One share per sector, or two?

[pyad]

[Mod. - quoting Stephen Bland] The total is then split equally between them so as not overweight the sector at cost.

That's interesting. Why "at cost" and not "at value" if building over many years?

I assume that he refers to the initial cost of the multiple holding within said sector being equal to the cost of single holdings in other sectors so as to retain balance. Thereafter is another matter as share price movements will tweak the balance. But that's what I think, not what I know.

The amount to invest in a new sector added to an existing portfolio is always the current average value so as not over or under weight it according to the present structure of the port. It is the present state of the port that matters. If you are adding a sector having two or more new shares then you split the sum equally between them.

Similar thinking applies to topping-up existing sectors, where the maximum sum to add is that required to bring it up to current average value but not more.

Original cost applies only to a new portfolio being set up but for a portfolio already in existence, it becomes wholly irrelevant for the purpose of calculating how much to add to create a new sector or top-up existing sectors.

I worded my previous comment badly so apologies for that. I should have said original average cost or current average value for new or existing ports respectively.

[Gengulphus]

I can say from personal experience that with very large ports that started off much smaller and grew substantially over time, what happens is that you increase your individual sector or share comfort zone value, rather than increase the number of holdings just so as to reduce the average holding size to the earlier CZ level. There may be some increase in the number of sectors or shares but not in proportion to the growth of port value.

Thus if at some earlier stage you were comfortable with an average £10,000 per sector or share and the port later doubles in value you don't suddenly double the number of holdings. What happens is that your CZ doubles to £20,000 per sector or share and so on. Well that's how things panned out for me.

Just to say that my personal experience was similar, apart from the fact that when my total wealth increased by a huge amount very quickly during the tech boom, there was a definite 'lag': it took considerably longer for the holding size I felt comfortable with to rise in accordance than it did for the holding size I should have been comfortable with. Not by any means an issue that every HYPer will face, though other things besides stockmarket bubbles can cause a big and rapid wealth increase, such as a Lottery jackpot win. I would guess the most likely one is a big inheritance - but the chance of a HYPer having a relative wealthy enough in comparison to them to cause such an increase still won't be all that high.

But mainly, the point of what I said about pushing up the number of holdings is a reductio ad absurdum argument to show that fixing on an absolute holding size limit is an unsustainable position to take, unless it is chosen to allow for all the future portfolio growth that is likely to occur. Unless only a reasonably small amount of future portfolio growth is likely to occur, such a big absolute limit will do little (if anything) to keep the portfolio diversified now, so one either needs something else as well to do that job or not to fix the absolute limit but instead review and update it regularly. But that's unnecessary complexity, as a holding size limit set relative to the portfolio size rather than absolutely will do the job both now and in the future, without any reviewing needed.

Such a relative holding size limit adjusts to portfolio size in the way you describe.

Gengulphus

[moorfield]

The amount to invest in a new sector added to an existing portfolio is always the current average value so as not over or under weight it according to the present structure of the port. It is the present state of the port that matters. If you are adding a sector having two or more new shares then you split the sum equally between them.

Similar thinking applies to topping-up existing sectors, where the maximum sum to add is that required to bring it up to current average value but not more.

Original cost applies only to a new portfolio being set up but for a portfolio already in existence, it becomes wholly irrelevant for the purpose of calculating how much to add to create a new sector or top-up existing sectors.

I worded my previous comment badly so apologies for that. I should have said original average cost or current average value for new or existing ports respectively.

pyad, thanks again for clarifying. My own build method differs here in that I currently work with an upper % limit rather than an average value for sectors. Unsurprisingly, I'm beginning to see the concentrating effects of that already - my 5 largest sectors of 16 / 9 shares of 20 account for ~47% of portfolio value - so I'm going to take on board trying average value instead next year.