Premium Bond new "interest" rate

Mikeyorks

Not quite sure where all the alleged 'deviousness' comes from ? This seems clear enough on the NS&I site :-

To fund the extra jackpots, we are increasing the prize fund interest rate by 0.15% pa for the December 2006 and June 2007 draws. We will adjust the percentage shares of the prize fund allocated to each prize band to ensure that the overall distribution of prizes remains roughly as it would have been.

The overall number of prizes in the anniversary prize draws will be the same as they would have been, depending on the underlying interest rate and odds at the time. The table above shows an estimate of the number of prizes of each value for the first anniversary prize draw in December, based on a prize fund interest rate of 3.55% pa (ie the current rate of 3.40% pa plus 0.15% pa to fund the extra three £1 million jackpots) and odds of 24,000 to 1. The prize fund interest rate and the odds of winning are both variable, so they may change from time to time.

I've italicised one of the bits that seemed to be deemed contentious?

Stonk

Mikeyorks wrote:

Not quite sure where all the alleged 'deviousness' comes from ?

The 'deviousness' derives from the fact that they state one known future interest rate, but not another, on the page called "our interest rates". I'm not sure where the text you quote comes from, but it's on another page from the one that people will naturally look if they want to know about interest rates!

This is analogous to a bank deciding to have an interest rate of 25% for a single day, and then publishing this fact on their interest rate tables as "from 1 Dec 2006, 25%", but neglecting to mention that "from 2 Dec 2006, 0.1%".

nrsql

They could argue that the interest rate is for the prize fund not what the PB owner expects to receive.
The importatnt figure is the 24000-1 cahnce of winning which doesn't change (as indicated on the site).
I think it's misleading to call it an interest rate in the table rather than interest rate for prize fund - but that is stated just above it.

Dan29

nrsql wrote:

>> AER = 11/12 of the interest rate because you are left our of the draw for one month.
Not really - you are included in the draw if you hold the bonds at 1st of month. So can hold for 12 months and a couple of days to get included in 12 draws.

Not true. "Your Bonds become eligible to take part in the prize draws once you have held them for a full calendar month following the month you bought them." In other words if you hold the bonds for between 12 months and 13 months you will only be entered into 11 draws.

Stonk

nrsql wrote:

The importatnt figure is the 24000-1 cahnce of winning which doesn't change (as indicated on the site).
I think it's misleading to call it an interest rate in the table rather than interest rate for prize fund - but that is stated just above it.

I agree it isn't really an interest rate - hence my quotes around it (when I remember).

You do need to know the "interest rate" though; the probability figure does not represent the situation completely. For example, it has been 24,000-1 for quite a while now, even though the "interest rate" has changed several times.

When the "interest rate" is higher, although your chance of winning is the same, there is a better chance that, given you do win, you will win a larger prize. In particular, the ratio of £100's to £50's increases. That is vital when determining the expected value of the investment (simplifying it to, for instance, "I have a 24,000-1 chance of winning £50" does not capture the true position).

nrsql

Dan29 wrote:

Not true. "Your Bonds become eligible to take part in the prize draws once you have held them for a full calendar month following the month you bought them." In other words if you hold the bonds for between 12 months and 13 months you will only be entered into 11 draws.

no.
You only have to hold the bonds on the 1st of the month to be included in the draw (as long as you have held them for a calander month or reinvested winnings)

Buy bonds 31 jan sell mar 2
held for 1 month and a couple of days - 1 draw
Buy bonds 31 jan sell apr 2
held for 2 months and a couple of days - 2 draws
....
held for 12 months and a couple of days - 12 draws

nrsql

Stonk wrote:

When the "interest rate" is higher, although your chance of winning is the same, there is a better chance that, given you do win, you will win a larger prize. In particular, the ratio of £100's to £50's increases. That is vital when determining the expected value of the investment (simplifying it to, for instance, "I have a 24,000-1 chance of winning £50" does not capture the true position).

I would disagree - hence my comment about valuations being personal.
I only consider the 50-100 prizes as that's all I expect to win. Anything else would be nice but I don't expect it and would just be inclued in the volatility of the return if I had a measure for it.
There are one and a quarter million ish 50-100 prizes and about ten thousand others.
If you ignored the 50-100 prizes it would be very misleading to try and correlate the interest rate of the prize fund for the remainder with the expected return.

As I say - valuation is different for everyone so go with what you feel comfortable - that's just how I value the investment (I'm not far out but that's just a coincidence).

Stonk

Fair enough, then - consider only the 50's and 100's. My comments still apply. Now that the "interest rate" has increased, you'll receive a higher ratio of 100's to 50's than you did before. If you think only about the odds (24000-1), then you aren't capturing that effect.

Consider an extreme example. Right now, there are far fewer 100's than 50's. But suppose that they decide to double the "interest rate", but leave the odds the same, by paying a higher ratio of 100's. You will win the same number of prizes, but more of them will be 100's, and your winnings will be almost twice as much. But according to you, it's still 24000-1 and nothing has changed.