Page 2 of 3

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 5:55 pm
by Gilgongo
TheRIT wrote:Looking back the end of 1994 was a very good time to retire.


I just checked this too. With respect, the fact that Dod's not heard of sequence risk until now isn't surprising :D I also see that that by some estimates, it's not much of a problem beyond about the first 10-15 years of a 30-40 year retirement, and of course with a sufficiently large portfolio it'll take some pretty terrible management to run into much serious trouble anyway (just ask Donald Trump!).

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 6:00 pm
by Gilgongo
tikunetih wrote: Every so often I top-up the cash from asset sales, selling down overweight positions to return them to target allocations.


Would you mind explaining how you determine "overweight" in that case? How many assets are you dealing with?

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 6:12 pm
by TheRIT
Gilgongo wrote:I just checked this too. With respect, the fact that Dod's not heard of sequence risk until now isn't surprising :D I also see that that by some estimates, it's not much of a problem beyond about the first 10-15 years of a 30-40 year retirement, and of course with a sufficiently large portfolio it'll take some pretty terrible management to run into much serious trouble anyway (just ask Donald Trump!).


IMHO this is THE big problem with a self-funded I'm on my own retirement vs say a 'pooled group fund' designed to support multi-aged people for perpetuity with new contributors, FIRE entrants through to "those who no longer need the money". The day you retire decides your fate - you now have YOUR sequence of returns baked in. Let's use history, the US proxy again, my 2% example from above and let's say I have 40 years of retirement ahead of me. Get a favourable sequence and my wealth increases by x40, get an unfavourable sequence and it's less than x2. That is a ridiculously large divide.

That is the problem. To cover all historic eventualities as a proxy for what might happen in the future you need so much wealth that in the majority of cases you leave a 'metric truckload' on the table when you no longer have a need for it. Unless of course one wants to play the 'do you feel lucky punk' game...

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 6:33 pm
by Dod101
TheRIT wrote:
Dod101 wrote:...or try to live beyond their financial ability. ...

This is a critical point and as somebody who has had plenty of years in living a self funded retirement your experience (both psychological and mechanical) is invaluable to us "newbies". Would you be prepared to share in % terms what annual income to assets ratio you started with at the end of 1994 and what that % looks like today?


Well that is easy today. It is something over 5% currently which slightly surprises me but then there are some high income and very low growth shares around in the UK at the moment. I also hold four bond funds but they are in total a very small proportion of my total assets.

I really do not know what the percentage would have been at the end of 1994 because, like many who suddenly had a largish lump sum, I really did not know much about what I was going to do with it except that I knew I was not going to buy an annuity. I spent quite a bit of capital on upgrades to housing and so on and frankly probably lived on some of the capital for a while. After what I can only describe as a bit of trial and error (probably more of the latter) I realised that if I concentrated on shares that paid a half decent dividend at least the dividends would be fairly secure and so my version of a HYP was born.

Over that length of time, I have given away a significant amount, mostly to my children to help with house purchase, but certainly I now have much more capital than I had on 1 January 1995. I was never afraid of money but was always fairly cautious with it. I know some of my colleagues and friends retiring on much the same terms were, and some made the most awful mess of it. I think knowing your self and your capabilities is essential because some who got in a mess should never have had that sort of money in their hands in the first place.

As for being a good time to retire, compared to retiring on 31 December 2019, yes I would agree. The Covid crisis would certainly be a shock to any system, but do not forget that we had the tech bubble in 2000 which made and lost many fortunes, and then the 2008/9 crisis and now this. These are facts of life which occur every so often and we need to be prepared for them.

Dod

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 6:47 pm
by TheRIT
Dod101 wrote:Well that is easy today. It is something over 5% currently which slightly surprises me but then there are some high income and very low growth shares around in the UK at the moment. I also hold four bond funds but they are in total a very small proportion of my total assets.

