Losing at Premium Bonds...

Milarky · 24 September 2004 at 9:05PM

Right, this is meant to "set the cat amongst the pigeons"..

This was prompted by an item in last Sundays Times' 'Money' section about Premium Bonds, which was fairly well informed.. unlike some coverage. At one point, a 'spokesman for National Savings' was quoted coming out with some strange rationalisations for why the prize distribution has changed [there are relatively fewer larger prizes than before, so that the chances of winning anything other than £50 is more remote than before]. He said something like: "People have expressed a preference for winning a smaller prize rather than no prize at all"[!] and "Because so much money has been pouring into Premium Bonds lately, we had to stretch the odds on the largest prizes - eg £1,000,000, £100,000, £25,000, £10,000"[!]

Anyway, he went on to describe what a Bondholder with "average luck" could expect to win, simply by reference to whatever the "interest rate" [the value of the prize fund divided by the value of all PB holdings - it's about "2.8%" at the moment] is at the time, and it was this statement that got me thinking....

Q. So, how 'lucky' does a Bondholder have to be to benefit from "average luck"? 50%, 30%, 20%?.. It's less than you think...

A. First. The average prize in a PB draw is £56..

In a 'typical' draw there are

1 prize of £1,000,000
4 prizes of £100,000
6 prizes of £50,000
15 prizes of £25,000
37 prizes of £10,000
73 prizes of £5,000

These 'large' prizes make up 10% by value of the total prize fund

Then there are..

1,123 prizes of £1,000
3,369 prizes of £500

which also make up 10% by value of the prize fund

Finally there are about...

12,672 prizes of £100, AND
985, 974 prizes of £50

These are by far the bulk of the prizes and contain 80% by value of the prize fund.

From these figures there is about £56.2 million in the prize fund divided between 1,003,274 winners - which is £56.

To have 'average luck' at Premium Bonds is to win prizes with at least an average value of £56.

The smallest prize is £50 - representing about 56 of every 57 prizes in each draw.

Second. Assume that you are a 'large' Bondholder - whose holding allows them to win with 'average frequency'. If you win prizes more or less often than than the 'average' [which you may do] then will not be winning with 'average luck'

To 'break even', and win an average of £56 per prize, you will need to 1 prize in every 8 to be at least £100 in value..

But this is unlikely, and the approximate likelihood of this happening is 9.8% that the 'eighth' prize will be £100, and about 3.4% that the 'eighth' prize will be over £100...

Calculations: [skip if bored..]

since the chances of this happening in any ONE draw are actually between 1 in 78 [the ratio of £100 prizes to all prizes] and 1 in 57 [the ratio of any prizes greater than £50 to all prizes] the chances of this happening AT LEAST once in the required 1 draw in 8 are something like 100% minus the chances of it NOT happening. For £100 prizes that is about 9.8%, or

100% x [1 - (77/78 )⁸],

and for prizes above £100 that is about 3.4%, or

100% x [(77/78 )⁸ - (56/57)⁸]

So there is actually about a 1 in 7 chance of having 'average luck' at Premium Bonds... assuming that you win prizes with only 'average' frequency...

[ ..Resume here]

Conclusion

Less that one Bondholder in 6 with average winning frequency will therefore 'not lose' [compared to the equivalent amount of tax free interest they could have earned just by having the same holding in another National Savings account] The remaining bondholders [about 86%] will 'realise' a small average deficit of not less than £6 per prize.. So they will get only '50/56ths' of 'average luck' instead..

The 'chances' of a 'large' Bondholder having average luck are just 13.2%... Their chances of having 'more than average luck' are just 3.4%..

And Bondholders with the largest holdings [of £30,000] have the highest probability that they will win with average frequency in the first place.

pheww..!

That's my 'best' answer I have been able to come up with so far...

Can anyone come up with an answer that substantially alters the conclusion that PBs really are a waste of money if you have a large sum to 'invest' ?

lswwong · 25 September 2004 at 12:01AM

I have this article right in front of me!

Alas, the sums are beyond me. But I take the figures quoted in the article in good faith. My mind only took in the facts that:

Odds of winning £1m PB prize in October 2004: 1 in 25bn.

Odd of winning lotto jackpot at present: 1 in 14m.

(Virtually) No point at all buying PBs in the hope of scooping the £1m jackpot ... (cathy might have something to say about this).

The article says PBs might be worth looking at if one is a higher rate taxpayer. So I suppose that rules me out immediately .......

Milarky · 25 September 2004 at 1:43AM

Yes LSW, higher rate taxpayers do 'benefit' more than basic taxpayers - but that is true of any 'tax free' product, and National Savings offers [upto certain limits] alternative 'tax free' products - such as index-liked and fixed-interest certificates. They also offer a 'guaranteed equity bond' but if this were also 'tax free' not only would it be very 'popular' with savers [as it is really a 'savings' product] it would also be completely 'unfair competition' to every other 'bank' offering the same thing as a 'taxed' product..

