What "Release of reserve" really is ?

[Duc Thinh Vu]

Hi everyone,

I'm struggling with the concept of "release of reserve". Could you please explain me what "release of reserve" really is ? What do they mean by the "release" of reserve ?

Is that reserve is "released" only when we no longer need it (i.e. when the policyholder died, or when the insurance policy ends) ?

Thank you very much for your help!

 

[Lindsay Smitherman]

Hi: 'release of reserve' is the same thing as 'reduction in reserve held' or, alternatively, is the negative of 'increase in reserve'.

Therefore yes: when a policy ends for whatever reason, the amount of reserve held will reduce to zero - and so the 'release of reserve' at that point will equal the reserve held immediately before the policy ends.

But we can also have a release of reserve for a policy remaining in-force but where the reserve held is decreasing, such as for a term assurance at later durations. If the reserve at the start of year t is 100 and at the start of year t+1 is 60, there has been a release of reserve of 40 over year t. And this 40 falls into profit (or surplus) as it no longer has to be tied up to back the policy.

 

[Duc Thinh Vu]

Hi: 'release of reserve' is the same thing as 'reduction in reserve held' or, alternatively, is the negative of 'increase in reserve'.

Therefore yes: when a policy ends for whatever reason, the amount of reserve held will reduce to zero - and so the 'release of reserve' at that point will equal the reserve held immediately before the policy ends.

But we can also have a release of reserve for a policy remaining in-force but where the reserve held is decreasing, such as for a term assurance at later durations. If the reserve at the start of year t is 100 and at the start of year t+1 is 60, there has been a release of reserve of 40 over year t. And this 40 falls into profit (or surplus) as it no longer has to be tied up to back the policy.

Hi, thank you very much for your reply. I return to this question after 1 month because I still don't catch the idea of "Release of reserve" in the calculation of profit.

My question is that "Why is release of reserve considered as an element of profit?" (or in other words, why should we care about release of reserve when calculating the profit of a period ?)

The context is as follow: Intuitively, as I understand, Profit of a year t is Income(t) - Outgo(t), with Income(t) = Premium(t) + Investment income(t), Outgo(t) = Claims(t) + Expense(t) (commission, administration fee,...). And yes, as you said in your previous reply, the release of reserve is considered to be a source of profit because we no longer hold as much reserve as the previous period.
However, in my point of view, reserve is actually an "artificial thing" in the liability side, which means that, there isn't indeed any cashflows generated that have something to do with reserve. In other words, I think, from time to time, to calculate profit of a period (a year for example), we shouldn't take reserve into consideration, because: 1) it is just a "artificial number" that tells the company that "hey, you should have XX £ to meet your obligation in the future" and 2) a change in reserve (i.e. release in reserve) does not generate anything to the company (no cashflow, no income, no outgo).

Maybe, the problem is that I don't really understand the relationship between "reserve" which is an element of balance sheet and "profit" which is an element of the income statement.

Thank you so much for your help!

 

[CapitalActuary]

P&L just breaks down the different between one balance sheet and the next. An increase (decrease) in assets is a profit (loss) and a decrease (increase) in liabilities is a profit (loss). If you look at profit net of tax over a year, minus dividends to shareholders, this will be the difference between the net assets at the start of the year and the end of the year.

It sounds like it would be useful to revise the concepts of accrual accounting from CT2 (or CB1 or whatever it's called now). The accounting principles of revenue recognition, matching, accrual (and others) explain a lot of why insurance accounting, and indeed all accounting, is done the way it's done rather than purely based around cashflow statements.

 

[Duc Thinh Vu]

P&L just breaks down the different between one balance sheet and the next. An increase (decrease) in assets is a profit (loss) and a decrease (increase) in liabilities is a profit (loss). If you look at profit net of tax over a year, minus dividends to shareholders, this will be the difference between the net assets at the start of the year and the end of the year.

It sounds like it would be useful to revise the concepts of accrual accounting from CT2 (or CB1 or whatever it's called now). The accounting principles of revenue recognition, matching, accrual (and others) explain a lot of why insurance accounting, and indeed all accounting, is done the way it's done rather than purely based around cashflow statements.

Thank you for your reply, it is quite clear. I have already reviewed the CT2 long before coming here to ask such a question.

But in fact, I think, in accounting, the profit is literally "What I receive" minus "What I have to pay" because it is intuitive like that no matter which accounting principle we use. So therefore, I think we should only consider "what I have to pay" and "what i receive".

