Pensions already exceed LTA - How do BCEs and Taxation apply

zagfles

rayzjay said:

jamesd said:

To avoid, take DB now as income and start crystallising 80k a year, the tax free bit going into an ISA, taxable left in flexi-access drawdown until later. 

Is there some significance to the amount of £80k a year or is that just an example?

Presumably because it releases £20k TFLS which can go into an ISA (the annual ISA limit).

But I think jamesd's market timing strategy is flawed, as above. If you expect investments to grow, then full crystallisation would usually be the best idea, to avoid the LTA charge increasing, which would usually be more than the tax on the inevitable unwrapped assets from the PCLS, although would depend on other income. Also inheritance issues etc.

Basically you need to either research all aspects in depth or get advice. In your situation paying for one off tax advice would probably be an idea, doesn't necessarily need to be independant, your SIPP provider probably offer it.  

Dead_keen

rayzjay said:

zagfles said:

As the PTM says, it's growth minus withdrawls, ie it simple subtracts the value at 75 from the value when crystallised under BCE1 and if it's greater you'll pay 25% tax on the difference as a one-off, usually from the SIPP. Plus of course income tax as usual on drawdown.

So does that mean that I can avoid any additional tax on the growth (other than income tax) providing I take out the amount of any growth as pension payments prior to age 75?

Yes, that is right.

Albermarle

zagfles said:

rayzjay said:

jamesd said:

To avoid, take DB now as income and start crystallising 80k a year, the tax free bit going into an ISA, taxable left in flexi-access drawdown until later. 

Is there some significance to the amount of £80k a year or is that just an example?

Presumably because it releases £20k TFLS which can go into an ISA (the annual ISA limit).

But I think jamesd's market timing strategy is flawed, as above. If you expect investments to grow, then full crystallisation would usually be the best idea, to avoid the LTA charge increasing, which would usually be more than the tax on the inevitable unwrapped assets from the PCLS, although would depend on other income. Also inheritance issues etc.

Basically you need to either research all aspects in depth or get advice. In your situation paying for one off tax advice would probably be an idea, doesn't necessarily need to be independant, your SIPP provider probably offer it.  

By coincidence I was looking today at my SIPP providers financial/retirement  advice charges ( Fidelity).
They charge £4K for one off advice relating to LTA strategy only . 

jamesd

zagfles said:

jamesd said:

... Continue with this while waiting for a market drop. Say 200k drop required on 900k is 23% and that's moderately frequent. Growth and the gradual crystallising makes it harder each year but the prospects for getting it all with 25% tax free and without any lifetime allowance charge are at least reasonable. If you see a nice correction maybe use that along the way.

Have you actually backtested this sort of strategy you keep mentioning? A drop of 23% might be quite common. For the market to be 23% lower in the future than at this point in time isn't. A dip of 23% after growth of 30% is still growth.

Look at any stockmarket chart. Then backtest - pick dates and see if the stockmarket is ever 23% lower in the future than on that date.

Looking at a random one, eg Dow Jones, from 1982 to 1998 there is no date on which the stockmarket was ever 23% lower on any future date, except for a few months in 1987. Same for 2009 to about 2017. So for most dates between 1982 and now, the Dow Jones was never 23% lower at any point in the future than on that date. And that doesn't account for dividends.

Of course and I'll illustrate three and a half recent years below. Part of the purpose of crystallising gradually is to get the crystallising done if it does turn out to be the case that there is no future lower opportunity.

Let's look at the FTSE100, data at https://m.uk.investing.com/indices/uk-100-historical-data and I set it to monthly values from March 2009, that being around the last long term bottom and the start of a long mostly bull market. Look for runs of red to see time periods to check for possible opportunities.

In March 2020 it shows 5671 and in October 5577 with intervening months only a little higher.
 5577 is lower than every month value since June 2012 when it was 5571. 7242 would be a level where a 23% drop would go to 5577 and you can see that from Feb 2017 to September 2018 and again from March 2019 to January 2020 the index was at least that high in the monthly values. The daily low was 4994 on 23 March 2020 but it was 5237 on the 12th so there was time to act. 6801 is where 5237 is 23% down and from September 2016 the index was normally higher than that.