I really do not know what the percentage would have been at the end of 1994 because, like many who suddenly had a largish lump sum, I really did not know much about what I was going to do with it except that I knew I was not going to buy an annuity. I spent quite a bit of capital on upgrades to housing and so on and frankly probably lived on some of the capital for a while. After what I can only describe as a bit of trial and error (probably more of the latter) I realised that if I concentrated on shares that paid a half decent dividend at least the dividends would be fairly secure and so my version of a HYP was born.

Over that length of time, I have given away a significant amount, mostly to my children to help with house purchase, but certainly I now have much more capital than I had on 1 January 1995. I was never afraid of money but was always fairly cautious with it. I know some of my colleagues and friends retiring on much the same terms were, and some made the most awful mess of it. I think knowing your self and your capabilities is essential because some who got in a mess should never have had that sort of money in their hands in the first place.

As for being a good time to retire, compared to retiring on 31 December 2019, yes I would agree. The Covid crisis would certainly be a shock to any system, but do not forget that we had the tech bubble in 2000 which made and lost many fortunes, and then the 2008/9 crisis and now this. These are facts of life which occur every so often and we need to be prepared for them.

Dod


Thanks for sharing. Your post reminds me of the saying "life is what happens when you're busy making plans". It's important to plan but not over plan because "stuff" happens.

Using your 5% as a proxy starting drawdown plus my previous example of historic returns shows that while you have more capital than you started because of a favourable sequence in 16% of cases a 5% starting withdrawal only increasing with inflation would now have £0. A big difference...

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 7:02 pm
by dealtn
TheRIT wrote:Using your 5% as a proxy starting drawdown plus my previous example of historic returns shows that while you have more capital than you started because of a favourable sequence in 16% of cases a 5% starting withdrawal only increasing with inflation would now have £0. A big difference...


Do you have a source for that?

I suspect it is over-simplified, so would like to be proved wrong.

For instance does that 5% withdrawal rate apply to all time periods, or is it statistically adjusted for market movements. That is to say does it treat 5% withdrawal the same for all market condition start points, or does it adjust for recent market change?

Imagine a 4% withdrawal rate starting "last year", and also a 5% withdrawal rate on the same basis. Now a year later a different retiree with a 5% withdrawal rate, but the market has fallen 20%. Which of the 2 previous scenarios does it match up with?

You need a pretty complex Monte-Carlo type analysis I would suggest to reach a conclusion of 16% so I am intrigued as to the study. It will be worthy of reading.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 7:13 pm
by TheRIT
dealtn wrote:Do you have a source for that?

I suspect it is over-simplified, so would like to be proved wrong.

For instance does that 5% withdrawal rate apply to all time periods, or is it statistically adjusted for market movements. That is to say does it treat 5% withdrawal the same for all market condition start points, or does it adjust for recent market change?

Imagine a 4% withdrawal rate starting "last year", and also a 5% withdrawal rate on the same basis. Now a year later a different retiree with a 5% withdrawal rate, but the market has fallen 20%. Which of the 2 previous scenarios does it match up with?

You need a pretty complex Monte-Carlo type analysis I would suggest to reach a conclusion of 16% so I am intrigued as to the study. It will be worthy of reading.


Not doing anything clever and it's not over simplified. I'm just using previous historic sequence of returns. As I've mentioned I've used US data as a proxy. This one will give you the story https://calculator.ficalc.app/ The only changes I made are set the Length of Retirement to 25 years (Dod's length of retirement so far) and the Constant Dollar Retirement Strategy to $50,000 (ie a starting drawdown of 5% increasing with inflation). 20 of the past 125 years (16%) would have failed.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 7:33 pm
by dealtn
TheRIT wrote:
dealtn wrote:Do you have a source for that?

I suspect it is over-simplified, so would like to be proved wrong.

For instance does that 5% withdrawal rate apply to all time periods, or is it statistically adjusted for market movements. That is to say does it treat 5% withdrawal the same for all market condition start points, or does it adjust for recent market change?

Imagine a 4% withdrawal rate starting "last year", and also a 5% withdrawal rate on the same basis. Now a year later a different retiree with a 5% withdrawal rate, but the market has fallen 20%. Which of the 2 previous scenarios does it match up with?

