How Much to Accumulate for Retirement? (Excel analysis)

Thrugelmir

Audaxer said:

pensionpawn said:

Audaxer said:

pensionpawn said:

Linton said:

BritishInvestor said:

Audaxer said:

SouthCoastBoy said:

It seems most people take the swr as the way forward. My understanding of swr is that the pot will be preserved at the swr? Is that incorrect? 

Whether or not the pot will be preserved will depend on the Sequence of Returns. If for example you get a poor sequence of returns in the first decade of retirement, the pot might not last throughout retirement if you don't have a cash buffer to draw from in loss years.

I'm not sure the cash buffer is going to be a massive help (certainly not if you take it from the equity allocation).

https://finalytiq.co.uk/cash-reserve-buffers-withdrawal-rates-old-wives-fables-retirement-portfolio/

 - None of the strategies tested matched what it is proposed here that one does with a buffer. ie use it for covering crashes
 - The strategies were mostly tested when bond returns were of some significance - one of the major reasons for choosing a cash buffer is that safe bonds now seem to be sub-optimal.  If bond were returning say 5% the need for cash would be significantly reduced.
 - In retirement I would think most people are prepared to sacrifice returns if necessary to reduce risk

If having a large cash buffer enables you to sail through a crash worry-free when you would otherwise lie awake at night then a cash buffer is a massive help.

Essential for the early years however as time (decades) go by and your funds hopefully continue to grow significantly, you can start to whittle down the cash holding as with decreasing years ahead it has performed it's role.

If your investments grow significantly in retirement, I would say that is a good reason for converting more to cash to enable you to lower your risk profile and spend more of it if you want. As others have previously said, there is no point in being the richest person in the graveyard.

I think it depends on the value of your holding compared to what you need to withdraw pa. If you have £400k in funds and £100k in cash, are 65 and drawdown £20k per year it may be worth keeping your 5 year cash safety net. However if you now have £800k in funds and still £100k in cash at 75 you may well be asking what sort of catastrophic world event is going to leave you struggling to get by at 80 after you've used all your cash? Bearing in mind you have a full SP and say £0.5M equity in your house. What we don't spend from our drawdown each year will go into savings such as an ISA, however besides that I'm not going to significantly add to my cash buffer.

That's fine, but if your investments have grown to that extent, I'm not looking at it from the point of view of being more cautious. In your example, if your investments have gone up to £800k and you had £100k cash, you could if you wanted to, convert a bit more of the £800k to cash, say another £100k, to spend it on luxuries over the next few years or give some to good causes.  You would still have £700k in investments and £100k cash, which would still be enough to drawdown £20k per year for your remaining years. My point being, you can't take it with you.   

Periods of good returns are more often or not followed by famine. There are two types of investors those that have been humbled, and those that are yet to be. When markets are high it's easy to become intoxicated with the thought of having reached the mountain peak and standing there admiring the view. 

DT2001

Linton said:

DT2001 said:

If you employ a strategy that requires no chance of downward adjustment of income you will be overly cautious.
Flexibility allows you to have a greater overall level of income.

If retiring before SPA without any guaranteed income maybe a combination of cash reserve and flexibility is ideal. It is the time when supposedly you’ll be spending most in retirement (unless needing care later on) but if you can tweak the timing of say long haul holidays by a year or so you must improve the longevity of your pot.

It isnt a matter of being cautious but rather of using your assets  to their best advantage.  In particular a 60/40 portfolio is seen as reasonable for retirement.  There should be plenty of money in the 40% to avoid selling equities for at least 5 years and for taking your long-booked luxury holiday.  Better to see the 40% as part of your armoury rather than simply as padding .  This implies that you should devote as much time to the allocation of your non-equity assets as you do to your equities.

I can't see that the savings from simply cutting excess expenditure for the short term without seriously reducing your standard of living are going to be significantly large to justify the pain. Rather than reacting in a panic to a crash by cutting all your expenditure by as much as say 25% for a year or two it would be better to plan to reduce expenditure by a few % for the rest of your life which is something you probably would not notice.  Such a small change would probably be reversed when your equities recovered anyway.

