Pensions already exceed LTA - How do BCEs and Taxation apply

jamesd

EdSwippet said:

jamesd said:

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.

Taking opportunistic advantage of any dips that come along seems to encapsulate the idea, and that seems fine in theory. I think the main question here is how to make it actionable while minimising regret at missing later and better opportunities.

Would you be entirely happy to have saved 10% of the LTA charge if you had pulled the trigger early on a market drop that, had you waited, could have saved you 50%, 75%, or even the entire LTA charge? Without resorting to attempting market timing (or worse, voodoo charting), how would someone set the trigger points here?

I'd say that the biggest issue is deciding whether circumstances are favourable for using the approach at all, as zagfles has helpfully illustrated. Fortunately the cyclically adjusted P/E provides us with help on that front and the ten year inverse correlation is also a convenient time period. Now happens to be favourable in many major equity markets so I made the suggestion.

In the past I've occasionally posted more detailed suggestions for how to handle less than sufficient dips. But the time periods involved are in the range where market movements more resemble a random walk than something with a useful correlation to give guidance. 

Given the lack of short term predictability then unless the total dip needed is quite low I think it's best to pick some sort of individual compromise that does exploit dips to some extent. Then simply accept that there is no knowable in advance optimal strategy for the time periods involved and try not to kick yourself when you find that if you'd have known the future you could have done better.

I'm not a fan of charting either, though momentum might have some value. All we have to do is look at the 2020 dip to see not a lot of momentum when it came to staying down, though do something every month would have worked for a while in many cases.

jamesd

Dead_keen said:

I think it's helpful to compare how much smaller the OP's DB pension would be over a range of investment returns for the SIPP (up to the date that the SIPP is crystallised - which on the OP's current plan would seem to be immediately before the DB is taken).

As jamesd says, in a very narrow range of circumstances (i.e. a substantial fall in value of the SIPP) there will be no reduction in the DB pension because it will suffer no LTA charge.  But in all other circumstances there may be a reduction in the DB pension.

So for me, the key issues would be:

1. To quantify the likely reduction in the DB scheme over a range of outcomes and decide what is acceptable.  If a high probability of a large reduction in the DB scheme is unacceptable then something may need to be done.  That might be crystallising the SIPP now or it might be something else.  If the likely range of reduction in the DB pension is acceptable (e.g. it is pretty much rounding on all OP's other sources of income) then there is no need to do anything.  

2. If it is unacceptable then there is a question of what could be done.  This could be drawing the DB pension early, crystallising the SIPP later, transferring the DB to a DC scheme or something else.  Some of these might help other things (e.g. what's available for the OP's spouse on death) and some might not. 
 
3. How to manage the OP's effective income tax rate going forward vs the age 75 BCE on the SIPP.  But we don't have enough information to comment on that as we don't know why the OP expects to be a higher rate / additional tate taxpayer "for the foreseeable future".

Agreed about range - and do pay attention to starting conditions.

But recall also that the initial proposal was "use the LTA to cover the SIPP in full and allow the DB scheme to suffer crystallisation taxes" and in context I think the intent there was the whole DC at once. Crystallising the DC first ensures crystallising the main variable but there's then little scope for future outcome variation, limited to just when to take the DB and whether the LFA changes.

If DC crystallisation is partial and intermingled with DB crystallisation the analysis gets way more complex.

Agree that 3 is an issue and given the amounts involved it's likely that at a minimum withdrawing from the DC the whole remaining basic rate band every year will be appropriate. And actually I suspect that higher rate to some extent for a few early years will also be merited but we just don't know enough about the overall income situation to actually recommend with any great confidence.

jamesd

zagfles said:

But how does it identify a dip is coming (rather than a plateau for a while while returns catch up with capital, or lower growth than would otherwise have occurred?) And the size of the dip? And you've still not explained how you identify the bottom of a dip, as per Ed's post.

There has been some work done on correlating the chance of a big dip with cyclically adjusted P/E but I haven't looked into it in much detail, just enough to note that there does seem to be a potentially useful correlation. Regrettably since I haven't looked much I don't have handy any references for you. Anyone else have some?

The main point of interest for me is the "are conditions favourable for the approach?" question and hopefully after fair bit of reading you'll agree that CAPE10 is viable for that.

