SWR Question

michaels

Back to the question of when to take a DB - lets assume a fully indexed DB to avoid questions around max indexation clouding the picture.

For me the question is how much the 'bridge' will cost.  If this period is say 12 years then you are in a no mans land situation where you are still at considerable risk of inflation but for inflation protection you need equities where 12 years leaves you with considerable volatility risk.  Thus your only 'safe' option is index linked govt bonds which generally means locking in 0% or less real growth over that period.

Thus if you take the DB early you reduce the amount of gap needed to be filled by 'short term safe' investments and increase the amount of your DC that can be invested longer term for (hopefully) above inflation returns thus generating on average a higher return on the DC component.  Is this incorrect?

Pat38493

michaels said:

Back to the question of when to take a DB - lets assume a fully indexed DB to avoid questions around max indexation clouding the picture.

For me the question is how much the 'bridge' will cost.  If this period is say 12 years then you are in a no mans land situation where you are still at considerable risk of inflation but for inflation protection you need equities where 12 years leaves you with considerable volatility risk.  Thus your only 'safe' option is index linked govt bonds which generally means locking in 0% or less real growth over that period.

Thus if you take the DB early you reduce the amount of gap needed to be filled by 'short term safe' investments and increase the amount of your DC that can be invested longer term for (hopefully) above inflation returns thus generating on average a higher return on the DC component.  Is this incorrect?

This is certainly along the lines that I am thinking.

Another point is that if I am in my 50s and early 60s, I can always take on some work to help with any problems.

The other context that I should probably disclose is that when I look at my household as a couple, we are in the very fortunate position that if I take the DB late, by the time I am 67 as the younger partner we will have about 75K (real terms) of guaranteed income which is obviously more than enough to cover essential expenses and even all of my current estimate of money to maintain same lifestyle.

Even taking the DB early, our joint guaranteed income when I am 67 covers almost 90% of the lifestyle maintain budget.

The only thing that sometimes nags at me is that even these figures that most people can't dream of, it's still only about 40% of what was our gross income together when we were both working full time (I guess we are in the top 1% that we hear so much about, or almost so).  The numbers seem to work though if I look at it on net and not gross levels.

My wife is one of the lucky few to be on a full NHS MHO pension which I think (I would need to check) is uncapped.  As mentioned above my DB is 5% RPI capped.

I am still collecting detailed data on my spending to prepare for retirement (7 months of data so far so I have to annualise it), so it's not yet exact but it's probably not far off.

It's in that context that I might seem a bit gung ho about taking the DB early and "gambling" my DC fund over a longer period in the hope of coming out on top, or even in very optimistic scenarios becoming rather rich and having money to pass on to kids etc.  Even if my DC is exhausted by the bridging process I am still not going to starve, but I will miss out on the potential opportunity of a good returns if I am lucky with SOR and so on.

[Deleted User]

michaels said:

Pat38493 said:

michaels said:

I find this tool useful as you can put in all sorts of other cashflows, lump sums etc and it tells you what your SWR (100%, 99%, 95% etc no failures) and annual DC withdrawals for a length of time you choose (I tend to use to age 95) would have been given your chosen asset split over the last 120 odd years.

An Updated Google Sheet DIY Withdrawal Rate Toolbox (SWR Series Part 28) – Early Retirement Now

I was looking at this sheet this evening.  On the first sheet what is the meaning of the sections called “safe cons amounts conditional on S&P 500 drawdown”, with various columns for different % levels.  Is this relating somehow to the market performance during the start year or suchlike?

I think it is suggesting that if you start your drawdown during a period when the S&P has fallen back from its peak level then historically you would have been able to withdraw slightly more  (logically it sounds reasonable and helps with the situation where if you retire in a year when markets are high your big pot gives a high annual amount at a given SWR whereas if you retire a year later when markets are say down 20% then your safe withdrawal amount based on a percentage alone is also 20% smaller - ie if markets are not as 'high' then they are likely to perform 'better' going forward - of course it contradicts efficient market theory)

The problem is that “peaks” are only ever identifiable after falls. One could be waiting for a 20% drawdown for 10 years. 

Pat38493

Deleted_User said:

michaels said:

Pat38493 said:

michaels said:

I find this tool useful as you can put in all sorts of other cashflows, lump sums etc and it tells you what your SWR (100%, 99%, 95% etc no failures) and annual DC withdrawals for a length of time you choose (I tend to use to age 95) would have been given your chosen asset split over the last 120 odd years.