Sorry, I didn’t understand that your cash buffer was part of the 40%.

Guyton-Klinger would give you a better starting withdrawal rate which even after a 10% reduction would still be more than your ‘normal’ SWR.

If it is part of your plan to react to crashes there shouldn’t be a panic.

When I first calculated my ‘number’ I had 50% of my annual budget in my holiday fund so adjusting for prevailing market conditions wouldn’t worry me. Maybe having been self employed for 25 years with quite a variable income it is ‘normal’ for me to continually adjust expenditure without a feeling of lowering my living standards.

BritishInvestor

DT2001 said:

Linton said:

DT2001 said:

If you employ a strategy that requires no chance of downward adjustment of income you will be overly cautious.
Flexibility allows you to have a greater overall level of income.

If retiring before SPA without any guaranteed income maybe a combination of cash reserve and flexibility is ideal. It is the time when supposedly you’ll be spending most in retirement (unless needing care later on) but if you can tweak the timing of say long haul holidays by a year or so you must improve the longevity of your pot.

It isnt a matter of being cautious but rather of using your assets  to their best advantage.  In particular a 60/40 portfolio is seen as reasonable for retirement.  There should be plenty of money in the 40% to avoid selling equities for at least 5 years and for taking your long-booked luxury holiday.  Better to see the 40% as part of your armoury rather than simply as padding .  This implies that you should devote as much time to the allocation of your non-equity assets as you do to your equities.

I can't see that the savings from simply cutting excess expenditure for the short term without seriously reducing your standard of living are going to be significantly large to justify the pain. Rather than reacting in a panic to a crash by cutting all your expenditure by as much as say 25% for a year or two it would be better to plan to reduce expenditure by a few % for the rest of your life which is something you probably would not notice.  Such a small change would probably be reversed when your equities recovered anyway.

Sorry, I didn’t understand that your cash buffer was part of the 40%.

Guyton-Klinger would give you a better starting withdrawal rate which even after a 10% reduction would still be more than your ‘normal’ SWR.

If it is part of your plan to react to crashes there shouldn’t be a panic.

When I first calculated my ‘number’ I had 50% of my annual budget in my holiday fund so adjusting for prevailing market conditions wouldn’t worry me. Maybe having been self employed for 25 years with quite a variable income it is ‘normal’ for me to continually adjust expenditure without a feeling of lowering my living standards.

As has been discussed on other threads, GK can see some pretty big cuts if unfavourable markets are encountered - might not be suitable for everyone.

michaels

Spreadsheet: https://docs.google.com/spreadsheets/d/1QGrMm6XSGWBVLI8I_DOAeJV5whoCnSdmaR8toQB2Jz8

Good explanation of cash buffer utility (zero): Our emergency fund is exactly $0.00 – Early Retirement Now

OldScientist

BritishInvestor said:

OldScientist said:

BritishInvestor said:

SouthCoastBoy said:

It seems most people take the swr as the way forward. My understanding of swr is that the pot will be preserved at the swr? Is that incorrect? 

Yes, incorrect. SWR allows the pot to be drawn down if required (unless you have explicitly specified a legacy amount in your modelling). 

While the UK SWR for a 30 year retirement period is around 3.0-3.5% depending on asset allocation (e.g. see paper by Estrada at https://blog.iese.edu/jestrada/files/2018/03/MaxWR.pdf), historically the final pot ranged from 0 to more than 6 times the original (in real-terms) again dependent on asset allocation. Historically, in order to leave a minimum of about half the original pot (in real terms), the withdrawal would have had to have been  1.7-2.2% although the largest pot would then have been about 10 times the initial value. The latter values are my own calculations using UK returns and inflation from https://www.macrohistory.net - this dataset gives slightly more pessimist outcomes than that mostly used in the literature, see https://www.london.edu/-/media/files/faculty and research/subject areas/global investments yearbook.pdf, but $6k per year is a bit much for my retirement budget!