In July 2018 the index was 7748. 6728 in December 2018 with daily low 6584, a 15% drop to potentially exploit along the way because the market was higher since June 2016. Of course not 15% down for most of that time but still exploitable to reduce the bill.

There are no guarantees that a suitable drop will come along. At present the US cyclically adjusted price/earnings ratios implies a high chance of a large drop but it still may not happen in time to matter.

jamesd

rayzjay said:

jamesd said:

To avoid, take DB now as income and start crystallising 80k a year, the tax free bit going into an ISA, taxable left in flexi-access drawdown until later. 

Is there some significance to the amount of £80k a year or is that just an example?

Chosen to get 20k tax free for reinvesting inside an ISA each year.

How much you should crystallise at the start depends on your assessment of the prospect of a suitable drop in the future. The more you crystallise initially, the larger the drop needs to be to eliminate the whole bill because the same Pound decrease has to be made from a smaller amount not crystallised.

If you want to maximise your chances you'd have high equities uncrystallised and low equities crystallised.

There's no guarantee that a drop will come along.

zagfles

jamesd said:

zagfles said:

jamesd said:

... Continue with this while waiting for a market drop. Say 200k drop required on 900k is 23% and that's moderately frequent. Growth and the gradual crystallising makes it harder each year but the prospects for getting it all with 25% tax free and without any lifetime allowance charge are at least reasonable. If you see a nice correction maybe use that along the way.

Have you actually backtested this sort of strategy you keep mentioning? A drop of 23% might be quite common. For the market to be 23% lower in the future than at this point in time isn't. A dip of 23% after growth of 30% is still growth.

Look at any stockmarket chart. Then backtest - pick dates and see if the stockmarket is ever 23% lower in the future than on that date.

Looking at a random one, eg Dow Jones, from 1982 to 1998 there is no date on which the stockmarket was ever 23% lower on any future date, except for a few months in 1987. Same for 2009 to about 2017. So for most dates between 1982 and now, the Dow Jones was never 23% lower at any point in the future than on that date. And that doesn't account for dividends.

Of course and I'll illustrate three and a half recent years below. Part of the purpose of crystallising gradually is to get the crystallising done if it does turn out to be the case that there is no future lower opportunity.

Let's look at the FTSE100, data at https://m.uk.investing.com/indices/uk-100-historical-data and I set it to monthly values from March 2009, that being around the last long term bottom and the start of a long mostly bull market. Look for runs of red to see time periods to check for possible opportunities.

In March 2020 it shows 5671 and in October 5577 with intervening months only a little higher.
 5577 is lower than every month value since June 2012 when it was 5571. 7242 would be a level where a 23% drop would go to 5577 and you can see that from Feb 2017 to September 2018 and again from March 2019 to January 2020 the index was at least that high in the monthly values. The daily low was 4994 on 23 March 2020 but it was 5237 on the 12th so there was time to act. 6801 is where 5237 is 23% down and from September 2016 the index was normally higher than that.

In July 2018 the index was 7748. 6728 in December 2018 with daily low 6584, a 15% drop to potentially exploit along the way because the market was higher since June 2016. Of course not 15% down for most of that time but still exploitable to reduce the bill.

There are no guarantees that a suitable drop will come along. At present the US cyclically adjusted price/earnings ratios implies a high chance of a large drop but it still may not happen in time to matter.

That a good data source - easy to copy and paste into a spreadsheet and do some calculations on.

Right - so I've used monthly data as far back as it goes, a bit over 20 years to Feb 2001.

Column to show price col * (1-0.23)

Another col to give min future value "=MIN(B$2:B247)" filled up from bottom.

Another to indicate if min future value is the higher.