Moreover, I think it is unreasonable to consider "An increase in asset is a profit" (or the equivalent statement with liability) because for example, the company purchase a building on credit, the asset increases (and the liability - account payable also increases), but the company does not make any profit or loss. So it is much more complicated than your statement about the relationship between increasing/decreasing of asset/liability and profit.

Could you please explain to me in a little bit further details ? Thank you a lot!

 

[CapitalActuary]

Thank you for your reply, it is quite clear. I have already reviewed the CT2 long before coming here to ask such a question.

But in fact, I think, in accounting, the profit is literally "What I receive" minus "What I have to pay" because it is intuitive like that no matter which accounting principle we use. So therefore, I think we should only consider "what I have to pay" and "what i receive".

Moreover, I think it is unreasonable to consider "An increase in asset is a profit" (or the equivalent statement with liability) because for example, the company purchase a building on credit, the asset increases (and the liability - account payable also increases), but the company does not make any profit or loss. So it is much more complicated than your statement about the relationship between increasing/decreasing of asset/liability and profit.

Could you please explain to me in a little bit further details ? Thank you a lot!

You just explained yourself why the company doesn’t make a profit or loss immediately on purchasing a building with credit. There is both an increase in assets and an increase in liabilities at the same time, which cancel each other out.

The idea that P&L is just an expression of the difference between two balance sheets is absolutely the case and not a simplification or throwaway comment. And it explains exactly why a reserve release contributes to profit.

Profit is not always simply things you receive less things you have to pay, precisely because of concepts in accrual accounting like revenue recognition and matching.

 

[Kamina]

A release of reserves needs to be taken in the context of what else is going on in the income statement.

If everything goes as per assumptions used to calculate the reserve for an annuity book, then a release in reserves (I. E. reduction in liabilities) will be offset by an equal reduction in assets (I. E. Cash going out the door to policyholders). So there would be no profit, as expected.

However, if things don't go as was assumed on the reserves, then you do get a profit/loss element arising

For example, on an annuity book, if you saw higher deaths than expected in the period, then you will end up with a greater release in reserves vs change in assets, hence giving you a profit. In this case you'll have an excess of assets you don't need to hold to support the liabilities, and are hence booked as profit.

This is why when working out profit /loss, you take into account the change in the value of assets vs change in liabilities, as well as the "cash flow" items such as income vs outgo
.

 

[Lindsay Smitherman]

These are great answers to this question - thanks everyone!

Just to add that you should recall from CP1 that profit (for an insurance company) is equivalent to 'surplus arising', which is the change in surplus = change in {value of assets - value of liabilities} over the period. So profit = increase in value of assets - increase in value of liabilities.

And to give a quick example of the release of reserves idea: consider a simplified case where you were holding a reserve of 100 for a policy with only a short period left to run, and then it actually costs you 80 to pay out on that policy, then the extra 20 can be released into profit.

 

[Duc Thinh Vu]

A release of reserves needs to be taken in the context of what else is going on in the income statement.

If everything goes as per assumptions used to calculate the reserve for an annuity book, then a release in reserves (I. E. reduction in liabilities) will be offset by an equal reduction in assets (I. E. Cash going out the door to policyholders). So there would be no profit, as expected.

However, if things don't go as was assumed on the reserves, then you do get a profit/loss element arising

For example, on an annuity book, if you saw higher deaths than expected in the period, then you will end up with a greater release in reserves vs change in assets, hence giving you a profit. In this case you'll have an excess of assets you don't need to hold to support the liabilities, and are hence booked as profit.

This is why when working out profit /loss, you take into account the change in the value of assets vs change in liabilities, as well as the "cash flow" items such as income vs outgo
.

Thanks Lindsay, CapitalActuary and Kamina for your helps. It is quite clear for me what is happening in 2 sides of the balance sheet.

Just to clarify a little bit further, I want to ask "What is the meaning of the change in reserve (i.e. the difference between the reserve(t+1) and reserve(t)) ?"