"historically the final pot ranged from 0 to more than 6 times the original (in real-terms) again dependent on asset allocation"

That's the key takeaway - the vast range of historical (and therefore future?) outcomes. We just don't know whether we are going to be the lucky 6x starting pot, or running out way before we wanted to.






"but $6k per year is a bit much for my retirement budget!"

I wonder if there's enough of a market for DIY investors to pay for a tool that contained this data....

There may very well be - I guess it depends on the pricing (and how much Credit Suisse charge to lease the data for for onward commercial use - if that isn't $6k per year). 

Thrugelmir

OldScientist said:

BritishInvestor said:

OldScientist said:

BritishInvestor said:

SouthCoastBoy said:

It seems most people take the swr as the way forward. My understanding of swr is that the pot will be preserved at the swr? Is that incorrect? 

Yes, incorrect. SWR allows the pot to be drawn down if required (unless you have explicitly specified a legacy amount in your modelling). 

While the UK SWR for a 30 year retirement period is around 3.0-3.5% depending on asset allocation (e.g. see paper by Estrada at https://blog.iese.edu/jestrada/files/2018/03/MaxWR.pdf), historically the final pot ranged from 0 to more than 6 times the original (in real-terms) again dependent on asset allocation. Historically, in order to leave a minimum of about half the original pot (in real terms), the withdrawal would have had to have been  1.7-2.2% although the largest pot would then have been about 10 times the initial value. The latter values are my own calculations using UK returns and inflation from https://www.macrohistory.net - this dataset gives slightly more pessimist outcomes than that mostly used in the literature, see https://www.london.edu/-/media/files/faculty and research/subject areas/global investments yearbook.pdf, but $6k per year is a bit much for my retirement budget!

"historically the final pot ranged from 0 to more than 6 times the original (in real-terms) again dependent on asset allocation"

That's the key takeaway - the vast range of historical (and therefore future?) outcomes. We just don't know whether we are going to be the lucky 6x starting pot, or running out way before we wanted to.






"but $6k per year is a bit much for my retirement budget!"

I wonder if there's enough of a market for DIY investors to pay for a tool that contained this data....

There may very well be - I guess it depends on the pricing (and how much Credit Suisse charge to lease the data for for onward commercial use - if that isn't $6k per year). 

Not Credit Suisse. 

The DMS dataset used in the Yearbook The DMS dataset provides annual total returns for stocks, government bonds, treasury bills, inflation rates, exchange rates, and maturity and equity premiums for 37 markets. 28 of these start in 1900, with 121 years of data; nine start later, but typically have more than 50 years of data. Morningstar is the exclusive distributor of the DMS dataset. The annual licence fee is USD 4000 for existing clients of certain Morningstar products or USD 6000 to others. Please contact Masa Krzic at Masa.Krzic@morningstar.com or phone +44(0)20 3194 1451.

Assume that Morningstar have a licencing arrangement with the London Business School. 

Further background

Summary of the impact

The Dimson-Marsh-Staunton study of long-run global asset returns involved (i) the construction of a unique database of long-run returns on stocks, bonds, bills, inflation, currencies, GDP and population growth for 22 countries since 1900; and (ii) interrogation of this resource to answer contemporary questions in investment, finance, and regulation. It underpins a worldwide reappraisal of the size of the equity risk premium. The project has informed the strategy of the world's largest investors; influenced cost of capital estimates and real investment decisions in leading corporations; and guided regulation of financial institutions, utilities, and other businesses.

Underpinning research

The research was produced by Elroy Dimson, Paul Marsh and Mike Staunton (DMS). At the time of research and publication, Dimson and Marsh were Professors of Finance at LBS; they are now Emeritus Professors. Staunton was and is Director of the London Share Price Database at LBS.

Previous knowledge about long-run returns was almost exclusively for the USA. The University of Chicago's CRSP database showed that the historical annualised equity premium (the amount by which stocks beat treasury bills) was 6.25% over 1926-99. This became received wisdom in textbooks and business schools worldwide, and it was applied by practitioners as if it were universal. DMS conjectured that the US focus was likely to overstate global investment returns.

