For a one year commercial insurance policy - is the expiry date the same
date as the inception, but one year later? Or is it the day before exactly
one year later?
e.g. For an inception date of 5th July 2006, is the expiry date of that
policy 5th July 2007 or 4th July 2007?
The policy will expire 4th July 2007.
"Dave" <> wrote in message
news:7Uegg.3613$
> For a one year commercial insurance policy - is the expiry date the same
> date as the inception, but one year later? Or is it the day before exactly
> one year later?
>
> e.g. For an inception date of 5th July 2006, is the expiry date of that
> policy 5th July 2007 or 4th July 2007?
>
Dave wrote:
> For a one year commercial insurance policy - is the expiry date the same
> date as the inception, but one year later? Or is it the day before exactly
> one year later?
>
> e.g. For an inception date of 5th July 2006, is the expiry date of that
> policy 5th July 2007 or 4th July 2007?
On our business policy it has something like 00:00 05/07/2005 through to
23:59 on 04/07/2006 as an example.
"Colin Forrester" <> wrote in message
news:
> Dave wrote:
>> For a one year commercial insurance policy - is the expiry date the same
>> date as the inception, but one year later? Or is it the day before
>> exactly one year later?
>>
>> e.g. For an inception date of 5th July 2006, is the expiry date of that
>> policy 5th July 2007 or 4th July 2007?
>
> On our business policy it has something like 00:00 05/07/2005 through to
> 23:59 on 04/07/2006 as an example.
This is why I am confused. I understand your definition is common practise
but in my opinion running from 05/07/2005 to 05/07/2006 makes more sense.
This is because at the instant in time when it becomes 05/07/2006 (i.e.
05/07/2006 00:00:000000000000000...), the previous policy expires and the
new one begins (assuming the policy is renewed).
If using your method, the policy expires at 23:59 on the previous day, then
technically that means there is a small period (e.g. 1 minute) without cover
(again assuming a renewal). I realise in practise that this will make no
difference, but it's not a very exact method, is it? Even if you say it
expires at 23:59:9999999999999999999999999999999999999999, there is still a
tiny period of time without cover.
"Dave" <> wrote in message
news:6ufgg.1656$
> "Colin Forrester" <> wrote in message
> news:
>> Dave wrote:
>>> For a one year commercial insurance policy - is the expiry date the same
>>> date as the inception, but one year later? Or is it the day before
>>> exactly one year later?
>>>
>>> e.g. For an inception date of 5th July 2006, is the expiry date of that
>>> policy 5th July 2007 or 4th July 2007?
>>
>> On our business policy it has something like 00:00 05/07/2005 through to
>> 23:59 on 04/07/2006 as an example.
>
> This is why I am confused. I understand your definition is common practise
> but in my opinion running from 05/07/2005 to 05/07/2006 makes more sense.
> This is because at the instant in time when it becomes 05/07/2006 (i.e.
> 05/07/2006 00:00:000000000000000...), the previous policy expires and the
> new one begins (assuming the policy is renewed).
>
> If using your method, the policy expires at 23:59 on the previous day,
> then technically that means there is a small period (e.g. 1 minute)
> without cover (again assuming a renewal). I realise in practise that this
> will make no difference, but it's not a very exact method, is it? Even if
> you say it expires at 23:59:9999999999999999999999999999999999999999,
> there is still a tiny period of time without cover.
On top of this, it also makes mid-term adjustment calculations more
complicated. For example a policy that runs from 05/07/2005 to 05/07/2006
mathematically has a difference of 365 days between the inception and expiry
date. However a policy that runs until 04/07/2006 has 364 days. In order to
do mid-term adjustment calculations, a day has to be added onto the
days-remaining count.
Dave wrote:
> "Dave" <> wrote in message
> news:6ufgg.1656$
>> "Colin Forrester" <> wrote in message
>> news:
>>> Dave wrote:
>>>> For a one year commercial insurance policy - is the expiry date the
>>>> same date as the inception, but one year later? Or is it the day before
>>>> exactly one year later?
