policyholders to healthcare providers has dropped from 75% to 73%. As stated in SMA’s position statement, “In USA, the Affordable Care Act or “Obamacare” formally states that Insurers must spend at least 80-85% of premium on health costs; rebates to policyholders must be issued if this is violated. As one can see from the table above, only 75% of premiums collected were paid as claims in 2019 for the IP sector”. We have since gotten even lower than 75% and now stands at 73%, suggesting that the effectiveness or utility of IP as a funding mechanism for private health care has deteriorated further.

The column on “Change in Reserves and Other expenses” is highly variable and subject to individual IP insurer’s actions. For example, in 2020, AIA noted “Long Term Individual Medical Expense: The underwriting result includes a one-off impact of S$(697)m from reinsurance. Excluding this one-off impact, the underwriting result in 2020 is S$(3)m. “. In other words, if we took out this $697m item from AIA, the industry would have made a healthy profit and AIA would not have made a 700m loss in 2020. We do not really know why AIA had to book this one-off 693m item into the books. Is this an accounting treatment and nobody burned through $697m of cash or they really lost all that money in cold, hard cash. This hobbit suspects it is some form of accounting treatment but admittedly, this is just a guess.

The Other Side Of The Coin: Value in Investment (Or the lack of it)

As with all insurance activities, insurers take our money in the form of premiums and surpluses and then invest the premium monies in the quest for good returns (and hopefully these returns will relieve some of the pressure to raise premiums).

In additional to the datasets that SAS has discussed in their 2020 paper, there is actually more information that can be gleaned from Form A5 submissions in the form of Nett Investment Income. This is where the insurer invests the premiums and underwriting gains to generate more income for the company. Whether they are successful or not is another story, as Table 3 will show:

Table 3: Underwriting Gain/Loss and Nett Investment Income: 2019 to 2022

IP insurer	Underwriting Gain/(Loss)	Nett Investment Income/(Loss)	Operating Results (Underwriting Gain/Loss + Investment Income/Loss)
AIA	($664.7m)	($19.4m)	($684.1m)
GE	$38.8m	$23.4m	$62.2m
HSBC_AXA	($13.4m)	$0.2m	($13.2m)
Income	$56.8m	($56.1m)	$0.7m
Prudential	$231.1m	($5.7m)	$225.4m
Raffles	($11.4m)	$0.1m	($11.3m)
Singlife_Aviva	$20.9m	1.6m	$22.5m
Total	($341.9m)	($55.9m)	($397.8)
Total Less AIA’s 697m	$355.1m	($55.9m)	$299.2m

What are the possible conclusions we can draw from Table 3? Here is what this hobbit thinks:

• If we ignore or remove the effect of the one-time AIA’s 697m reinsurance figure, then AIA probably made some money from underwriting (?+32M) and only two insurers lost money in underwriting; i.e. the two smallest ones.
• Insurers often claim that they know how to grow our money, but judging from the data above, maybe they aren’t so good after all.
• Let’s take the example of a real IP insurer. We will call this insurer – Insurer A, a relatively large IP insurer. They were profitable from underwriting activities consistently for the four years in question: from 2019 to 2022. They made money from investment activities from the first two of the four years. Then they lost 13M in 2021 and a whopping 144M in 2022! Basically they had practically lost all their money that they made in underwriting in this period and then some. One can only wonder what they were doing to lose so much in investment activities in one year.
• The smaller IP insurers, such as HSBC_AXA, Raffles and Singlife_Aviva are the only insurers that made money from investment activities in all four years. One large insurer only made money in one of the four years, another insurer made money in three of the four years and yet another in two out of the four years. Many of the seven insurers would have been better off putting the premiums in a bank fixed deposits or buying T-bills.

One may counter that insurance premiums are calculated based on actuarial principles and not based on investment outcomes. This hobbit concedes he does not know enough about the insurance industry and maybe that is true. But having said that, if I were the board director of an insurance company or even just a shareholder, one would want to know how will management address the losses arising from investment activities. How does one cover the big hole of say, ($144m)?

Back To Premiums

And now we come back to the issue of IP premiums. Notice whenever anybody wants something more from the insurers, the insurers will respond by saying whatever you ask for will lead to higher premiums? They seem to be looking for every opportunity to repeat this mantra of higher premiums. This is especially so when doctors lobby for more, whether in terms of asking for more doctors to be empanelled or raising fee schedules.