The DMS research began with an analysis of long-run UK asset returns (2001, Journal of Business). This utilised the comprehensive London Share Price Database (LSPD), created and maintained by LBS, and available to researchers worldwide.

Beyond the important evidence for the UK, there was a need for a more comprehensive global study. DMS painstakingly assembled a multi-country database of asset returns, initially for 16 countries since 1900. Their analysis was published as Triumph of the Optimists (2002). The DMS global database has been updated, extended, and analysed intensively; see the Global Investment Returns Yearbook and Sourcebook. It spans 22 countries, covering 96% of global stock market capitalisation in 1900. It embraces countries where returns were adversely impacted by world wars (Germany, Japan, Austria-Hungary), civil war (Spain), and revolutions (Russia, China).

Some key findings from Triumph and related research (referenced below) are that (i) the equity premium has historically been 4.1%, much lower than previously thought; (ii) after adjusting for unrepeatable factors, the prospective premium is lower still at 3-3.5%; (iii) in every country equities outperformed bonds and bonds outperformed bills over the long term; (iv) equity risk does not appreciably decline with longer holding periods; (v) over the long run, purchasing power parity has held to a close approximation, making hedging unnecessary for long horizons; (vi) equities were a poor hedge against inflation, and bonds were the only hedge against deflation; (vii) there is no positive relation between GDP growth and equity returns, and so investor favouritism to fast-growing countries like China is mistaken; and (viii) whereas asset returns may show some weak tendency to mean revert, efforts to take advantage of this are likely to be counterproductive.

References to the research

"UK Financial Market Returns," Elroy Dimson and Paul Marsh, Journal of Business 74(1), January 2001, pp. 1-31. dx.doi.org/10.1086/209661

Linton

DT2001 said:

Linton said:

DT2001 said:

If you employ a strategy that requires no chance of downward adjustment of income you will be overly cautious.
Flexibility allows you to have a greater overall level of income.

If retiring before SPA without any guaranteed income maybe a combination of cash reserve and flexibility is ideal. It is the time when supposedly you’ll be spending most in retirement (unless needing care later on) but if you can tweak the timing of say long haul holidays by a year or so you must improve the longevity of your pot.

It isnt a matter of being cautious but rather of using your assets  to their best advantage.  In particular a 60/40 portfolio is seen as reasonable for retirement.  There should be plenty of money in the 40% to avoid selling equities for at least 5 years and for taking your long-booked luxury holiday.  Better to see the 40% as part of your armoury rather than simply as padding .  This implies that you should devote as much time to the allocation of your non-equity assets as you do to your equities.

I can't see that the savings from simply cutting excess expenditure for the short term without seriously reducing your standard of living are going to be significantly large to justify the pain. Rather than reacting in a panic to a crash by cutting all your expenditure by as much as say 25% for a year or two it would be better to plan to reduce expenditure by a few % for the rest of your life which is something you probably would not notice.  Such a small change would probably be reversed when your equities recovered anyway.

Sorry, I didn’t understand that your cash buffer was part of the 40%.

Guyton-Klinger would give you a better starting withdrawal rate which even after a 10% reduction would still be more than your ‘normal’ SWR.

If it is part of your plan to react to crashes there shouldn’t be a panic.

When I first calculated my ‘number’ I had 50% of my annual budget in my holiday fund so adjusting for prevailing market conditions wouldn’t worry me. Maybe having been self employed for 25 years with quite a variable income it is ‘normal’ for me to continually adjust expenditure without a feeling of lowering my living standards.

Why should the cash buffer not be part of the 40%?  It seems a great waste for the 40% just to be sitting there not doing very much.  Another part of my 40% is in corporate bonds providing income.  Every fund must have a positive purpose, there is no money for mere padding.  The 60/40 split is simply a measure of overall diversification.

I dont believe in SWRs other than to provide a sanity check, but rely more on a year by year spreadsheet with pessimistic assumed long term return, inflation rates, and normal living costs.  So Guyton Klinger is not applicable either. A measure of the risk is the amount of money left over at the end of plan.  Major one-off expenditure is justified by the proposed expenditure not reducing the safety margin to a worrying level.  Only known expenditures are explicitly planned.

