ISAs should be more aggressive than sipps

bowlhead99

TheTracker wrote: »

SIPP tax efficiency is mostly on contribution. MAs argument is what to do about allocation post-contribution.

Re "The wealth-maximising option is the one that has your highest growth pot be the cheapest to acquire, or put another way, maximising the growth of the free money." is not true post contribution. The wealth maximising option is to maximise growth in the pot with least post-contribution tax. I'm sure you are aware of the mistaken assumption people make about compounding of relief inside a wrapper, it doesn't happen.

If this is all about post contribution allocation (i.e. after getting the tax relief on the way in but before paying some (lower rate of) tax on the way out, then yes I do agree. If you are going to vastly grow just one of two pots, it makes sense to grow the one that doesn't have any tax to pay on it. Otherwise you are giving away tax on the growth when you could instead give away tax on the non-growth, which is relatively less tax leakage to HMRC.

Funny world where I'm siding with MA vs b99 logic!

As usual, I wasn't quite clear on where MA was going to take his thread and what he was getting at, it can be a leap in the dark sometimes And reference to bearing in mind the LTA seemed like it is a pre-contribution decision, notwithstanding the fact that MA's own financial situation is not in danger of breaching annual or lifetime allowances.

As you say, if this is all about post-relief allocation then that's a different question to maximising the relief that you get from end to end. The figures put forward by Linton show the effect of that and come down in favour of taking the tax boost (i.e. SIPP) on the higher growth assets ; my 'overstating the case' as he mentioned was just to get big numbers out there as examples so the result would be less subtle.

In the OP, he suggested 'the SIPPs have *all* the bonds and the ISAs have *all* the equities'. Most people Matthew's age on a 30-40-50 year view to reach mid-retirement would have the vast majority of their assets in equities and so if all the equities went through ISAs and SIPP was only used for the bond component, they would be throwing away valuable tax relief on perhaps four-fifths of their portfolio, which doesn't seem like a smart thing to do.

However, coming up with random brainwaves about using an unconventional investment strategy that most seasoned investors would shun, and then defending that brainwave to death on the ensuing thread, is the sort of thing that Matthew does all the time, so I just assumed it was going to be one of those threads

With the further clarifications, seems like you've correctly surmised that he is only looking at the post-relief allocation once as much pension tax relief at source has already been bagged. In which case it makes sense to hold the bonds in the SIPP along with equities where they will get taxed in the future, and then have the ISA be equities only (if it's reasonably certain that no access would ever be needed on the volatile ISA pot). Though he does mention in the OP that the ISA does need bonds in the time before SIPP can be accessed, implying he is thinking of SIPP as the longer term solution and the ISA is a shorter term solution ; as such, the thread title 'ISAs should be more aggressive than SIPPs' does not really compute, in that context.

In the past, MA has argued against all comers that his solution of having his SIPP invested exclusively in a global developed world small companies equities tracker is the best thing to do and keep his other, multi-asset investing in the ISA, so this is a complete U-turn. And we know from other threads that he is currently getting a very good rate of SIPP relief due to being on tax credits while there's not currently much danger of him being high rate taxpayer in retirement. So, a switch to suddenly favouring the equities investing being outside a SIPP and having his relatively few bonds all being done inside the SIPP does not make a lot of sense. Probably the safest option when I see a badly explained opening post from MA in future is just not to get involved

Thrugelmir

MatthewAinsworth wrote: »

Fast growth could put one into the 40% income tax band or over the lifetime allowance...

For the average investor this is their ultimate dream. When you find the Holy Grail let us know.

System

Bowl -

so this is a complete U-turn

I do that a lot, I'm open to changing my mind dramatically, I used to be very risk adverse, and still am in non financial aspects, I used to think I'd save up to buy a house without buying a flat first, I used to think I'd never ride a motorbike.

As for "defending to the death" - I just like to fully settle all issues in my mind before I consider complete U turns. Sometimes I have trouble concisely explaining my line of thought, as you see

Don't dismiss me breaching the LTA, it's possible if 12% continues over the next several decades, I can afford to be completely aggressive with tax credits and a dB pension, it's only a wedding and nursery fees and clearing a stooze currently slowing me down, after that I can get serious

Thrug- just be really aggressive, if you have more than 5 years you probably can be

grey_gym_sock

Linton wrote: »

I am nervous saying this but I think Bowlhead's post greatly overstates the case....

IMHO, it doesn't overstate, it's wrong - see below ...

Assuming basic rate tax throughout....