Results are, for the 246 months since Feb 2001, 162 of then there was never any point in the future where the FTSE was 23% or more lower. 82 of them there was.

So your strategy of waiting for a 23% fall fails two thirds of the time, based on FTSE since 2001.

Now try with the Dow. That goes back further, to Feb 1985.

Results are 328 months there was never any point in the future where the index was 23% or more lower, 110 there was.

So your strategy fails three quarters of the time with the Dow.

If you add in dividends, it'll be even worse.

There's no point disingenuously pointing out 15% falls. You said "but the prospects for getting it all with 25% tax free and without any lifetime allowance charge are at least reasonable. If you see a nice correction maybe use that along the way"

So if you see a 15% fall, that wouldn't do to get it out "without any LTA charge", you'd need wait till it fell further to achieve your goal. Except, based on above, it probably wouldn't. And the result would likely be never getting there, and as a result paying far more LTA than if crystallising fully now.

jamesd

Thanks for doing that work.

That looks like an "at least reasonable" prospect to me. In addition there's the chance of getting it out with a lower charge rather than none. And of course I provided for the no drop case as well with regular crystallising.

Starting time matters and it'd be a bad idea to do it just after a major and unrecovered market drop but that's not the situation those making the decision today are in. For them, they are starting close to or at long term market highs.  That improves the odds but it's still a case of probabilities rather than certainty, as with so much of DC retirement planning.

It's interesting to look at the probabilities and perhaps also see how they vary with cyclically adjusted price/earnings ratios to see whether that can help to refine when things are particularly favourable or unfavourable.

One other important factor to consider is the investment mix. Bonds can be expected to make it increasingly hard as their percentage increases and at the levels needed here it's very much a case of extracting bonds first to increase the equity percentage. As you say, dividends also reduce the success rate.

zagfles

jamesd said:

Thanks for doing that work.

That looks like an "at least reasonable" prospect to me. In addition there's the chance of getting it out with a lower charge rather than none.

How? That would mean you'd need to call the bottom. If you're capable of calling the bottom of any dip (or the top of any peak), then you don't need to worry about trivialities like the LTA, you're admiring the view of your private island from your new luxury yacht.

For a mortal without the ability to see the future, the only way is to set a target. Above you set the target of a 23% drop to avoid the LTA. So a 15% dip would be no use, you'd wait for it to drop further to 23%.

And of course I provided for the no drop case as well with regular crystallising.

Which also means in a rising market, the 23% drop required to avoid the LTA would be increasing all the time as you use a greater % of LTA each time

Starting time matters and it'd be a bad idea to do it just after a major and unrecovered market drop but that's not the situation those making the decision today are in. For them, they are starting close to or at long term market highs.  That improves the odds but it's still a case of probabilities rather than certainty, as with so much of DC retirement planning.

It's interesting to look at the probabilities and perhaps also see how they vary with cyclically adjusted price/earnings ratios to see whether that can help to refine when things are particularly favourable or unfavourable.

One other important factor to consider is the investment mix. Bonds can be expected to make it increasingly hard as their percentage increases and at the levels needed here it's very much a case of extracting bonds first to increase the equity percentage.

So more market timing talk. The stockmarket tends to rise over the long term, therefore you'd expect it to spend a lot of time at a "long time market high".

You don't seem to get the point that trying to time the market wrt LTA taxation is no different in principle to trying to time the market for anything else, like buying/selling. The gear is different, ie it's tax rather than the whole amount, but it's the exact same principle.

If you think the market will be lower in the future, then sell equities now and rebuy then. If you think the market will be lower in the future, then wait before crystallising to reduce LTA change. Exactly the same principle, different scale.

As you say, dividends also reduce the success rate.

Indeed it does, massively. I've amended to use an adjusted index with an assumed 2% dividend return after charges. The results are now (in terms of successes in waiting for a 23% fall):

FTSE: 3 success, 243 fails

DOW: 16 success, 422 fails

So under 4% chance of ever seeing the stockmarket 23% lower inc an assumed 2% net dividend return after charges.