OldScientist

FWIW, it might be interesting that holding bills (or cash either as easy access or in short term fixed) generally improved the historical SWR for a 30 year retirement in the UK when the SWR was small (in the bottom panel below, most of the SWRs below about 3.5% are higher with Bills than bonds), although, unfortunately, the worst case (1937) had the same outcome.

This also happens for most other countries in the dataset including the US (but not all, e.g. Germany, France, Italy, and Japan)

Dataset: JST (https://www.macrohistory.net) UK bond, bill and stock returns 1872-2015

In other words, holding short term (duration of a few years) fixed income often improved poor retirement periods (at the expense of making good retirement periods worse). According to Pfau ( https://retirementresearcher.com/4-rule-work-around-world/ ) “In terms of the SAFEMAX, 50% stocks and 50% bills generally outperforms 50% stocks and 50% bonds in the dataset. For bonds, the additional returns over bills failed to compensate for their additional volatility in a retirement income plan …”

However, this is not a separate cash buffer, but is cash held as part of the portfolio (so will be included in rebalancing) . Also worth noting that the future may be different from the past. Personally, we hold 50-60% of our fixed income in cash deposits (a small ladder of two year fixed rate accounts).

Audaxer

Thrugelmir said:

Audaxer said:

pensionpawn said:

Audaxer said:

pensionpawn said:

Linton said:

BritishInvestor said:

Audaxer said:

SouthCoastBoy said:

It seems most people take the swr as the way forward. My understanding of swr is that the pot will be preserved at the swr? Is that incorrect? 

Whether or not the pot will be preserved will depend on the Sequence of Returns. If for example you get a poor sequence of returns in the first decade of retirement, the pot might not last throughout retirement if you don't have a cash buffer to draw from in loss years.

I'm not sure the cash buffer is going to be a massive help (certainly not if you take it from the equity allocation).

https://finalytiq.co.uk/cash-reserve-buffers-withdrawal-rates-old-wives-fables-retirement-portfolio/

 - None of the strategies tested matched what it is proposed here that one does with a buffer. ie use it for covering crashes
 - The strategies were mostly tested when bond returns were of some significance - one of the major reasons for choosing a cash buffer is that safe bonds now seem to be sub-optimal.  If bond were returning say 5% the need for cash would be significantly reduced.
 - In retirement I would think most people are prepared to sacrifice returns if necessary to reduce risk

If having a large cash buffer enables you to sail through a crash worry-free when you would otherwise lie awake at night then a cash buffer is a massive help.

Essential for the early years however as time (decades) go by and your funds hopefully continue to grow significantly, you can start to whittle down the cash holding as with decreasing years ahead it has performed it's role.

If your investments grow significantly in retirement, I would say that is a good reason for converting more to cash to enable you to lower your risk profile and spend more of it if you want. As others have previously said, there is no point in being the richest person in the graveyard.

I think it depends on the value of your holding compared to what you need to withdraw pa. If you have £400k in funds and £100k in cash, are 65 and drawdown £20k per year it may be worth keeping your 5 year cash safety net. However if you now have £800k in funds and still £100k in cash at 75 you may well be asking what sort of catastrophic world event is going to leave you struggling to get by at 80 after you've used all your cash? Bearing in mind you have a full SP and say £0.5M equity in your house. What we don't spend from our drawdown each year will go into savings such as an ISA, however besides that I'm not going to significantly add to my cash buffer.

That's fine, but if your investments have grown to that extent, I'm not looking at it from the point of view of being more cautious. In your example, if your investments have gone up to £800k and you had £100k cash, you could if you wanted to, convert a bit more of the £800k to cash, say another £100k, to spend it on luxuries over the next few years or give some to good causes.  You would still have £700k in investments and £100k cash, which would still be enough to drawdown £20k per year for your remaining years. My point being, you can't take it with you.   