You need a pretty complex Monte-Carlo type analysis I would suggest to reach a conclusion of 16% so I am intrigued as to the study. It will be worthy of reading.


Not doing anything clever and it's not over simplified. I'm just using previous historic sequence of returns.


Ah!

I will revert to the original study it purports to be from in that case. Thanks anyway.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 8:58 pm
by Gilgongo
TheRIT wrote:That is the problem. To cover all historic eventualities as a proxy for what might happen in the future you need so much wealth that in the majority of cases you leave a 'metric truckload' on the table when you no longer have a need for it. Unless of course one wants to play the 'do you feel lucky punk' game...


Yes, and this is also why I would suggest that anyone on these boards saying "Just draw down on the dividends" or anything to do with natural yield in order to live in retirement is revealing they don't actually have skin in the game. With a pot big enough to do that (basically a million quid), you can give any advice you want. Unless of course it's not that big, in which case you'll start playing the same game as the rest of us:

https://finalytiq.co.uk/natural-yield-t ... -strategy/

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 9:13 pm
by TheRIT
Gilgongo wrote:Yes, and this is also why I would suggest that anyone on these boards saying "Just draw down on the dividends" or anything to do with natural yield in order to live in retirement is revealing they don't actually have skin in the game. With a pot big enough to do that (basically a million quid), you can give any advice you want. Unless of course it's not that big, in which case you'll start playing the same game as the rest of us:

https://finalytiq.co.uk/natural-yield-t ... -strategy/

Full disclosure: As I mentioned above I'm trying to do just that. 3 years of cash + enough investments to spend "only" 85% of the dividends and to make matters worse 5% of my portfolio is gold which doesn't pay much of a dividend.

Still feels like I have plenty of skin in the game...

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 10:10 pm
by Gilgongo
TheRIT wrote:As I mentioned above I'm trying to do just that


Not to press the point, but on what basis would you therefore disagree with the article I link to when it says, "... a natural yield approach is a bonkers retirement income strategy for virtually all but very wealthy retirees, who mostly rely on other sources of steady income." Is it that your 85% rule means you'll need to operate a parallel strategy to mitigate the fluctuations?

Re: Minimising sequence risk in a mixed portfolio

Posted: August 15th, 2020, 10:53 pm
by TheRIT
Gilgongo wrote:Not to press the point, but on what basis would you therefore disagree with the article I link to when it says, "... a natural yield approach is a bonkers retirement income strategy for virtually all but very wealthy retirees, who mostly rely on other sources of steady income." Is it that your 85% rule means you'll need to operate a parallel strategy to mitigate the fluctuations?


I think the first point is that drawdown is both a psychological thing as well as a mechanical thing. All these articles that keep getting regurgitated in various forms only ever address the second. I'm yet to meet one of these authors (and that includes all the famous FIRE bloggers) who's actually living what they write about when it comes to 4% rules, SORR etc. When setting my initial strategy I addressed each in turn:

- Psychological. My wealth came from genuine hard graft (long hours, very high stress, etc) which I then put into low cost tracker investing. I know what it took to build what I have and as I travelled along my road to FIRE I just knew I wouldn't be able to sell what I had worked so hard to accrue. I can handle the market gyrations no problem but actively selling anything to eat is a completely different thing. I therefore did some backtesting (admittedly with US) data that showed if I started out with spending 85% of the dividends plus 3 years in cash that by the end of the worst bear markets I had been able to hold my spending substituting the drop in dividends with cash. It's the cash buffer that helps me hopefully mitigate the dividend fluctuations and the 85% helps me rebuild my cash fund after a bear market. As mentioned above that worked out at about a 2% withdrawal when I FIRE'd.

- Mechanical. For this I looked wider than just the US studies as that's just 1 dataset. For example 50% UK equities : 50% UK bonds from a Wade Pfau study only allowed a withdrawal of 3.05% to not run out of money after 30 years. There are also worse developed country examples that haven't been ravaged by a world war. For example Australia was 2.9%. In contrast a UK investor holding 50% international equities : 50% international bonds could support 3.26%. I first retired at 46 so likely needed a lot longer than 30 years and I also have investment expenses of about 0.25%. So for me 2.5% felt about right.