If you had £200K net to invest, £100K in an ISA and £100K in a SIPP then you would end up with £100K in the ISA and £125K in the SIPP.

bzzzt! ... the problem is that you don't have equal values in the ISA and the SIPP at the start. the value of the SIPP is not what it cost you in contributions, but what you'll get out after paying any tax.

your assumption is that the effective tax rate on exit is 15% (because you're assuming 20% tax is paid on the 75% that is tax-free, and 20% of 75% is 15%). therefore the current value of £125K in a SIPP is 0.85 * £125K = £106.25K.

the SIPP (£106.25K) is effectively bigger than the ISA (£100K). so when you put equities in the SIPP + bonds in the ISA, you have a higher equity allocation than when you put bonds in the SIPP + equities in the ISA. and a higher equity allocation gives you higher expected returns.

Suppose the choice was either equity with a total of 100% return or bonds with a 0% return at retirement after a number of years.

1) Bonds in ISA, Equity in SIPP

ISA income =£100K
SIPP income=0.25*£250K+0.75*0.8*£250K=£212.5K

Total=£312.5K

here you start with an equity allocation of £106.25K / £206.25K = 51.52%

and you end with an equity allocation of £212.5K / £312.5K = 68%

2) Equity in ISA, Bonds in SIPP

ISA income=£200K
SIPP Income=0.25*£125K+0.75*0.8*£125K=£106.25K

Total=£306.25K

and here you start with an equity allocation of £100K / £206.25K = 48.48%

and you end with an equity allocation of £200K / £306.25K = 65.31%

case (2) has lower returns than case (1) because it uses a lower equity allocation throughout. (in both cases, the equity allocation drifts up over time, because we aren't rebalancing. this doesn't affect the comparison between the 2 cases, though.)

let's confirm by trying an example in which the ISA and SIPP are genuinely equal in size at the start ...

suppose the SIPP contains £200K, and (as before) we expect to pay 15% tax on that, so it's worth a net £170K. so let's suppose the ISA also contains £170K. and we still assume that bonds make no return, equities double in value.

A) Bonds in ISA, Equity in SIPP

ISA income =£170K
SIPP income=0.25*£400K+0.75*0.8*£400K=£340K

Total=£510K

Equity in ISA, Bonds in SIPP

ISA income=£340K
SIPP Income=0.25*£200K+0.75*0.8*£200K=£170K

Total=£510K

so that gives the same total either way, as predicted.

but ..........

what about matthew's original point, that we might breach tax brackets?

for instance, if your SIPP will be right on the cusp, where any growth will be taxed with higher-rate income tax (allowing for the 25% tax-free part, that's effectively a rate of only 30%), but any losses will be softened only by a saving of only basic-rate tax on the money you are no longer going to draw out of your SIPP because you've just lost it (effectively, 15% tax). in that case, you keep 70% of any further gains you make, but lose out to the value of 85% of any losses. so avoiding losses is more important than making gains.

there are similar issues if you end up near the LTA.

however, that is a bit specialized. most of us will have pension pots that land us in the middle of some tax bracket (though it won't be the same bracket for all of us), with no plausible chance of that changing. so does this matter for most of us?

well, that depends where we end up.

if you end up as a non-income-tax-payer in retirement, then it doesn't matter how you place equities and bonds in ISA vs SIPP: the SIPP is worth 100% of the investments in it (0% effective tax rate, instead of 15%). (however, you still might as well prefer to put any bonds you want to hold in the SIPP (and only put equities in it if it's too big to hold just your desired allocation of bonds), because if the SIPP grows too much then getting it out within the personal allowance could become a bottle neck.)

but in any other case - i.e. some income tax paid in retirement - it does matter how you place equities and bonds in ISA vs SIPP. ....... why so? .........

well, consider the common case, of being a basic-rate taxpayer in retirement, paying an effective 15% on what you will take out of a SIPP. that 15% is your marginal rate of tax, but you probably pay 0% of some of what you take out of the SIPP, because your state pension is likely to be less than your personal allowance. (but note: this could be different if you also have DB pensions, or have other kinds of taxable income when you're retired.)

let's estimate that the first £100K in a pension will come out with no tax due (perhaps after taking the £25K tax-free lump-sum, you drawdown 4% of the remaining £75K per year; that's £3K a year, and perhaps your state pension is £8K and personal allowance is £11K).

so the first £100K has 0% effective tax rate, and everything over that has 15% effective tax rate. (and we're confident that you won't have enough to reach any higher rate beyond that.)

let's say you have £200K in a SIPP now. (so clearly you're into the 15% effective marginal tax rate.) the current value of the SIPP is £100K + 0.85 * £100K = £185K.

to make your SIPP and ISA equal in size (initially), let's say you have £185K in an ISA.

start the fans! .........