I don't think knowing in advance the bottom of a dip is possible. The best that I think is achievable is approaches that minimise an individual's regret while recognising that optimal is unachievable. That IMO calls for seeking to use dips even though the optinal answer might turn out to have been not to have done so, for any particular magnitude of dip.

zagfles

jamesd said:

zagfles said:

jamesd said:

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.

Sorry but that's rubbish. 

Unfortunately for you, you're wrong about that and it's been checked and found true for every significant equity market in the world.

And since it is true, it's entirely sensible to use it to narrow down the use of the approach to only relatively favourable starting times, including now for many major markets.

To get started you might usefully read this post and the pages it links to, then move on to the more academic paper referenced by Guyton: https://forums.moneysavingexpert.com/discussion/comment/70696736/#Comment_70696736  .

Try reading the next sentence in my post instead of just the first 4 words. Then it will become blatently obvious that what I referred to as "rubbish" isn't the "inverse correlation between cyclically adjusted prjce/earnings ratios of markets and their ten year returns", it's "...your arguments at their heart seem to largely largely rely on that fallacy."

zagfles

jamesd said:

zagfles said:

But how does it identify a dip is coming (rather than a plateau for a while while returns catch up with capital, or lower growth than would otherwise have occurred?) And the size of the dip? And you've still not explained how you identify the bottom of a dip, as per Ed's post.

There has been some work done on correlating the chance of a big dip with cyclically adjusted P/E but I haven't looked into it in much detail, just enough to note that there does seem to be a potentially useful correlation. Regrettably since I haven't looked much I don't have handy any references for you. Anyone else have some?

The main point of interest for me is the "are conditions favourable for the approach?" question and hopefully after fair bit of reading you'll agree that CAPE10 is viable for that.

I don't think knowing in advance the bottom of a dip is possible. The best that I think is achievable is approaches that minimise an individual's regret while recognising that optimal is unachievable. That IMO calls for seeking to use dips even though the optinal answer might turn out to have been not to have done so, for any particular magnitude of dip.

"Conditions being favourable" might mean it's now more likely than normal. So what, unless that's quantified it's not much use. It could mean there's now a 1 chance in 5 instead of a 1 chance in 10. So although better than normal, still far more chance of losing than winning.

jamesd

My response was for both aspects that I quoted.

You also wrote that "Your only answer to that seems to be based around market timing." when PE10 isn't about market timing and did a great deal to focus on the wrong side outcomes in unfavourable starting circumstances in previous posts, strongly suggesting to me that at least at the time of those posts you didn't accept the PE10 inverse correlation with returns.

However, we're chatting here so I can ask you:

Do you accept that research has shown that PE10 is inversely correlated with ten year equity returns?

Do you accept that that inverse correlation makes it possible to increase the chance of picking good starting points and reduce the chance of picking bad ones?

Thrugelmir

I started investing in 1999 when UKX was around 6400 and it is only 11% higher today after 22 years so it would not have been a good investment over that time.

Never ceases to amaze me how many people totally ignore dividends (and the reinvestment of, which results in compounding) when comparing market indices. One of the most fundamental aspects of investing is to understand what you are investing in, whatever it might be. 

zagfles

jamesd said:

My response was for both aspects that I quoted.

You also wrote that "Your only answer to that seems to be based around market timing." when PE10 isn't about market timing

Of course it's about market timing! It's the very definition of market timing. "...unfavourable starting circumstances" below is proof

and did a great deal to focus on the wrong side outcomes in unfavourable starting circumstances in previous posts, strongly suggesting to me that at least at the time of those posts you didn't accept the PE10 inverse correlation with returns.

I focused on all conditions, not just "unfavourable" ones.

Do you accept that research has shown that PE10 is inversely correlated with ten year equity returns?

Do you accept that that inverse correlation makes it possible to increase the chance of picking good starting points and reduce the chance of picking bad ones?

Possibly. But I don't believe it's significant enough to make a material difference to any decision. If it did, it would have been expolited enough by professional market timers to get them new yachts and private islands until the effect was cancelled out. Just like any other indicator of market direction. 

Are you going to sell all your equities and wait for this dip?

jamesd

zagfles said:

Are you going to sell all your equities and wait for this dip?

I'm not in a position to potentially exceed the lifetime allowance so this approach has no value for me.

But yes, I've reduced my equity holdings as suggested in Guyton's work. Not eliminated them, of course, because my job is to cover both possibilities of long growth or the more likely prospect of a short term decline within the next few years.