minerjoe wrote:So i have been unitising my portfolio (I should have done this ages ago!) and I decided as one of the data learnings I would assess all of the cash put into the HYP-ISA vs dividends generated by it over the few years its been running.
The results are very interesting to me and show a large growth in the contribution dividends are making YoY. This will be skewed this year due to the limit going to £20k (i'll try to use the whole thing). But then it got me thinking, how does everyone else's HPY look? Are you at the point where new cash actually makes up less than the dividends reinvested? and at what point do you just keep the cash? ...
GDHYP should give you a useful idea of what can be expected from a constant rate of saving into ISAs, as I've run it on a constant £7,200 per year since its start in April 2008 and without any Income Tax or CGT being taken from it, and it's tracked pretty accurately using a Halifax ShareBuilder account run with real money, at 1/8th of the portfolio value (*).
It reached a milestone in April 2016 when its forecast dividend income passed £3,600 per year. Before making its 37th purchase, the forecast income was £3,567.33 (see
https://web.archive.org/web/20161213055103/http://boards.fool.co.uk/gdhyp-37th-purchase-13358323.aspx?sort=whole). The purchase itself was of £1,719.12 worth of Aberdeen Asset Management shares, on a forecast yield of about 7.2% (
https://web.archive.org/web/20161213055743/http://boards.fool.co.uk/poll-gdhyp-37th-purchase-final-run-off-poll-13360493.aspx?sort=whole). That will have added about £120 to the forecast income and so have pushed it over £3,600.
So that's 8 years to reach the halfway point to your target of dividend income exceeding contributions, producing a crude estimate of 16 years to reach the target. That's a bit pessimistic, though, because as the dividend stream rises, money is being invested at an increasing rate - so if the average yield for the new purchases averages about the same, the income will rise more quickly. A less crude estimate that takes account of that effect (not totally accurately, but it's pointless trying for great accuracy in the face of the vagaries of dividend cuts, etc) is that during those first 8 years, the rate of investment averaged about 1.25 times the initial savings rate, and during the years it then takes to get up to twice the initial savings rate it should average about 1.75 times the initial savings rate, so it should take about another 8*1.25/1.75 years, or about another 6 years, for about 14 years in total. On the other hand, there has been some evidence that the forecasts tend to be a few percent optimistic, and in any case there's no real point in trying to give more than a ballpark round-number figure, so I would probably just settle for "very roughly 15 years".
The assumption that contributions are being made as a constant rate might of course not be realistic, given the existence of inflation and rapidly-increasing ISA allowances. But you can achieve just about any answer you like by changing the contribution schedule to keep contributions ahead of dividends or drop them below dividends... I.e. the question basically doesn't have an answer without some sort of assumption about the contribution schedule!
(*) E.g. the notional ISA version gets £7,200/12 = £600 per month savings put into it, so the Halifax ShareBuilder account gets £600/8 = £75 per month. There are some small adjustments and inaccuracies involved in running the Halifax ShareBuilder account - they are:
* I only actually started running that account around the beginning of July 2008, and I 'kick-started' it with 3 months of contributions. This didn't involve backdating any investment decisions, as those first three months were just spent accumulating enough cash for its first purchase, It did involve a very minor inaccuracy on the interest earned on those first three months' contributions.
* The Halifax ShareBuilder account supports fractional shares, to a granularity of a millionth of a share - which is precise enough to never leave any 'change' from a purchase. A real ISA would buy in whole shares, generally getting 'change' from each purchase and putting it towards the next purchase - so some of its purchases will be slightly smaller than in strict proportion (if they receive less 'change' from the previous purchase than they pass on to the next purchase) and some slightly larger (if they receive more 'change' from the previous purchase than they pass on to the next purchase). This will produce some pretty minor inaccuracies.
* Amounts of cash in the Halifax ShareBuilder account are 1/8th of what the corresponding amounts would be in the ISA, rounded to the nearest penny. That rounding produces very minor inaccuracies in the representation of the ISA - no more than 4p on any particular dividend, etc.
* The Halifax ShareBuilder buying commission was originally £1.50, exactly 1/8th of the £12 commission I was assuming in the ISA. I adjust for this by putting an extra 50p into the Halifax ShareBuilder account for each purchase (and on the one occasion that I've needed to make a sale with commission, I've made a similar adjustment).
* Generally, I just pay the small amounts of tax on GDHYP's dividend income from outside the Halifax ShareBuilder account, so that account is effectively free of tax. The one exception so far is PIDs from the portfolio's one REIT, namely British Land: I can't stop the 20% withholding tax being taken from the account, so instead I adjust for it by putting an extra amount equal to the withholding tax into the account each time British Land pays a PID.
Gengulphus