As for me, I often imagine, for example, the reserve reported at 1/1/2020 as some kind of "account payable" in the liability side of the balance sheet (although it is not exactly the same thing) that the insurer would have to pay back the policyholders from 1/1/2020 to the end of the coverage period. Therefore, the difference between Reserve(1/1/2020) and Reserve(1/1/2021) is "the amount of money the insurer is expected to payback the policyholder in the operating period of the year 2020". During the year 2020, if things go well (i.e. the reality is favorable than expectation), I imagine that the insurer, say, gets a "discount" in the amount of "account payable" (i.e. like at first you should pay X to buy a laptop on credit, but eventually, the supplier only demands Y < X to settle the transaction) and this "discount" will go to some account in the Stockholder's equity part of the balance sheet.

So, reconsider the equation to calculate the profit for a particular year (assuming no investment income): Profit = Premium - Claims + Change in reserve. In this equation, "Premiums - Claims" is the "real amount of payment made during the period" and "change in reserve" is indeed the "amount of payment that insurer is expected to pay". This equation of profit then in fact compares the "real payment made" and the "payment expected to be made", if it is positive, then it is a real profit.

Could you please tell me if my reasoning is persuasive or it should be improved in another way ?

Thank you very much for your help.

 

[CapitalActuary]

Therefore, the difference between Reserve(1/1/2020) and Reserve(1/1/2021) is "the amount of money the insurer is expected to payback the policyholder in the operating period of the year 2020".

That's actually not the case. For example, if when I set the reserves at 1/1/20 I thought there was a 25% chance of a £100 claim on 1/1/2030, I might set the reserve for this at £25 (ignoring discounting etc). When I get to 1/1/21 I revise the probability of the £100 claim on 1/1/2030 to 50% based on some new knowledge / experience / analysis - so the reserve as at 1/1/21 is £50. It's increased by £25 but this has nothing to do with payments over 2020.

Perhaps it would be useful for you to think about how to break down the reserves at 1/1/20 into different bits:

• one bit we expect will be paid out over 2020
• another bit we expect will be paid out after 2020

Then in the 1/1/21 reserves, we'll have:

• the bit from the 1/1/20 reserves that we expected to be paid out after 2020 +/- revaluations to this based on experience (like my example above)
• a new bit in respect of new business we've written over 2020

We'll also have paid some claims over 2020 and these amounts may or may not have been the same as we expected in our 1/1/20 reserves.

Hope this continues to clarify things for you.

 

[Duc Thinh Vu]

That's actually not the case. For example, if when I set the reserves at 1/1/20 I thought there was a 25% chance of a £100 claim on 1/1/2030, I might set the reserve for this at £25 (ignoring discounting etc). When I get to 1/1/21 I revise the probability of the £100 claim on 1/1/2030 to 50% based on some new knowledge / experience / analysis - so the reserve as at 1/1/21 is £50. It's increased by £25 but this has nothing to do with payments over 2020.

Perhaps it would be useful for you to think about how to break down the reserves at 1/1/20 into different bits:

• one bit we expect will be paid out over 2020
• another bit we expect will be paid out after 2020

Then in the 1/1/21 reserves, we'll have:

• the bit from the 1/1/20 reserves that we expected to be paid out after 2020 +/- revaluations to this based on experience (like my example above)
• a new bit in respect of new business we've written over 2020

We'll also have paid some claims over 2020 and these amounts may or may not have been the same as we expected in our 1/1/20 reserves.

Hope this continues to clarify things for you.

Hi, thank you so much for your clarification. It helps a lot.

However, I still don't understand the meaning of the "change in reserve" ? Can you please tell me what does this term mean to you ?

Denote N the end date of the insurance contract and ignore new written policy in 2020. As for me, the reserve in 1/1/20 is the expected amount paid from 1/1/20 to N. The reserve in 1/1/21 is the expected amount paid from 1/1/21 to N. Therefore, the difference should be interpreted as the expected amount paid in the operating year 2020. I think it is reasonable right ?

Thank you one more time for your help!

 

[Lindsay Smitherman]

That was a great response from Capital Actuary - thank you! Recall from CP1 (Chapter 37) that surplus arises due to various items, not just the difference between actual and expected claims and expenses, but also from the inclusion of new business during the period, changes in the valuation assumptions etc.

I would add that the reserve will also change over the year due to the future cashflows that are being valued becoming one year closer (so being discounted by less). So profit / surplus will arise due to the actual investment return achieved on the assets backing the reserves differing from the discount rate used in the reserve calculation.

 

[Lindsay Smitherman]

However, I still don't understand the meaning of the "change in reserve" ? Can you please tell me what does this term mean to you ?