Periods of good returns are more often or not followed by famine. There are two types of investors those that have been humbled, and those that are yet to be. When markets are high it's easy to become intoxicated with the thought of having reached the mountain peak and standing there admiring the view. 

I know that, but there are also some of us that will just continue to accumulate in retirement, when in some cases spending rates could be comfortably increased. In the example given by the previous poster, £100k could have been removed from that portfolio, for example to fund a major house move, and there would still have been more than enough to fund the rest of their retirement. On the basis that the £20k annual income required was still under 3% of the remaining portfolio, that is a pretty safe withdrawal rate, especially with the addition of a 5 year cash buffer. 

Thrugelmir

Audaxer said:

Thrugelmir said:

Audaxer said:

pensionpawn said:

Audaxer said:

pensionpawn said:

Linton said:

BritishInvestor said:

Audaxer said:

SouthCoastBoy said:

It seems most people take the swr as the way forward. My understanding of swr is that the pot will be preserved at the swr? Is that incorrect? 

Whether or not the pot will be preserved will depend on the Sequence of Returns. If for example you get a poor sequence of returns in the first decade of retirement, the pot might not last throughout retirement if you don't have a cash buffer to draw from in loss years.

I'm not sure the cash buffer is going to be a massive help (certainly not if you take it from the equity allocation).

https://finalytiq.co.uk/cash-reserve-buffers-withdrawal-rates-old-wives-fables-retirement-portfolio/

 - None of the strategies tested matched what it is proposed here that one does with a buffer. ie use it for covering crashes
 - The strategies were mostly tested when bond returns were of some significance - one of the major reasons for choosing a cash buffer is that safe bonds now seem to be sub-optimal.  If bond were returning say 5% the need for cash would be significantly reduced.
 - In retirement I would think most people are prepared to sacrifice returns if necessary to reduce risk

If having a large cash buffer enables you to sail through a crash worry-free when you would otherwise lie awake at night then a cash buffer is a massive help.

Essential for the early years however as time (decades) go by and your funds hopefully continue to grow significantly, you can start to whittle down the cash holding as with decreasing years ahead it has performed it's role.

If your investments grow significantly in retirement, I would say that is a good reason for converting more to cash to enable you to lower your risk profile and spend more of it if you want. As others have previously said, there is no point in being the richest person in the graveyard.

I think it depends on the value of your holding compared to what you need to withdraw pa. If you have £400k in funds and £100k in cash, are 65 and drawdown £20k per year it may be worth keeping your 5 year cash safety net. However if you now have £800k in funds and still £100k in cash at 75 you may well be asking what sort of catastrophic world event is going to leave you struggling to get by at 80 after you've used all your cash? Bearing in mind you have a full SP and say £0.5M equity in your house. What we don't spend from our drawdown each year will go into savings such as an ISA, however besides that I'm not going to significantly add to my cash buffer.

That's fine, but if your investments have grown to that extent, I'm not looking at it from the point of view of being more cautious. In your example, if your investments have gone up to £800k and you had £100k cash, you could if you wanted to, convert a bit more of the £800k to cash, say another £100k, to spend it on luxuries over the next few years or give some to good causes.  You would still have £700k in investments and £100k cash, which would still be enough to drawdown £20k per year for your remaining years. My point being, you can't take it with you.   

Periods of good returns are more often or not followed by famine. There are two types of investors those that have been humbled, and those that are yet to be. When markets are high it's easy to become intoxicated with the thought of having reached the mountain peak and standing there admiring the view. 

I know that, but there are also some of us that will just continue to accumulate in retirement, when in some cases spending rates could be comfortably increased. In the example given by the previous poster, £100k could have been removed from that portfolio, for example to fund a major house move, and there would still have been more than enough to fund the rest of their retirement. On the basis that the £20k annual income required was still under 3% of the remaining portfolio, that is a pretty safe withdrawal rate, especially with the addition of a 5 year cash buffer. 

If equity markets falter and inflation persists. Then the overall return on the portfolio may disappoint.  Holding cash is no cert substitute for the positive returns generated by Government Bonds over the past 40 years. Seems to be more a case of investors convincing themselves it is.