I) Bonds in ISA, Equity in SIPP

ISA income =£185K
SIPP income=£100K+0.25*£300K+0.75*0.8*£300K=£355K

Total=£540K

II) Equity in ISA, Bonds in SIPP

ISA income=£370K
SIPP Income=£100K+0.25*£100K+0.75*0.8*£100K=£185K

Total=£555K

so equity in the ISA wins! matthew was right after all!

what is going on here?! ... the £200K SIPP has a tax liability of £15K, giving a blended tax rate of 7.5%; if you put equities in the SIPP, and it grows to £400K, it then has a tax liability of £45K, giving a blended tax rate of 11.25% ... growing the SIPP (when you have the choice of growing the ISA instead - none of this applies in case your ISA is 100% equities and you want more equities: in that case, do put them in the SIPP!) is sub-optimal because it increases the effective blended tax rate on the SIPP, even when it doesn't increase the effective marginal tax rate.

note: i haven't considered inheritance tax. pensions can reduce your estate's IHT bill. but providing for your own retirement should be a higher priority than reducing IHT. and avoiding IHT is IMHO somewhere between pointless and anti-social, anyway. so i won't attempt to cover that.

(thanks to all contributors in this thread - i didn't know some of this before i started reading it )

bowlhead99

MatthewAinsworth wrote: »

Don't dismiss me breaching the LTA, it's possible if 12% continues over the next several decades, I can afford to be completely aggressive with tax credits and a dB pension, it's only a wedding and nursery fees and clearing a stooze currently slowing me down, after that I can get serious

As mentioned on other threads, 12% over several decades isn't realistic; you can't just extrapolate what you saw on your global smallcap fund from the bottom of a market trough to a market peak during a period when sterling weakened 20%, and presume that will continue as a long term average over three or four decades.

The only way you will get 12% on an ongoing basis in nominal terms is if inflation is rampant. And if it is, you would expect the LTA to be increased. At the moment, a million pounds at 4% drawdown gives you £40k spending money each year. However, after 4 decades of 3.5% inflation that would only be £10k in real terms. And with the sort of higher inflation that would allow 12% returns from mainstream investments, it would be even less, so there is a general assumption that the LTA will be increasing over time.

System

12% is I believe a long term historical average for small caps. From bottom of the trough it's not 12%, it's 15+%, which I'm not expecting.
It is an assumption of course but isnt a long term historical average a reasonable assumption for a broad global sector?

Malthusian

Not for small caps. It is difficult / inadvisable to invest in small caps using an index-tracking approach (because of the illiquidity of some small caps and the amount of complete junk that makes up the lower end of something like the FTSE AIM) so the actual experience of most investors in small caps can be expected to diverge heavily from the index, as it depends on how many unicorns they or their fund manager pick and how many duds that go bust. This is not the case when you are talking about a relatively stable index like the FTSE All Share.

12%pa with a very aggressive portfolio may well be possible (in a statistical sense), and since we are talking about the Lifetime Allowance it may even make sense to at least consider the possibility of such high growth. I certainly wouldn't base my tax planning decisions on such an assumption, however - i.e. I wouldn't use a less tax-efficient solution because of the fear that if I get 12% per annum growth I may end up paying a lifetime allowance charge. That comes under the heading of "extremely nice problems to have".

System

Malthusian - illiquidity would genuinely be a problem for a tiny individual investor? I could just exit gradually when I do at the end and that way I don't need to divert from the index? Also being in an income version might help somewhat

To me the relatively high p/e of 18 of my fund suggests I should find a buyer?

bowlhead99

MatthewAinsworth wrote: »

It is an assumption of course but isnt a long term historical average a reasonable assumption for a broad global sector?

Well, you have to start somewhere. I'm assuming you have plucked the popular 12% figure from the "US Smallcaps" index performance info in the annual Ibbotson SBBI report which goes back to the 1920s. For each of the last few years it has been reporting something like 11.9 to 12.1% as the annualised return for that asset class.

However, there are a few flaws with seizing that as an assumption on a go-forward basis.

- it's based on US returns in dollars.

- US is pretty much the most capitalist place you can find and US market returns for a lot of the 20th century were in line with market conditions of what we might see these days as "emerging markets". US was still developing its markets from the '20s to the end of the century. The market really developed properly and boomed for multi decade periods, asserting its global dominance. Now it is already there, it's developed, waiting to hand over power to the next generation or two of other competing economies such as China or India with their larger populations and cheaper workforce.