Denote N the end date of the insurance contract and ignore new written policy in 2020. As for me, the reserve in 1/1/20 is the expected amount paid from 1/1/20 to N. The reserve in 1/1/21 is the expected amount paid from 1/1/21 to N. Therefore, the difference should be interpreted as the expected amount paid in the operating year 2020. I think it is reasonable right ?

No, this isn't correct for the reasons that Capital Actuary explained.

Also be careful with the phrase 'Change in reserve'. In your earlier post, you added this into profit statement. If we define 'change in reserve' as 'reserve at period end minus reserve at start period' then this should be deducted from the profit calculation.

Simplistically (for conventional without-profits business), profit = premiums - claims - expenses + investment earnings - increase in reserve

Or, rather than 'minus increase in reserve' substitute 'plus release of reserve' or (equivalent) 'plus reduction in reserve.'

 

[Lindsay Smitherman]

If reserve at year end = 100 and reserve at start of year = 80 then 'increase in reserve' = 20. So profit would be reduced by 20 during that year as a result of having to set aside more funds to meet those reserves.

The reserve could have increased, for example, due to writing a lot of new business during that year (which would not have been 'expected' at the start of the year, as reserves do not include future new business). And/or because the assumptions used in the reserving calculation were strengthened. And/or because the cashflows are now one year closer and so are discounted by less.

 

[Duc Thinh Vu]

No, this isn't correct for the reasons that Capital Actuary explained.

Also be careful with the phrase 'Change in reserve'. In your earlier post, you added this into profit statement. If we define 'change in reserve' as 'reserve at period end minus reserve at start period' then this should be deducted from the profit calculation.

Simplistically (for conventional without-profits business), profit = premiums - claims - expenses + investment earnings - increase in reserve

Or, rather than 'minus increase in reserve' substitute 'plus release of reserve' or (equivalent) 'plus reduction in reserve.'

Hi, thanks for your reply and your patience. However, I would to clarify what I mean with the interpretation of "change in reserve" (or increase in reserve).

Perhaps it would be useful for you to think about how to break down the reserves at 1/1/20 into different bits:

• one bit we expect will be paid out over 2020
• another bit we expect will be paid out after 2020

Then in the 1/1/21 reserves, we'll have:

• the bit from the 1/1/20 reserves that we expected to be paid out after 2020 +/- revaluations to this based on experience (like my example above)
• a new bit in respect of new business we've written over 2020

We'll also have paid some claims over 2020 and these amounts may or may not have been the same as we expected in our 1/1/20 reserves.

As in the qouted post of CapitalActuary, Reserve in 2020 is combined of (1) payment expected to be paid out over 2020 and (2) payment expected to be paid after 2020. We have the same thing for Reserve in 2021. Assuming no change in reserving basis, no new policy written, the increase in reserve is calculated as:
Reserve Increase = Reserve_2021 - Reserve_2020 = Reserve_2021 - (Expected_Payment_2020 + Expected_Payment_after_2020) = - Expected_Payment_2020
(because by assuming no change in reserving base and no new policy written, we have Reserve_2021 = Payment_After_2020)

Therefore, Reserve Increase = - Expected_Payment_2020 which can be interpreted as the "expected" payment in 2020. Maybe I misuse the word "payment" in this situation but what I mean is that if we look at a portfolio of 1000 policyholders for example, then the Reserve Increase can really be interpreted as the "expected" payment made in 2020. By comparing the Reserve Increase and the real amount made in 2020, we can see if the company has profit or loss in 2020 (if ignoring investment income and expense). By this interpretation, we can think about the difference between the Reserve Increase and the real experience as a discount or an increasing in expense in that period. (i.e. Reserve Increase is the expected expense, real experience is the real expense)

My ultimate purpose is to simply understand the intuition of Reserve Increase (or "Reserve release") and why we should take it into account in the calculation of profit when it has nothing to do with any real payment made during the period.

Thank you very much if you can provide me with your thoughts. I am very appreciated.

 

[CapitalActuary]

So, you’re saying that if we don’t write any new business *and* we don’t revalue our reserves at all, then the new reserve is just the part of the old reserve relating to payments after 2021? Sure, that’s true.

And yeah, the reduction in the reserve would contribute to a profit. You say want to understand why it’s got anything to do with P&L when it isn’t a payment or money we receive, i.e. a cash-flow.

So it sounds like your question is… how can things be profits and losses if they aren’t cash-flows? Hence my original answer that pointed you toward CT2.