>>>>
>>>> e.g. For an inception date of 5th July 2006, is the expiry date of that
>>>> policy 5th July 2007 or 4th July 2007?
>>>
>>> On our business policy it has something like 00:00 05/07/2005 through to
>>> 23:59 on 04/07/2006 as an example.
>>
>> This is why I am confused. I understand your definition is common
>> practise but in my opinion running from 05/07/2005 to 05/07/2006 makes
>> more sense. This is because at the instant in time when it becomes
>> 05/07/2006 (i.e. 05/07/2006 00:00:000000000000000...), the previous
>> policy expires and the new one begins (assuming the policy is renewed).
>>
>> If using your method, the policy expires at 23:59 on the previous day,
>> then technically that means there is a small period (e.g. 1 minute)
>> without cover (again assuming a renewal). I realise in practise that this
>> will make no difference, but it's not a very exact method, is it? Even if
>> you say it expires at 23:59:9999999999999999999999999999999999999999,
>> there is still a tiny period of time without cover.
There is no minute without cover. Cover is provided for whole-minute
slots, of which there are 60x24x365 per year, and which are identified
by their start time. Hence a policy deemed to end at 23:59 is deemed to
end at the *end of* minute 59 of hour 23.
> On top of this, it also makes mid-term adjustment calculations more
> complicated. For example a policy that runs from 05/07/2005 to 05/07/2006
> mathematically has a difference of 365 days between the inception and
> expiry date. However a policy that runs until 04/07/2006 has 364 days. In
> order to do mid-term adjustment calculations, a day has to be added onto
> the days-remaining count.
No. A policy which runs from 0000 on 5/7 to 2359 on 4/7 has 365 days in
exactly the same way as one which runs from 1200 on 5/7 to 1159 on 5/7.
The only place it makes a difference is that policies which run noon to
noon would, for the purpose of calculating prepayments or accruals of
premiums, be split to X and a half days to one year and Y and a half days
to the other (where X+Y=364 if the period does not contain a leap day).
"Ronald Raygun" <> wrote in message
news:O_fgg.80247$
> Dave wrote:
>
>> "Dave" <> wrote in message
>> news:6ufgg.1656$
>>> "Colin Forrester" <> wrote in message
>>> news:
>>>> Dave wrote:
>>>>> For a one year commercial insurance policy - is the expiry date the
>>>>> same date as the inception, but one year later? Or is it the day
>>>>> before
>>>>> exactly one year later?
>>>>>
>>>>> e.g. For an inception date of 5th July 2006, is the expiry date of
>>>>> that
>>>>> policy 5th July 2007 or 4th July 2007?
>>>>
>>>> On our business policy it has something like 00:00 05/07/2005 through
>>>> to
>>>> 23:59 on 04/07/2006 as an example.
>>>
>>> This is why I am confused. I understand your definition is common
>>> practise but in my opinion running from 05/07/2005 to 05/07/2006 makes
>>> more sense. This is because at the instant in time when it becomes
>>> 05/07/2006 (i.e. 05/07/2006 00:00:000000000000000...), the previous
>>> policy expires and the new one begins (assuming the policy is renewed).
>>>
>>> If using your method, the policy expires at 23:59 on the previous day,
>>> then technically that means there is a small period (e.g. 1 minute)
>>> without cover (again assuming a renewal). I realise in practise that
>>> this
>>> will make no difference, but it's not a very exact method, is it? Even
>>> if
>>> you say it expires at 23:59:9999999999999999999999999999999999999999,
>>> there is still a tiny period of time without cover.
>
> There is no minute without cover. Cover is provided for whole-minute
> slots, of which there are 60x24x365 per year, and which are identified
> by their start time. Hence a policy deemed to end at 23:59 is deemed to
> end at the *end of* minute 59 of hour 23.
OK, that explains it. Thanks.
>> On top of this, it also makes mid-term adjustment calculations more
>> complicated. For example a policy that runs from 05/07/2005 to 05/07/2006
>> mathematically has a difference of 365 days between the inception and
>> expiry date. However a policy that runs until 04/07/2006 has 364 days. In
>> order to do mid-term adjustment calculations, a day has to be added onto
>> the days-remaining count.