-With a capitalist society and encouraging entrepreneurship, US has been a good place for smallcaps and some have really thrived, on average outperforming largecaps by 2% or more over those long timescales. However, these days it is perhaps a little different. When Apple or Microsoft were a smallcap, they could thrive. They carved new niches. These days if you're a player in the tech space you may get squeezed out or gobbled up by those guys or Facebook or Google or Tencent, Alibaba etc before you have any time to make your investors money as part of a smallcap index. If you look at other countries like Mexico or Malaysia, largecaps were ahead for the first fifteen years of the current century (I haven't got more recent data) as they have large oligopolies which can crowd out smaller companies. These days the premium returns from US smallcap are not a massive amount more than largecap or midcap, so what worked in 1950 doesn't necessarily work today.

- And the global financial markets support really big IPOs these days; Facebook, Snapchat etc just IPOd at midcap /largecap level instead and if Uber were to IPO it would be in the $50bn+ range. The concept of institutional private equity funds and VC houses and tech incubators didn't really exist for the first half of the timescale from where the "average 12%" returns are being taken. But now a company can stay in private hands or have a succession of off-market owners, before IPO at 20 billion plus, and go up in value five or ten or twenty-fold without the smallcap fund getting a sniff.

- So, those are just a few reasons why basing your faith in a single country experience of a single asset class could give you fundamental misplaced confidence in the future returns because you are looking back at the last 90 years and not forward to the next 40 years.

FWIW, I don't have any issue with saying smallcap should return slightly better than the slow moving cash cow behemoths over the next few decades. But if the returns are "largecap plus 2%" you will only get to 12% for smallcap if largecap is 10%, which is aggressively ambitious from here. You will struggle to find someone projecting that for the global developed markets. And most do not try to build a 100% smallcap pension portfolio because it's a wild ride with no guarantee of the best result.

Some of the niches that might do better than generalist smallcap include microcap and emerging markets but as we have discussed on other thread, you don't want them because you perceive them to be risky or unnecessary. So, your unwavering faith for smallcap index fund all the way, is not something I endorse, but we've been here before in other threads.

MatthewAinsworth wrote: »

Malthusian - illiquidity would genuinely be a problem for a tiny individual investor? I could just exit gradually when I do at the end and that way I don't need to divert from the index? Also being in an income version might help somewhat

The "illiquidity as a problem" refers to having large investment funds with billions under management all trying to deploy money into small companies with limited capital in free float on the stock market. There may not be sufficient liquidity for them to buy or sell the amount they want/need efficiently. The will be a bid-offer spread when acquiring and disposing of shares in the companies and in a down market they might have to take a massive loss to dispose of the assets, as the action of a large seller trying to exit moves the price down while they're trying to do it. And they might be a settle even if they like the investee companies, just because they have investors that want their money back and they have no choice but to dump stock to get their hands on the necessary cash.

So, it's not saying that you personally will suffer illiquidity issues when wanting to get money out of a large fund, but that the fund itself suffers losses when you and others want to exit. This may or may not translate to the fund freezing redemptions or passing the costs to the leaving investors but as an investor generally you will at least indirectly suffer from it. Also the fund has to pay high costs to deploy money when you want to buy in to these illiquid companies. Some of these things are expensive or unhealthy for your return potential. The 12% return in the studies was with no deductions for real world costs and fees. Or taxes for that matter, given that you are not buying individual company shares through your pension, but instead buying a fund which may suffer unavoidable foreign withholding taxes on its divi income from its investees.

To me the relatively high p/e of 18 of my fund suggests I should find a buyer?

Without commenting on that specific exact number and whether it's high or not: a fund you own having a high P/E means that you and some others were willing to pay a high ratio of "buy price to earnings of the companies" to buy in today -or at least to hold on to them today - rather than selling.

But that in itself doesn't suggest you'll find a buyer in the future at that ratio at any time in the next few decades. Maybe in 2047 when you want to retire, people will only be willing to pay 12x earnings for a broad smallcap fund, which would be a third less than they are paying now -so a large amount of the profit growth of the constituent holdings would be offset by market sentiment for smallcap vs othercaps.

System

On the back of the last 90 years being different I'm willing to switch my small forecast to large+2%, it seems reasonable and more conservative

It doesn't really matter whether I hit LTA, might do, who knows, might not

Would currency really matter to average growth over 90 years?

There may be liquidity problems faced by fund, but if at the moment tracking error is negligible can I assume it's ok at the moment?

Fully accept that current p/e isn't a guide to future, but assumption is all I have

I have been revisiting emerging but it doesn't seem reliable enough, and maybe tracking the s&p 500 is a better way overall of doing that