State Saving Certificates

Discussion in 'Deposits' started by Logo, Jan 30, 2017.

  1. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    Just stumbled on this thread. Can I suggest that dub_nerds excellent post from a previous thread is promoted to Key Post status. This is by way of update. I will work with 100k. Assuming all prizes are €50 the annual expected return is 0.60% i.e. 12 wins. So taking Rabo 90 day deposit just for argument I only need to win 5 times in a year to beat that. Using the Poisson distribution (which is a perfectly acceptable approx for the binomial here) I reckon the chances of not beating that are less than 1%.

    I agree that 10k Prize Bonds should not be compared with bank deposits. But these days we are talking about interest amounting to the cost of one lotto ticket per month and Prize Bonds are far better than lotto tickets in terms of windfall payout ratio.

    In summary, for large amounts use Prize Bonds as alternative to bank deposits and for small amounts use them as alternative to the Lotto:D
     
  2. dub_nerd

    dub_nerd Frequent Poster

    Posts:
    1,358
    Actually, I'd like to suggest that dub_nerd gets sent to the dunce's corner! :(

    The Poisson distribution just reproduced all my carefully calculated discrete odds to four decimal places! Looks like my efforts were a bit misdirected.

    But thanks for the enlightenment anyway!
     
  3. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    I don't know why you are beating yourself up. Your original post is a classic.
     
  4. dub_nerd

    dub_nerd Frequent Poster

    Posts:
    1,358
    Last edited: Apr 6, 2017
    You mean this one. Yeah, I'm just embarrassed how much work it took to produce those graphs when your Poisson distribution does it perfectly well.

    EDIT: on the up side, I can finally produce a Google Sheet that let's people do the calculation for themselves. My original method involved factorials too large for any spreadsheet.
     
    Last edited: Apr 6, 2017
  5. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    Yep, that's the one. Posting an online link using 300,000! could crash the World Wide Web.:rolleyes:

    Are you saying you can now post a Poisson friendly tool? That certainly merits a Key Post.
     
  6. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    Last edited: Apr 7, 2017
    Here is a Poisson calculator.

    In the box named x enter your target number of €50 wins in a year and in the box named average rate of success enter the average number of Prize Bonds i.e. currently .60% of your investment amount divided by €50 or more simply 12 per €100,000.

    The respective figures for 100k are x = 5 to beat Rabo 90 day notice saver and 12 for average number of Prize bonds. For other sums simply pro rata these numbers e.g. 500k (max joint holding) enter 25 (or whatever your target) and 60.

    Notes: multiply the probabilities by 100 to get percentages. X must be a whole number (you can't half win) but average rate of success can be any number.

    (Apologies if any French slips into the above link, I am doing this in the South of France:p)
     
    Last edited: Apr 7, 2017
  7. dub_nerd

    dub_nerd Frequent Poster

    Posts:
    1,358
    Last edited: Apr 7, 2017
    Excel and Google Sheets both croak above 170! :mad:

    Here's a first cut. Click tabs along the top for graphs. Any suggestions for improvement?

    (P.S. Now that I recall, I did know the Prize Bond return was shaped like a Poisson distribution ... I've no idea why I didn't cop that it could actually be computed that way: http://www.askaboutmoney.com/thread...atives-to-deposits.182819/page-8#post-1386534).
    (P.P.S. Pity I'm figuring it out just as the return from Prize Bonds hits a level where it's hardly worth the hassle to invest).
     
    Last edited: Apr 7, 2017
  8. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    Last edited: Apr 8, 2017
    Good stuff dubnerd. Possibly higher sums in multiples of 100k up to the max of 500k could replace the smaller sums for which we agree it is not an alternative to deposits. However the range of outcomes increases proportionately and so hard to fit on one graph. An alternative is to show the 1, 5, 10, 25, 50, 75, 90, 95, 99 centiles.

    I know it sounds ridiculous to talk of half a million in PB but it does beat deposits hands down. Also IMHO it is safer than deposits (I have argued that the 11bn or so of State savings are the last to be welched on if it ever comes to that).

    Your data is in tabular form, I have supplied a calculator for individual sums, albeit it needs a bit more effort by the user.
     
    Last edited: Apr 8, 2017
  9. dub_nerd

    dub_nerd Frequent Poster

    Posts:
    1,358
    Finally! Just found the piece of the puzzle I'd been missing ... the Poisson Limit Theorem.
    (And I get it, so maybe exiting the dunce's corner ;))
     
  10. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    As an actuary the duke is well used to using Poisson for example in modelling mortality outcomes:rolleyes:
     
  11. dub_nerd

    dub_nerd Frequent Poster

    Posts:
    1,358
    Last edited: Apr 9, 2017
    Ok, try this one for size. Enter a new value in the investment amount cell. May take a couple of seconds for the chart to be re-centred to show columns with values > 0.1%. Anyone can edit, so it may behave a bit strangely if multiple people do it at the same time.
     
    Last edited: Apr 9, 2017
  12. Duke of Marmalade

    Duke of Marmalade Frequent Poster

    Posts:
    1,632
    Excellent, and I now have Google sheets on my iPad:)
    The graph for 500000 looks like the half skeleton of a fish which in French is poisson, just saying:rolleyes:

    Users should note that in effect the input units should be thought of as €-years. Thus 100000 can be thought of as 100k for 1 year or, for example, 20k for 5 years. This is not an entirely trivial observation.
    The graph of 20k for 5 years is tighter in shape than 20k for 1 year, albeit the centre of gravity adjusts proportionately.
     
  13. Ger1234

    Ger1234 New Member

    Posts:
    4
    Prize Bonds are an integral part of the wider range of NTMA State Savings products offered to personal savers by the National Treasury Management Agency (NTMA). When you save with NTMA State Savings products you are placing your money directly with the Irish Government. State Savings forms part of the sovereign debt of Ireland which is managed by the NTMA. The repayment of all NTMA State Savings money is a direct, unconditional obligation of the Government.
     
  14. Logo

    Logo Frequent Poster

    Posts:
    92
    Bank deposit rates are at an all time low. State saving products are a bit more attractive - but why have state savings been allowed to be more competitive than banks?
     
  15. CiaranT

    CiaranT .

    Posts:
    4,496
    Good question. The NTMA State Savings instant access product is not highly competitive but getting there! The 4 year, 5 year and 10 year products are market leading rates and the 3 year rate is highly competitive. The NTMA say they set rates based on sovereign yields and deposit rates. However, the NTMA do not frequently review their rates.

    Banks have every right to be livid about the NTMA State Savings rates and banks have canvased the government on this before. The 4 year, 5 year and 10 year products are priced above deposit market price and offer a state guarantee.

    Surely it will not be long before the NTMA reviews rates once again?