The LIA has just said a few days ago that empanelling more doctors will lead to higher premiums. If every empanelled doctor already sticks to being reimbursed at the rates prescribed by Insurer X’s fee schedule, how can more panel doctors lead to higher premiums for Insurer X’s policyholders? Can the newly-added doctors charge more than the fee schedule? Do more panel doctors mean more over-servicing? Can over-servicing happen frequently when insurers already hire armies of case managers to handle claims and ask questions that range from the clinically unsafe (e.g. why do you do two ward rounds a day instead of one) to the plain stupid (e.g. Why did you not try conservative treatment first before performing surgery on the patient with anal fistula?).

The fact is the majority of IP insurers are making decent margins and profits from underwriting IP policies in the last few years. Prudential, in particular, is doing exceptionally well and it contributes more than half of the underwriting profits of the entire group of seven IP insurers.

And if you look at MOH data released in the June 2023. (Please see chart 1), you will see that doctors’ fees have been rising significantly slower than hospital part of the bill and overall bill sizes.

Chart1 :

(The Straits Times 14 June 2023)

So let’s get things into perspective, the IP insurers aren’t doing badly underwriting IP policies and doctors are not the main culprits for rising healthcare costs in the private sector. Period. And some IP insurers can even do much better if they don’t make bad investment decisions. I think my grandmother can generate better investment returns than some of these folks.

Extracting More Value: Claims’ Ratio Going from 73% to 85%

The slowly declining claims ratio should be of particular concern to policy wonks. I think the world will be a happier place for policyholders and healthcare providers if the government just nationalised the whole IP sector (i.e. like Eldershield becoming Careshield) and raised the claims ratio to 85%. That would mean $1.1B (73% to 85% = 1.1B) more money over four years to pay for healthcare services, instead of now being spent on the supporting infrastructure of “management”, “commissions” and “other distribution expenses” of IP insurers.

On another note, this hobbit has been told that “other distribution expenses” include incentive overseas trips, conferences  and celebratory meals for insurance agents, and perhaps even media advertisements to show who are the most successful insurance agents. And these amounts are not small. For example, one company spent $31m on “additional distribution expenses” in 2022. These expenses are reflected under the “Changes in Reserves and Other Expenses” column in Tables 1 and 2 and are separate from the commissions that agents earn directly from the insurers.

If we can raise the IP sector’s claims ratio from 73% to 85% and extract about $275m a year more from the system, we can probably even have policy portability without significantly raising premiums. This would answer the question of “who pays for portability”. The answer would be, “Still the policyholders. Just cut out the middleman costs of private IP insurers and raise the claims ratio”.

85% is not a magic figure plucked from thin air. It is a figure mandated in Obamacare.  American healthcare is known to be notoriously expensive and inefficient. If they can mandate a figure of 80 to 85%, why should Singapore settle for anything less than 85%?

Then we can move from value-based healthcare to value-based insurance as well. The current IP milieu is failing on both counts. On one hand, the value in IP underwriting to policyholders can be significantly improved. Instead it has declined from 75% to 73%. On the other hand, it is also obviously not delivering enough value from its investment activities in many instances.

So instead of just talking about getting more value out of healthcare providers and always threatening to raise premiums, lets focus on getting more value out of IP sector and the insurers. The public and policyholders expect and need insurers to deliver value just as much as healthcare providers.

Portability and Choice

On 30 April, the President of SAS, Mr Alex Lee, weighed in on the debate in a letter to The Straits Times Form “Time for an inclusive and informed dialogued about IPs”.

He opined that “The data needed to support an inclusive and informed debate on portability, continuity of care, and other aspects of IP design is not available. That needs to change. Policyholders of different profiles must be heard”. He further commented that portability will lead to increase in costs and premiums because “choices have a cost” and indeed, this hobbit agrees that portability is about having a choice and there is a cost to having a choice, especially when you exercise that choice.

This hobbit has long argued that getting a restructured hospital to run four classes of bed leads to increased costs and less efficiency. It would be far better to have less choice for class of beds. We could derive some savings just by cutting down choice. The term management consultants use nowadays is not to “give a choice” but to “retain optionality”.

