Premium Bonds: Are they worth it? Discussion Area

ashm1

Hi,

The premium bond fund rapidly increased in value in recent years (according to NS&I) but the population is saving less taking on 'cheap debt'. I guess low interest rates might make PB's more of an option.

Is the PB fund being fueled by the property boom ?

I haven't got the data to look back at what happens to the PB fund in a property crash.....


Kind regards,

Ashley.

ReportInvestor

Hi Ashley,

IMHO the PB fund is being fuelled by fiscal drag i.e. the number of higher rate taxpayers has ballooned coz tax bands don't keep pace with wage inflation. So it's slightly better value for more people.

This links to the fact that the upper holding limit has been raised from £20K to £30K.

Then there's the the ubiqitous Mr. Sugar .
I can't think that people are using higher house prices to remortgage in order to invest in PBs . That would be crazy with knobs on.

PBs do little harm if you fancy a flutter; unless, of course, you have debts you need to pay off, or if you really need the extra interest / investment return for retirement essentials.

Here's the main Sunday Telegraph article on PBs. But I can't find a link to the subsidiary piece featuring my dentist & his wife

Gorgeous_George

It is possible that people use the feel good factor, created by low interest rates, to be more carefree with their money. For example, I have £500 of PBs and could use that money to pay off £500 of my mortgage thereby saving about £27.50. It's this £27.50 per year (excluuding compounding effects) that I am gambling on PBs. If IRs were 15%, I would not hold PBs at all.

For me, the main thing to understand is that PBs are a form of gambling rather than investing. The wager isn't the cost of the PB but is the interest that this money could earn. And you do not get your stake back.

Take £30K to your bookies and ask him to invest it for 1 month and to then withdraw the interest and stick it on number 5 in the first race at Ascot. If the horse wins, you receive the winnings. If not, you receive nothing. Next month, repeat. At any time, your £30K can be withdrawn. You've still lost your stake.



GG

ReportInvestor

That's well put, GG, and is probably as accurate a description as you could get of the link between high house prices / low interest rates / PB purchases.

Reestit_Mutton

I'm firmly in the camp of those who disagree with Martin's "mathematics".

From a mathematical point of view, if each individual prize is considered in isolation, you have a negligible chance of winning any of the prizes (even the £50 ones). However, all these negligible chances add up to something more significant when taken as a whole.

If Martin can choose to single out the two biggest prizes simply because it is unlikely that you will win them then he might as well exclude ALL prizes from his analysis for exactly the same reason, resulting in a big fat 0% return to the investor.

To only exclude the big prizes is tantamount to biasing the result - a big no-no in statistical circles. It's amazing how often statistics is abused by untrained practitioners these days.

I'm sorry, Martin. Much as I admire your eye for financial detail, you are no mathematician or statistician.

Reestit Mutton
(wearing his mathematician's hat)

nrsql

Afraid I sort of agree with Martin on this one.
When I calculate the expected rate of return from my PBs I only include £50 and £100 as that's all I expect to win.
Anything else is a nice surprise and will be well above my expected rate and the published rate.

Don't really care what anyone else thinks - that's just the way I am happy in making the decision on whether to invest. The decision is tailored by the possibility of a large prize but calculating variance? no point, I'm happy with my expected rate until I win £1m in dec (or more likely 2 I think).

As to comparing average luck with average savings accounts - not the same thing. There's not a lot you can do about your luck but there's a lot you should do about your average savings account.

p.s. I haven't read Martins article so sorry if I'm not agreeing with it after all.

MSE_Martin

[quote=Reestit Mutton] If Martin can choose to single out the two biggest prizes simply because it is unlikely that you will win them then he might as well exclude ALL prizes from his analysis for exactly the same reason, resulting in a big fat 0% return to the investor.[/quote]

Not at all - the chances of winning a million pounds are statistically negigible even for a £30,000 holder. The chances of winning £50 are very real. That's the difference.

As noted earlier I could've done this via looking at modal or median averages, which would've come up with a similar result.

[quote=nrsql]Afraid I sort of agree with Martin on this one.[/quote]

Why on earth would you be afraid?

MSE_Martin

Reestit_Mutton wrote:

If Martin can choose to single out the two biggest prizes simply because it is unlikely that you will win them then he might as well exclude ALL prizes from his analysis for exactly the same reason, resulting in a big fat 0% return to the investor.

Not at all - the chances of winning a million pounds are statistically negigible even for a £30,000 holder. The chances of winning £50 are very real. That's the difference.

As noted earlier I could've done this via looking at modal or median averages, which would've come up with a similar result.

Martin

Reestit_Mutton

MSE_Martin wrote:

Not at all - the chances of winning a million pounds are statistically negigible even for a £30,000 holder. The chances of winning £50 are very real. That's the difference.

As noted earlier I could've done this via looking at modal or median averages, which would've come up with a similar result.

Martin

Martin,

I don't think I explained myself properly...the point I was trying to make by arriving at a 0% payout was simply to show that by a valid extension of your statistical logic (induction) we can arrive at a mathematically contradictory conclusion ergo, your logic was incorrect in the first place. (standard logic theory)

EDIT: I guess it would be sensible to spell out explicitly the steps taken to reach the above conclusion so that there can be no misinterpretation about the point that I am trying to make.

1. You correctly surmised that the premium bond holder has many more chances of winning one of the thousands of £50 prizes on offer. However, each individual £50 prize (the result of a single number being selected from billions of numbers) occurs no more regularly than any of the the £1M prizes. Thus, from a logical point of view, if your reasoning is correct, it would be correct whichever two prizes are excluded from the analysis.

At this point I think it is important to make a clear distinction between the union of ALL £50 events (which is significant) and the individual £50 events (which, on their own, are negligible). You seem to be talking about the former whereas I am talking about the latter - this is where the misunderstanding lies.

2. Anyway, if we were to accept your logic in excluding two prizes from your analysis, then the same logic would allow us to exclude any third prize from the analysis because it too (as a single number selection from billions of remaining numbers) is unlikely to occur.

3. Equally, if you have excluded K prizes already, what's to stop you adding a K+1th prize to the excluded list as it too is unlikely to be won for exactly the same reason.

4. By a process of mathematical induction you will eventually have excluded ALL prizes from the analysis, leaving zero winnings to be distributed.

5. As someone has to win something, we have a contradiction. Therefore, the original logic must be wrong! i.e. all prizes must be included in the analysis for the analysis to be valid.

This argument is nothing more than standard mathematical proof.

I'm not against the basic thrust of your article (although, as with anything, I can think of one or two circumstances where investing the maximum would prove to be a beneficial thing to do from a purely pragmatic point of view) - I just disagree with trying to underline your point through an inappropriate use of statistics.

As for using a modal/median argument...maybe that's what you should have done in the first place (but don't forget to include ALL prizes). It's also worth considering how long you need to be invested before your statistical sample begins to look like the distribution of prizes.

However, to be perfectly pedantic about this, from a theorem proof point of view I can think of an infinite number of calculations with this data that would give me a similar answer (by virtue of the fact that there is an infinite number of ways to add and subtract numbers until you get what you want). Thus, while you may come up with a "similar" result via a modal/median analysis, it doesn't necessarily make your current analysis any more or less valid/invalid.

regards,
Reestit Mutton
(with his mathematician's hat on)

stevemcol

Reestit Munton
Martin's assessment might not stand up to pure mathematical analysis but is pretty much bang on as far as real world is concerned.
If you're looking at pre bonds as a source of garanteed or even probable interest, of course you should dismiss the high end prizes. An individual will not win either of the top two prizes. It is pragmatic then to take into account only probable wins to calculate probable interest.