Thread: Premium Bonds
View Single Post
  #17   Report Post  
Posted to uk.d-i-y
Robin Robin is offline
external usenet poster
 
Posts: 1,681
Default Premium Bonds

On 05/06/2021 21:14, John Rumm wrote:
On 05/06/2021 16:22, Andy Bennet wrote:
On 05/06/2021 15:56, Peter Johnson wrote:
On Sat, 5 Jun 2021 12:52:58 +0100, "Jim GM4DHJ ..."
wrote:

Well that is five £25 wins in five months since I started...is this
normal ?

Depends how big your holding is, I'd say,


agreed

and how recently it was
purchased.


Nope


Recent purchases win more often than older ones because
there are more of them, so a large holding purchased in the last 12
months will produce more winners than a similarly large holding
purchased, say, five or ten years ago.


Complete garbage. Each bond has exactly the same probability of
winning regardless of when it was purchased.


I suspect the logic is that it has the same chance of winning as any
other bond. However as the total number of bonds issued increases, the
chances of any bond winning gets smaller. e.g. if you bought 10K bonds
when there were 10M issued, you would be more likely to win than if you
bought 10K when there were 100M issued.


I think not...

(I presume they adjust the payout from time to time to accommodate that
and the fluctuating returns they get investing the money)

....because your presumption is not correct.

Newly purchased bonds do not enter the draw until they have been held
for a month. This allows the total amount of prizes when they enter the
draw to be recalculated /every/ month to take account of the newly
purchased bonds (and any bonds cashed in, and the then current interest
rate on the fund). So eg if an extra 100,000 people each bought £50,000
bonds in November then by the time of December draw their £5 billion
would add an extra £50 million in prizes available. So the odds of any
bonds winning - new or old - would accordingly be maintained[1]

This is not just any lottery. This is a government lottery set up when
the civil service[2] still knew how to define and execute a fair process

[1] maintained but not /exactly/ the same as befo calculating the
odds is well beyond my ability given, among other things, the effect of
the fixed number of £1 million prizes. But the way that is fixed means
a net growth in the total number of bonds actually /increases/ the
likelihood of anyone individual bond winning something. (I think. It's
a very long time since I heard all this from NS&I folk. And wines have
been taken.)

[2] to be fair, I am confident some still could - e.g. ONS. I'm less
sure about some other depts.


--
Robin
reply-to address is (intended to be) valid