If we change the required drop to 15%, results are:

FTSE: 21 successes, 225 fails

DOW: 31 successes, 407 fails

Above used monthly figures from the last day of the month. Now changed it to use the monthly low for measuring the drop, but still the end of month figure for the start date. That might improve things, if you're prepared to monitor daily:

For 23% drop:

FTSE: 25 successes, 221 fails

DOW: 28 successes, 410 fails

For 15% drop:

FTSE: 51 successes, 195 fails

DOW: 52 successes, 386 fails

jamesd

Have you never looked at the excellent inverse correlation between cyclically adjusted prjce/earnings ratios of markets and their ten year returns? It's very high and much of what I write below is informed by it so if you haven't, let me know and I'll provide some links.

Reading Interpreting Monte Carlo Analyses and the Wrong Side of Maybe Fallacy and the more generic The Wrong-Side-of-Maybe Fallacy would also be useful because your arguments at their heart seem to largely largely rely on that fallacy. I'm entirely happy to seek to exploit possible but uncertain outcomes, should they happen, and to explain their potential to others.


You're partly mixing up market timing and market trends, while also ignoring correlation between market cyclically adjusted price/earnings ratios and future performance. That correlation matters because it's what tells us how good or bad the prospects of a substantial drop are, as well as when drawdown rates are more or less likely to be at their lowest safe level. It's also why you're unlikely to see me suggesting this approach after substantial market drops, when conditions are less favourable than at the moment.

For a mortal to set a 23% target and ignore less would be to not do as I suggested and also exploit dips. It's not a target, it's sufficient to fully eliminate rather than just reduce the charge. Reducing is also a good outcome because we're comparing two alternative plans, one of which doesn't reduce it, the other which may. We don't know whether a sufficiently great outcome will come along, so take some benefit from whatever does.

My intended reading of "long term market high" is "setting new market highs or being at relatively high levels for a long time". Bull markets of record breaking duration happen but they also end.

Naturally and obviously I think that markets will be higher in the future. That doesn't prevent me from planning for the drops along the way. In another guise that's called sequence of return risk and it matters a lot in drawdown.

You're using absolute win everything or no benefit analysis in your statistics and also ignoring market conditions at the starting point. I'll be entirely happy to save 10% of the lifetime allowance charge bill and even 1% is an improvement.

For starting conditions you might look at what happens if you only follow the approach for markets which are first in the higher 50% of their cyclically adjusted price/earnings range, then starting only in the highest 10%.

In another context we've seen claims that (over the long term omitted) transfers out of DB are wrong 90% of the time and an IFA reporting them to be good (now) 40% of the time (because of current conditions). I expect that both are correct and the first is flawed if used for decision-making because conditions do affect outcomes.


Thanks again for the spreadsheet work, I like and appreciate it.

jamesd

What is success?

In my last post I mentioned reduced LTA charge as a success measure but it's not so simple.

We're considering two different strategies:

1. Original. This one reduces guaranteed income significantly by having the LTA charge come out of DB and does best if markets never dip during the time it takes to crystallise.

2. Modified. This one makes guaranteed income high and seeks to exploit the chance of a market drop to get a better outcome.

1 is a pure "good investment times" plan. Boom or just nothing significantly bad is where it excels.

2 is a mixed outcome plan. The guaranteed income provides a higher guaranteed income floor. If there's a big and sustained market drop or even just poor conditions this is the one that will excel.

Whether 1 or 2 is likely to produce the best outcome depends greatly on initial conditions.

I know the initial conditions in many equity markets are near or at record high cyclically adjusted price/earnings ratios and I know that this has an excellent inverse correlation with ten year expected returns as well as correlating with the probability of a big drop.

That causes me to think that current conditions are relatively favourable for 2 with its better OK to downside performance than for the good weather 1 approach. But it remains possible that the future could turn out to be one that makes 1 better.

In essence 2 has a good deal more hedging and resilience to less good outcomes than 1 but either could win.