Obviously National Savings can 'trade' upon the 'tax free' banner' and so is able offer otherwise mediocre accounts and rates.. I think that Premium Bonds are undeservedly pushed as a 'good idea' and are truely second rate...

There! see you in court NS&I! :P

smartsaver · 25 September 2004 at 2:26AM

I held £30.000 of stoozed money for 6 months or so but the return was quite a lot less than the highest e-saver . Most I won in a month was £100.00, some months nothing and a couple of £50.00. >:(

All very nice but I get much more in Cahoot 5.65%... In fact ING at 5% would beat the returns hands down.

Maybe you will be luckier.. :-/

Reaper · 25 September 2004 at 2:32AM

When people have asked I have always pointed out the low "interest rate" on premium bonds, but I haven't gone as far as to say don't buy them.

The important thing is to understand the low payouts as you correctly point out. I often hear people saying X wins at least £50 every month and assume X must be making big profits. Not true if it cost £30000 to get it.

On the other hand people like the idea of gambling and having the fun of win envelopes arriving through the letter box. While the odds are worse than the national lottery at least you can get your stake back whenever you want.

I would suggest only using spare cash for this. I know some people use the money from 0% credit cards for it and that seems reasonable as it hasn't cost them anything.

Personally I do not buy premium bonds and have no intention of doing so. I prefer the certain win of an offset savings account for my credit card money. But if people must gamble they could do a lot worse.

Loaded_Bachelor · 25 September 2004 at 3:56AM

I wonder why there is no competition in premium bonds unless its something to do with gaming laws?

Surely with competition companies would be able to offer the same type of product but with 3.8% net prize payout fund or a 2.8% net prize payout fund plus 1% actual guaranteed interest.

Maybe they could have even moreflexible options. Such as the choice of which premium for fund you go in. A high prize size, low number of prizes fund, a win more premium bonds fund, a win a higher guaranteed interest rate fund etc etc.

Bring it on......... an absolute fortune to be made by the providers if they can get past the regulators.

Loaded_Bachelor · 25 September 2004 at 4:21AM

(which premium bond fund you go in for - that should have said)

Going on from what I just wrote, I think the government have missed a trick with how they currently distribute the prize fund.

If they split up some of the £50 prizes, (let's say 20% of the £50 prizes), into £2 prizes and these £2 prizes were automatically converted into more premium bonds then they would give investors actual visible growth in their holdings.

I know that you can already take the option for more bonds rather than a cashable cheque (although I doubt most holders do this) but the main reason I am advocating some smaller automatically bond converted prizes is for the benefit of and the procurement of the smaller premium bond investor.

At the moment, as previously pointed out, it's near on having to own 28,000 bonds to on average get a monthly win. This means that someone who has £250 of bonds will be waiting over 9 years to get 1 win !!

By converting 20% of the £50 prizes into £2 prizes the 9 year figure would come down to about 18 months. The £2 prizes would add growth to people with large holdings as well as small ones.

I hear you cry "what about increased administration costs making this idea costly for National Savings to run?"

Well I think that these smaller prizes would only be notified by post once a year or maybe even every two years and/or they would only be visible on line or when you decide to liquidate your holdings.

Given that these £2 prizes get automatically re-invested National Savings would re-coup much of any additional admin cost this way and of course the main thing I believe is that many more small investors would take the plunge and less small investors would get fed up and bail out because of waiting years and years for a win.

I don't think the bigger investors would be too upset either because you would merely be substituting 20% of lets's face it pretty unexciting £50 wins for some smoother flowing growth in their holdings.

MJSW · 29 September 2004 at 8:53PM

Obviously National Savings can 'trade' upon the 'tax free' banner' and so is able offer otherwise mediocre accounts and rates.. I think that Premium Bonds are undeservedly pushed as a 'good idea' and are truely second rate

I agree that many of their rates on their own are poor, but obviously if the product is taxfree then that has to be taken into account as well. Whilst the 2.8% rate you are quoting for Premium Bonds is currently correct, it is slightly misleading in my opinion. The terms of Premium Bonds require that at least 3 months notice are given before any changes in the rate (which is obviously bad for the investor whilst rates are increasing, but good when they are falling). Consequently the current rate of 2.8% does not yet reflect the last 2 base rate increases. The increases already announced to the rates are 3.0% from 1 October and then 3.2% from 1 December.

For higher rate taxpayers, 3.2% tax free is equivilent to 5.33% gross. Whilst this isn't quite the best rate available, it is still pretty competitive, and better than the likes of Ing. (Although obviously because of the nature of the product you could end up with more or less than 5.33%, particularly for small holdings.)

As for the other products, I agree they are mostly fairly poor, with the obvious exception of the Index Linked Savings Certificates. These are in my opinion easily the most attractive deposit type product on the market for higher rate taxpayers, apart from cash ISAs. Inflation is currently 3.2%. The 3 year bonds pay inflation + 1.1%, making 4.3% tax free at current inflation rates. That is equivilent to a gross rate of 7.17% for a higher rate taxpayer. The 5 year bonds pay inflation + 1.25%, current gross equivilent is 7.42%.