I then work off the worst case of each when I set my spending each year. Currently that worse case is driven by the dividends. With time if I get a favourable sequence of returns I suspect it will switch to be the 2.5% of initial withdrawal increased annually for inflation.

Of course maybe the article is right as if history repeats I'm likely to die a very wealthy man. That is I'm bonkers which makes my strategy a bonkers retirement income strategy.

Full disclosure: Right now it's a bit mute for me anyway as I'm currently not testing it in these dividend reduced times as I'm doing a bit of work that can 100% be done from home, carries nominal flexible hours, carries no stress and leaves me plenty of time to do what I want when I want to do it. It's the FI bit of FIRE I have to thank for getting me into this position. I feel my point is still valid though as I did actually FIRE during the correction at the end of 2018 and live this before returning to FI.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 9:21 am
by Dod101
Gilgongo wrote:
TheRIT wrote:That is the problem. To cover all historic eventualities as a proxy for what might happen in the future you need so much wealth that in the majority of cases you leave a 'metric truckload' on the table when you no longer have a need for it. Unless of course one wants to play the 'do you feel lucky punk' game...


Yes, and this is also why I would suggest that anyone on these boards saying "Just draw down on the dividends" or anything to do with natural yield in order to live in retirement is revealing they don't actually have skin in the game. With a pot big enough to do that (basically a million quid), you can give any advice you want. Unless of course it's not that big, in which case you'll start playing the same game as the rest of us:

https://finalytiq.co.uk/natural-yield-t ... -strategy/


Gilgongo sounds a little bitter almost. If he has insufficient funds to see a decent retirement on the income his capital will produce maybe he needs to rethink the whole proposition of retiring early (which it sounds as if is what he wants) Anyone trying to live off the natural yield I can assure you has 'skin in the game', and I think it would be unwise for anyone to try to live off capital in any other way. I have never played any game with my investment funds and of course what is being ignored here are the investment returns. Living within your income is the only sensible mantra in retirement as at any other time and cutting your cloth to suit that is the only way it will work.

Sorry, probably unwelcome comments from an oldie who has been through it.

Dod

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 10:52 am
by Gilgongo
Dod101 wrote:I think it would be unwise for anyone to try to live off capital in any other way


I think we need to agree that this is an extreme position ;)

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 11:15 am
by Wuffle
'and I think it would be unwise for anyone to try to live off capital in any other way.'

Low earners in the benevolent European states are more likely to build up and run down capital.
The decision making process being 'I enjoy it now or it pays for my care - which ultimately I can get for nothing'.
It isn't as financially illiterate as at first appears.

The tricky place is in between rich and obviously not rich.

W.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 11:17 am
by Gilgongo
TheRIT wrote:I think the first point is that drawdown is both a psychological thing as well as a mechanical thing.


Thanks for detailing this. The psychological aspect is certainly significant, and a hybrid of natural yield with some capital erosion (and presumably keeping an even asset allocation along the way to mitigate sequence risk to an extent) is interesting.

So to return to the topic, because practice of maintaining asset allocation feels to be the same thing as what I was proposing.

Let's say we want to maintain the 50:50 stocks/bonds ratio you mention. How exactly would you choose to rebalance if not by looking at each asset (bond fund/ETF, HYP stock, whatever) historic price and dividend movements? Would that not imply a ranking operation and then selling the winners to bolster the losers? And by "losers" does that mean "divi cutters"?

This the detail I'd like to get into rather than generalising on broad strategy.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 12:21 pm
by 1nvest
Each and every asset have endured total net real losses over decade or longer periods. Drawing a income, no matter how drawn - dividends, interest, selling shares - would have induced further capital decline. Driving into the future looking in the rear view mirror - a exceptional historic period across the 1980's/1990's decades for instance (didn't really matter what stock/bond allocations you used, you made heaps of rewards fundamentally due to transition from very high to very low yields), and you run the risk of a crash.