So choice indeed costs. Henry Ford implied this truth in his autobiography regarding the production of the Ford Model T, when he said “Any colour the customer wants, as long as it’s black”. The retool his production line to provide for different colours for Model T will raise costs.

Similarly here, why are healthcare providers and some policyholders clamouring for portability? Portability is not an end in itself, it is really a means to the end of extracting more value through competition after a IP policy has been bought and especially if pre-existing diseases have developed.

The government’s disdain for riders is well known. Yet IP insurers still innovate and come up with all kinds of riders nowadays. This is because they need to compete for new business and so innovate to develop more perceived value for the prospective customer to sign up as their policyholder.

But once you have developed pre-existing diseases as a policyholder, you are kind of stuck. No competition occurs and IP insurers see no need to deliver more value, especially when the policyholder gets older. They are probably not very unhappy if you decide to drop your IP plan when are you are 75 years old, having paid IP premiums for decades and are now likely to start making claims.

Portability will force IP insurers to compete post-policy purchase and in doing so, hopefully deliver value. Yes, while this hobbit agrees with Mr Alex Lee, President of SAS that choice costs – But this is from the perspective of the manufacturer (like Ford) or the insurer. From the perspective of the consumer, retaining optionality (i.e. the threat of exercising choice) also leads to better value creation through competition on the other side of the equation.

In Search Of Value: Competition and Regulation

However, competition is not the only way to deliver more value and safeguard the interests of policyholders. Another way is regulation so as to address the “power imbalance” that the President of SMA, Dr Ng Chee Kwan, mentioned about in his letter on 27 April 24.

If some regulator stepped up to the plate and started to do most of the following (non-exhaustive):

• Mandate that the claims ratio for IP business should be at least 80%, preferably 85%
• Not allow IP insurers to change coverage anytime (remember the incident involving an insurer not covering diagnostic scopes suddenly a few years back?)
• Mandate that criteria for empanelment and exclusion from preferred panels be spelled out clearly and publicly
• Regulate the size of preferred panels with respect to the the number of policyholders
• Regulate on how insurers must grow their preferred panels with time
• Regulate what can be put into exclusion clauses for disease coverage. For example, they must be evidence based exclusion clauses. E.g. you can exclude coverage of colon cancer if the patient had colorectal polyps discovered previously, but you cannot exclude coverage of all GI cancers from the oesophagus to the anus just because the person had piles.
• Regulate what can be preauthorised and when preauthorisation can be denied
• Regulate IP insurers on what questions they can ask or cannot ask. Especially questions that subliminally promote a unsafe practice culture through under-servicing, e.g. “Why do you have to do two ward rounds a day instead of one”?
• Appointing a SMC-registered medical practitioner to oversee and be held accountable for decisions by insurers that affect patient safety and clinical outcomes adversely

If the above can be instituted for all insurers, then maybe we do NOT need portability so badly..

Bottomline is, other stakeholders in the IP sector, such as policyholders and healthcare providers are looking for a better deal from IP insurers i.e. they are searching for more value than what insurers are providing.

This can come through either by competition or regulation or both. But at present, both are not present. That is why “the current arrangements are untenable and will come to a boil soon enough” (Drs Jeremy Lim and Jake Goh).

Where’s The Value?

The days where the public and healthcare providers accept that IP insurers

• provide a claims ratio of only 73%
• do not compete after a policy is bought because there is no portability
• and are also at the same time largely not subject to adequate regulation

are probably over. We need to find ways to extract the kind of value from the IP insurers that our people need and deserves.

¹https://www.mas.gov.sg/statistics/insurance-statistics/insurance-company-returns

Postnote: Alert and mythical reader of this column, Dr Harry the Beng has alerted this hobbit that the figure of $1.1B could be extracted by raising the claims ratio from 73% to 85% over FOUR Years, not one year, or about $275m a year. The hobbit is sorry for these errors and has since corrected them. That’s what happens when you post at around midnight, I guess. In any case, this hobbit would like to suggest to the relevant authorities to consider creaming off any excess premiums that exist when the claims ratio falls below 85% and use the cream-off to fund Healthier SG activities Now, wouldn’t that be a much better way of using those excess premiums than letting the insurers keep the money?