I think if you want to get any additional value out of this you need to ask more pointed questions. Is there something specific you don’t understand? Some past paper question?

It feels like maybe there is an XY problem going on: https://en.m.wikipedia.org/wiki/XY_problem Or maybe not, but if you don’t understand something, you’ll need to help us understand exactly what you don’t understand.

 

[Duc Thinh Vu]

So, you’re saying that if we don’t write any new business *and* we don’t revalue our reserves at all, then the new reserve is just the part of the old reserve relating to payments after 2021? Sure, that’s true.

And yeah, the reduction in the reserve would contribute to a profit. You say want to understand why it’s got anything to do with P&L when it isn’t a payment or money we receive, i.e. a cash-flow.

So it sounds like your question is… how can things be profits and losses if they aren’t cash-flows? Hence my original answer that pointed you toward CT2.

I think if you want to get any additional value out of this you need to ask more pointed questions. Is there something specific you don’t understand? Some past paper question?

It feels like maybe there is an XY problem going on: https://en.m.wikipedia.org/wiki/XY_problem Or maybe not, but if you don’t understand something, you’ll need to help us understand exactly what you don’t understand.

Hi, thanks for your recommendation, I will come back to CT2 and revise more carefully. Yes, you are right about the original issue that I don't fully understand why things can be P&L if they aren't cash flows.

So yes, i will come back to CT2. Thank you very much for your help and for your patience!

 

[Lindsay Smitherman]

Reserve Increase = Reserve_2021 - Reserve_2020 = Reserve_2021 - (Expected_Payment_2020 + Expected_Payment_after_2020) = - Expected_Payment_2020
(because by assuming no change in reserving base and no new policy written, we have Reserve_2021 = Payment_After_2020)

For completeness (and for the benefit of others reading this thread), as well as assuming no new business & no change in valuation basis, for this statement to be correct we also have to be assuming a discount rate in the reserving calculation of 0%. [Because 'expected_payments_after_2020' in the reserve_2021 and reserve_2020 figures will be discounted to different points in time and therefore would not be directly cancellable - unless there is no discounting.]

 

[Lindsay Smitherman]

Yes, you are right about the original issue that I don't fully understand why things can be P&L if they aren't cash flows.

I tried to address this in what I wrote in this post:

If reserve at year end = 100 and reserve at start of year = 80 then 'increase in reserve' = 20. So profit would be reduced by 20 during that year as a result of having to set aside more funds to meet those reserves.

In other words, one way to think about this issue is in the following way:

If reserves are increased over the year, more funds have to be set aside to back those reserves and those funds therefore cannot be released as profit for the shareholders. Hence there is less profit arising in that period than otherwise would have been the case.

Hence the 'minus increase in reserves' item in the profit calculation.

 

[Lindsay Smitherman]

Essentially, I can see why the change in reserve coming through as profit feels suspicious, since (as you say) it doesn't generate any cashflows. Indeed, the amount of reserves held does not actually impact the total profit that will arise from a policy over its lifetime. That will depend only on the actual experience (claims, expenses, investment earnings) relative to what was loaded into the premium(s).

What the reserves do impact, however, is the pace at which that profit is allowed to emerge or is reported. [This is considered further at SA2 level when we look at different profit reporting approaches.]

The higher the reserves (eg more prudent), the more deferred or delayed the profit emergence is.

 

[Lindsay Smitherman]

For example, let's consider a single premium endowment assurance.

If we calculated profit as just cashflows (ie profit = premiums - claims - expenses + investment earnings) then at the point at which the policy is sold, there would be a large immediate profit recognised (= single premium - initial expenses). And at the maturity date, there would be a large loss recognised (= - maturity claim).

This is an unsatisfactory profit profile. Just considering cashflows in the profit calculation doesn't really work for life insurance business due its long-term nature. We don't know what profit we will actually end up getting until the very end of the policy, which could be in 30 years or more time. So in order to present profit emergence more appropriately, we allow for setting up and releasing the reserves.

Hence at the point at which the policy is sold, we deduct the initial reserve (ie 'minus increase in reserve'), giving a more appropriate initial profit/loss recognition. And at the maturity date, we release the final reserve (ie 'plus release of reserve'), offsetting the large cashflow loss mentioned above. [And for the intervening years, the change in reserves item smooths out the profit emergence similarly.]