If you own your own home then 'rent' is liability matched. The home value might also cover late life care costs for the longer survivor of a partnership (in most partnerships the first to die more often dies at home cared for by their partner). If your spending is £20,000/year then presently Linkers (Index Linked Gilts) come with around a 2.5%/year cost, so to buy a assured £20,000 in ten years times costs around 1.28 times more being needed to be deposited now - so £25,600 to buy £20,000 of inflation adjusted income in 10 years time, for £20,000 in a years time costs £20,500 ....etc. for all other years 1 to how-ever-many years out. A state pension might reduce that, if in x year times you receive £9000 state pension then that reduces the £20,000 amount having to be found down to £11,000.

For a 59 year old about to retire, who owned their own home so didn't have to find/pay rent, and whose spending was £20K/year, then a quick calculation indicates they'd need around £550K in a Linkers ladder assuming they'd receive £9K state pension at age 67, where that ladder assumes a 90 year age life expectancy. If instead invested in shares - spending a above average yield and that could work out fine, could lead to all sorts of problems/risks.

With Linkers providing a DIY annuity sorted, any amounts above/beyond that can be invested however preferred. If heirs are involved that surplus might grow fast enough to offset the decay/spending of the Linkers. If you die much sooner the value of the Linkers are also still largely there for heirs, unlike for a actual annuity where the value might be fully lost. If that surplus amount is £250K and that's invested in all-stock that earns 4% annualised real over the 30 years that the linkers are drawn-down/spent, then the combined portfolio value will have maintained its inflation adjusted value, Start with £800K, end with a inflation adjusted £800K still available in 30 years time. If limited to £550K and/or no heirs then you might just run the Linkers/drawdown. If less than £550K then you have to take on more risk and hit-n-hope.

The great case is where you have a decent inflation adjusted occupational pension alongside the state pension that in combination covers spending. In which case £0 savings is OK. Better still if you also have some savings as well.

In a bad case, perhaps £200K home value, £400K savings, I'd suggest pushing with the Talmud asset allocation, equal weightings of UK home, US stock, Gold in equal measure. £, $, global currency diversification, land, stock, commodity asset diversification. Drawing a 3.3% SWR from that (£20K/year). Likely that would be fine. As home values are illiquid, some UK stocks can be substituted in as part of that.

Yet another choice is to employ Harry Browne's Permanent Portfolio, 25% in each of UK stocks, gold, 30 year Gilt, cash deposits ... as a alternative to Linkers. Assume a conservative +2% real from that (likely results would be better), which is 4.5%/year more than Linkers, and you'd also likely be OK.

Fundamentally much of early years SoR risk can be largely mitigated via appropriate asset allocation. The risk is massively more pronounced with the likes of all-stock (HYP/whatever).

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 2:15 pm
by Gilgongo
1nvest wrote:With Linkers providing a DIY annuity sorted,


Since we're way off topic anyway... I'd been assuming that the growing pressure on the UK govt to stop using RPI means that anyone relying on index-linked gilts will need to prepare for up to a 30% drop in the future. I think there's been a stay of execution, but I'm not sure I have your confidence in Linkers being much of an answer in the long term.

Be that as it may, the same overall question applies to you too I think (although I'm not sure I understood a lot of what you were saying): how exactly would you go about re-blanancing?

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 2:56 pm
by TheRIT
Gilgongo wrote:Thanks for detailing this. The psychological aspect is certainly significant, and a hybrid of natural yield with some capital erosion (and presumably keeping an even asset allocation along the way to mitigate sequence risk to an extent) is interesting.

So to return to the topic, because practice of maintaining asset allocation feels to be the same thing as what I was proposing.

Let's say we want to maintain the 50:50 stocks/bonds ratio you mention. How exactly would you choose to rebalance if not by looking at each asset (bond fund/ETF, HYP stock, whatever) historic price and dividend movements? Would that not imply a ranking operation and then selling the winners to bolster the losers? And by "losers" does that mean "divi cutters"?

