Deep dive on ISK vs IT/CGT

[Duke of Marmalade]

I attach work in progress on this (usual caveats therefore).
The below exhibits show the superiority/inferiority of a possible 1% ISK over IT/CGT assuming current rates and marginal income tax rate of 50% (all adjustable within the spreadsheet)
ISK seems superior in most circumstances except in some negative capital gains scenarios. Of course negative capital gains every year for 40 years is hardly realistic and a proper analysis would require stochastic methods.
I show the situation where dividends are 0% or alternatively are not taxed as income - unrealistic IMHO.

 

[Brendan Burgess]

Can I make sure I understand this

After 4 years, if the average growth is 6% the ISK will return €88 more than CGT/IT?

Brendan

 

[Brendan Burgess]

I am doing something wrong

Despite dividends and capital growth, after 20 years, I am losing money on the RHS

The LHS is correct though

Ah, the figures in red are the differences. Not the absolute outcome.

 

[Brendan Burgess]

So you leave out the tax-free bit of the ISK which is correct.

Because, whichever one you opt for, you should always have the tax-free amount in the ISK.

And what this table shows us is how the balance over the ISK tax-free amount should be invested

 

[Duke of Marmalade]

After 4 years, if the average growth is 6% the ISK will return €88 more than CGT/IT?

Yep

 

[Brendan Burgess]

How have you managed the Deemed Disposal?

I thought it would be a lot more complex as you would have to pay the DD tax every 8 years?

 

[Brendan Burgess]

What is the table I have highlighted in green?

 

[Sarenco]

I don’t think it’s very helpful to use average returns to model the attractiveness of the ISK model v CGT ( or exit tax).

If you use a real world sequence of returns (or a Montecarlo simulation), the ISK gives a very unfavourable result if you have negative returns early on.

 

[Duke of Marmalade]

How have you managed the Deemed Disposal?

DD only works for 8 year duration.

 

[Duke of Marmalade]

What is the table I have highlighted in green?

That is the IT/CGT situation stand alone whereas the table above it is the difference between ISK and the IT/CGT situation.
If you look at the contents of Cell N16 by clicking on it you will see =D24. That means the table shows the contents of D24 in all the different combinations. N16 can simply be changed to =D15 to get a different table or even to =D15-D24

 

[Duke of Marmalade]

If you use a real world sequence of returns (or a Montecarlo simulation

Ok, I'm on the case.

 

[Brendan Burgess]

DD only works for 8 year duration.

So should it not be omitted from the tables?

 

[Duke of Marmalade]

So should it not be omitted from the tables?

it should but i am not posting any of it on the thread; I will try and get round to it

 

[Brendan Burgess]

I think it's clear that CGT/IT and the ISK dominate Deemed Disposal so it does not need to be included.

 

[Duke of Marmalade]

If you use a real world sequence of returns (or a Montecarlo simulation)

Here is my first cut at this. I used a 12% volatility and simulated 1,000 40 year periods.

The worst cases are generally negative for ISK vs IT/CGT though never by very much. The median and best cases are much better with ISK. The 50% tax on dividends is neutralising the 1% ISK tax so no surprise that ISK comes out on top. As a reality check I ran the same simulations with 0% dividends and 6% capital growth.

I puzzled over the -8% at 4 years; surely -4% is worst but the model is giving credit for CGT losses.
I attach the spreadsheet. You can play with the inputs but note that the response times are slightly slower than usual because of the heavy computations. 0% volatility gives an error so if you want the deterministic results key in a volatility of, say, .00001%

 

[Brendan Burgess]

So in the normal portfolio of 2% dividend and 4% capital growth, one should go for the ISK. You stand to gain a lot with a low risk of losing a little.

But in the somewhat artificial case of 0% dividends, the IT drag on the CGT/IT model is eliminated and so it often performs better than the ISK system.

Why is the ISK's advantage peak at 28 years and then decline before turning negative?

 

[Duke of Marmalade]

Why is the ISK's advantage peak at 28 years and then decline before turning negative?

That does seem wrong. I will have to think about it. Emphasises the caveat about this being work in progress.

 

[Duke of Marmalade]

Why is the ISK's advantage peak at 28 years and then decline before turning negative?

Not sure I have the full answer.
There are two trends.
The first trend is that ISK tax starts better than CGT at 1% vs 33% but this reduces steadily at 1% each year eventually turning negative.
The second trend is that the results are expressed as a % of the initial investment. Certainly for the median and 5% best the absolute amount of the gain will be growing so whilst ISK's inherent advantage is falling the growth in the actual gain affected is dominating that inherent decline.
There is a further effect in that by looking at the 5% best you are picking out those instances where sequence of return favours ISK and it seems better than CGT even beyond the natural point where 1% p.a. exceeds 33%. The same distortion is not present for the median.
I am sure that all seems like gobbledygook.

 

[Wolframius]

But in the somewhat artificial case of 0% dividends, the IT drag on the CGT/IT model is eliminated and so it often performs better than the ISK system.

Why artificial? You don't expect that low-cost accumulating index fund can be purchased on the future Irish ISK?

 

[Duke of Marmalade]

You don't expect that low-cost accumulating index fund can be purchased on the future Irish ISK?

I expect they will be in the majority. But Simon has stated that there will be steps against the anti avoidance implicit in accumulator funds, Possibly a look through to dividends reinvested. A big advantage of ISK is that taxation is without these complications.