Obviously the final return depends on future inflation levels, but if it stays around current levels, then the equivilent gross rate for higher rate taxpayers is around 2% better than the rates available on instant access accounts. Even if inflation drops by up to 1%, the rates will remain at least as good as current instant access rates. If inflation starts to drop below 2% or so, then you can always withdraw without notice (although you get no interest if you have held for less than 12 months, so you need to be confident that the money won't be needed that soon) . There is no penalty as such for withdrawals, but the average excess over the inflation rate will fall since the rates are stepped (eg on 3 year bond, it is 0.85% in year 1, 1.05% in year 2 and 1.41% in year 3).

With smart timing of investment and withdrawal, you can achieve even higher returns. Changes in RPI are not uniform during the year, there are fairly consistent patterns. The most obvious of these is the typical leap between the March to April RPI (presumably due to Budget increases in duty and rates rises). In 2004 the inflation in this one month alone was 0.6%, in 2003 0.7% and in 2002 also 0.7%. This may not sound a great deal, but if you get the timing right, you can invest for 13 months and the extra 0.6% plus additional interest in the extra month adds around another 1.2% to the equivilent gross return.)

Milarky · 29 September 2004 at 9:40PM

With smart timing of investment and withdrawal, you can achieve even higher returns. Changes in RPI are not uniform during the year, there are fairly consistent patterns. The most obvious of these is the typical leap between the March to April RPI (presumably due to Budget increases in duty and rates rises). In 2004 the inflation in this one month alone was 0.6%, in 2003 0.7% and in 2002 also 0.7%. This may not sound a great deal, but if you get the timing right, you can invest for 13 months and the extra 0.6% plus additional interest in the extra month adds around another 1.2% to the equivilent gross return.)

An excellent and highly informative reply, MJSW! :)

But can you enlarge on these final points, please? Are you saying that the rate of inflation [RPI] is 'paid' month by month of a holding - rather than simply being the change in the RPI in previous 12 months - and so if you take out a certificate for just 13 months [March to following April?] you would get 'inflation' for the 12 months + 'inflation' in the peak month of March + first year's bonus of 0.85%?

Thanks in advance

Milarky ;)

MJSW · 30 September 2004 at 12:02PM

Yes, the changes in RPI are calculated on a monthly basis. A few extracts from the Terms & Conditions:

9. The amount due when cashing in a Certificate which has been held for at least one complete month from an anniversary date will be the anniversary value on that anniversary date plus:

(a) index-linking for each complete month from that anniversary date to the date of repayment; and
(b) 1/12th of the annual interest for each complete month held from that anniversary date.

16. An index-linked value will be calculated as V x B/A where:

(a) 'V' is the value of the Certificate at the beginning of the index-linked period (this will be the purchase price or the value at an anniversary date);
(b) 'A' is the Index figure applicable to the calendar month in which the first day of the index-linked period falls (this day will be the purchase date or an anniversary of it); and
(c) 'B' is the index figure applicable to the calendar month in which the day after the final day of the index-linked period falls. This will be the maturity date, an anniversary date, or the day after the last completed month for which index-linking is earned.

Just to complicate matters further, if you want to buy based on the RPI in March, then you need to purchase the bonds in May. The initial RPI figure used is the RPI published in the previous calendar month. So a purchase in May is based on the RPI published in April, which relates to the RPI for March!

Another slight advantage is that you can earn 'double interest' for up to 10 days with careful timing. If you purchase online with a debit card, the start date of the Index Linked Bonds is the date of the online application. They then send you a written confirmation by post, which needs to be returned within 5 working days. Only once they have received this will they process the debit card transaction, and in my past experience this takes around another 3 or 4 days. So during this initial period, you are earning interest both on the Index Linked Bonds and then money is still in the current account earning interest too!

I'm not sure how much interest you would lose at the end though, as I haven't cashed any of my bonds yet. Payments are made by BACS transfer, and term 37 states: "The date a Certificate is cashed in will be taken to be the date on which, in the normal course, the payment would be credited to the nominated account following initiation of the electronic transfer by the Director. In the case of BACS, this is normally two banking days after the Director initiates the transfer." So there is possible no loss of interest. However, interest is only paid for complete months, so you would need to cash in just after the monthly anniversary to avoid losing part of a month of interest.

Robert_Sterling_3 · 30 September 2004 at 12:50PM

Odds of winning £1m PB prize in October 2004: 1 in 25bn.
Odd of winning lotto jackpot at present: 1 in 14m.

These stark figures do not take account of the fact that you can get 500 chances of winning ( one a month ) for each bond you hold.

I know I have had a bond for 40 years or more ... No luck yet though.

Losing at Premium Bonds...

Comments

Confirm your email address to Create Threads and Reply

🚀 Getting Started

Categories

Is this how you want to be seen?

Get our FREE Weekly email full of deals & guides - and it’s spam-free