This the detail I'd like to get into rather than generalising on broad strategy.


My strategy has me dividing my wealth into more than just stocks and bonds. Currently it calls for 15% UK Equity (my home country), 15% Asia Equity (my likely soon to be home - thank you FI again), 5% EM Equity, 25% International Equity (40% US, 40% Europe, 20% Japan), 5% Gold, 10% REIT's, 19% Bonds (Linkers and Corporate) and 6% cash (my 3 years of cash). Other than a HYP which I'm slowly exiting this is achieved with tracker funds/ETF's. So my rebalancing within each asset class is taken care of by the fund itself. QED.

So now I just have to rebalance the asset classes themselves remembering I'm only spending dividends so don't need to sell anything to eat (in theory). During accrual I was easily able to achieve this by just always buying the worst performing asset class with new money. During drawdown I positioned myself so that I would rebalance when any large holding moved by 5% from target (eg if UK equity moved by 15% to 10% or 20%) and when any small holding moved by 25% from target (eg if EM Equity moved from 5% to 3.75% or 6.25%) I'd rebalance. I have low trading expenses across the board and the plan was to do this rebalancing within tax friendly wrappers (ISA's/SIPP's) or within the annual CGT free allwance.

So far I haven't had to rebalance to the plan as my new work covers my expenses and while I'm not adding to my wealth anymore I'm able to rebalance by investing my dividends in the worst performing asset class. The only small exception so far has been in March I sold some gold and bought some Asia equities. Total trading cost £0, a little buy sell spread and no tax as done within an ISA. Minimising expenses and taxes is one of my mantra's.

When I do FIRE again I'm now seriously considering not even rebalancing ever again which takes the psychology right out of it. If the market performs averagely I know this will skew things but in my case I think I'm not far off it not even mattering as investing IMHO is about directionality rather than perfection. My reason is that as we saw with Dod's example if I get anything but the very worst sequence of returns my wealth is going to increase far faster than my spending meaning my withdrawal rate is going to decrease with time. This puts my sequence of returns risk within the first few years of FIRE which for me started in late 2018 and already I'm heading favourably with my wealth up 8%. So I started with a planned withdrawal of 2%. Using my previous historic examples as a quick proxy with a 2% withdrawal, if I don't rebalance, even just average performance will see my wealth increase by 50% in real terms in 8 years. Assuming dividends increase with wealth now my withdrawal is 57% of my dividends. How often can we expect that to occur and I still have my 3 years of cash if it does. At that point unless we have a black swan I've probably won the game and never need to look at my investments again. Then if we do have a black swan whether or not I rebalanced probably isn't going to matter.

Probably not the answer you're looking for but I think it's a nice example of the phrase mentioned earlier - life is what happens while you're busy making plans.

Re: Minimising sequence risk in a mixed portfolio

Posted: August 16th, 2020, 5:39 pm
by 1nvest
Gilgongo wrote:
1nvest wrote:With Linkers providing a DIY annuity sorted,

... but I'm not sure I have your confidence in Linkers being much of an answer in the long term.

Be that as it may, the same overall question applies to you too I think (although I'm not sure I understood a lot of what you were saying): how exactly would you go about re-balancing?

With a liability matched ladder the idea is that you have a bond that matures each year to the inflation adjusted value of your anticipated spending in that year, so there's no rebalancing, you just load the ladder from the offset, buying one that matures in a year that matches that years spending, another that matures in two years time that matches that years spending ...etc. The difficult part other than having the funds to load that, is predicting what your spending needs will be in future years. And yes there is a risk that the state can renege by such tactics as revising the rules (taxation, actual inflation rate ...etc.). The claims are often that the state doesn't default on its obligations, however they have done so numerous times in the past, not full defaults, but partial defaults in disguise. Henry VIII for instance was known as copper-nose-'enry because silver coins that represented the silver value of the coin were minted with copper interiors during his realm, and when the silver plating wore down on his image over time, the copper beneath would show through.