>
> No. A policy which runs from 0000 on 5/7 to 2359 on 4/7 has 365 days in
> exactly the same way as one which runs from 1200 on 5/7 to 1159 on 5/7.
Only if it runs in whole-minute slots, as you described above.
In message <6ufgg.1656$>, Dave
<> writes
>If using your method, the policy expires at 23:59 on the previous day, then
>technically that means there is a small period (e.g. 1 minute) without cover
No, it is the end of 59th minute, the smallest fraction nefore the start
of the next day.
>(again assuming a renewal). I realise in practise that this will make no
>difference, but it's not a very exact method, is it? Even if you say it
>expires at 23:59:9999999999999999999999999999999999999999, there is still a
>tiny period of time without cover.
No, it will be at the end of the period, not the start.
--
John Boyle
In message <jxfgg.1667$>, Dave
<> writes
>On top of this, it also makes mid-term adjustment calculations more
>complicated. For example a policy that runs from 05/07/2005 to 05/07/2006
>mathematically has a difference of 365 days between the inception and expiry
>date. However a policy that runs until 04/07/2006 has 364 days.
Why are you calculating the 'difference' - that is illogical. The
'difference' is not the same as the 'period' because the 'difference'
doesnt count the first day. In effect you are saying that 1 Jan 2006 ois
the 365th day after 1 Jan 2005, when in fact it is the 366th day.
> In order to
>do mid-term adjustment calculations, a day has to be added onto the
>days-remaining count.
>
>
Thats right!
--
John Boyle
In message <b1ggg.1763$>, Dave
<> writes
>> No. A policy which runs from 0000 on 5/7 to 2359 on 4/7 has 365 days in
>> exactly the same way as one which runs from 1200 on 5/7 to 1159 on 5/7.
>
>Only if it runs in whole-minute slots, as you described above.
No, the slots can be any length. i.e. a year or a micro second, it wil;l
always end at the very last moment of the period chosen.
--
John Boyle
John Boyle wrote:
> In message <jxfgg.1667$>, Dave
> <> writes
>>On top of this, it also makes mid-term adjustment calculations more
>>complicated. For example a policy that runs from 05/07/2005 to 05/07/2006
>>mathematically has a difference of 365 days between the inception and
>>expiry date. However a policy that runs until 04/07/2006 has 364 days.
>
> Why are you calculating the 'difference' - that is illogical.
On the contrary, it is quite logical. The difference between two
instants is a period. The only difficulty lies in deducing the
exact instants from the wishy-washy way these things are usually
specified.
> The
> 'difference' is not the same as the 'period' because the 'difference'
> doesnt count the first day.
Ah, but a day is itself a period, so it is illogical to take the
difference between two of them. You need to specify the instants
of the days in question.
> In effect you are saying that 1 Jan 2006 ois
> the 365th day after 1 Jan 2005,
which is true
> when in fact it is the 366th day.
Rubbish! By that reckoning the first day after 1 Jan would be 1 Jan,
but that is obviously not the case because 1 Jan isn't *after* 1 Jan.
The first day after 1 Jan is 2 Jan *and therefore* the 365th day after
1 Jan is 1 Jan of the following year.
>> In order to
>>do mid-term adjustment calculations, a day has to be added onto the
>>days-remaining count.
>
> Thats right!
Well, if you give each day a sequential number, then the period
comprising any number of *whole* days which begins on day A and
ends on day B will always be B-A+1 if it is to be assumed that the
period begins at the beginning of the start day and ends at the end
of the end day. Only if the period begins and ends at the same time
of day on both days (e.g. noon) can you measure the length of the
period as B-A.
In message <vPjgg.80348$>, Ronald
Raygun <> writes
>> In effect you are saying that 1 Jan 2006 ois
>> the 365th day after 1 Jan 2005,
>
>which is true
Rats! I should have said 'after the start of 1 Jan 2005'
--
John Boyle