An Updated Google Sheet DIY Withdrawal Rate Toolbox (SWR Series Part 28) – Early Retirement Now

I was looking at this sheet this evening.  On the first sheet what is the meaning of the sections called “safe cons amounts conditional on S&P 500 drawdown”, with various columns for different % levels.  Is this relating somehow to the market performance during the start year or suchlike?

I think it is suggesting that if you start your drawdown during a period when the S&P has fallen back from its peak level then historically you would have been able to withdraw slightly more  (logically it sounds reasonable and helps with the situation where if you retire in a year when markets are high your big pot gives a high annual amount at a given SWR whereas if you retire a year later when markets are say down 20% then your safe withdrawal amount based on a percentage alone is also 20% smaller - ie if markets are not as 'high' then they are likely to perform 'better' going forward - of course it contradicts efficient market theory)

The problem is that “peaks” or only ever identifiable after falls. One could be waiting for a 20% drawdown for 10 years. 

Yes this is what I was also thinking - this might be the case but you can never know for sure which situation you are in with hindsight and/or even if you want to risk this, you would have to be prepared to trigger your retirement according to market timing rather than when you actually want to stop working.

On another side, the spreadsheet posted by michaels is a completely different proposal in that it seems to propose that you judge your SWR at the time you start your retirement journey, and then you courageously stick to it throughout on the basis that history shows this will work out ok in the end.

The argument against this as pointed out earlier is that most of these models look at about 130 years of history, and even within that some of the data for some categories is not perfect, so you are only looking at a few discrete clear sequences of retirement.  However their argument is that it's valid to look at overlappiong sequences for any start month within that time.

I suppose you have to keep an eye on whether the market starts to do something it has never done before.   

[Deleted User]

michaels said:

Back to the question of when to take a DB - lets assume a fully indexed DB to avoid questions around max indexation clouding the picture.

For me the question is how much the 'bridge' will cost.  If this period is say 12 years then you are in a no mans land situation where you are still at considerable risk of inflation but for inflation protection you need equities where 12 years leaves you with considerable volatility risk.  Thus your only 'safe' option is index linked govt bonds which generally means locking in 0% or less real growth over that period.

Thus if you take the DB early you reduce the amount of gap needed to be filled by 'short term safe' investments and increase the amount of your DC that can be invested longer term for (hopefully) above inflation returns thus generating on average a higher return on the DC component.  Is this incorrect?

There are different scenarios, depending on the proportion of your needs covered from DB vs DC.The general rule is that as long as you can afford delaying DB (or state pension), you should.  Thats assuming your real objective is to derisk retirement and withdraw more SAFELY vs to maximize potential returns.  

In your scenario you want to take a bet on your particular  mixture of stocks and bonds beating the guaranteed returns on inflation linked gilts.  That’s far from guaranteed. An acquaintance retired in the 90s being fully invested in the  latter and his returns so far beat any possible mixture of stocks and bonds. Crucially, you are taking on unnecessary risk.

If you have to wait for 10 years before getting unreduced DB, my approach would be to have 5 years’ worth of guaranteed income in cash-like investments, including bonds of appropriate duration.  And then the rest could be kept invested with withdrawals from the investment portfolio made annually to cover 12 month expenditure.  

Linton

Pat38493 said:

Deleted_User said:

michaels said:

Pat38493 said:

michaels said:

I find this tool useful as you can put in all sorts of other cashflows, lump sums etc and it tells you what your SWR (100%, 99%, 95% etc no failures) and annual DC withdrawals for a length of time you choose (I tend to use to age 95) would have been given your chosen asset split over the last 120 odd years.

An Updated Google Sheet DIY Withdrawal Rate Toolbox (SWR Series Part 28) – Early Retirement Now

I was looking at this sheet this evening.  On the first sheet what is the meaning of the sections called “safe cons amounts conditional on S&P 500 drawdown”, with various columns for different % levels.  Is this relating somehow to the market performance during the start year or suchlike?

I think it is suggesting that if you start your drawdown during a period when the S&P has fallen back from its peak level then historically you would have been able to withdraw slightly more  (logically it sounds reasonable and helps with the situation where if you retire in a year when markets are high your big pot gives a high annual amount at a given SWR whereas if you retire a year later when markets are say down 20% then your safe withdrawal amount based on a percentage alone is also 20% smaller - ie if markets are not as 'high' then they are likely to perform 'better' going forward - of course it contradicts efficient market theory)

The problem is that “peaks” or only ever identifiable after falls. One could be waiting for a 20% drawdown for 10 years. 

Yes this is what I was also thinking - this might be the case but you can never know for sure which situation you are in with hindsight and/or even if you want to risk this, you would have to be prepared to trigger your retirement according to market timing rather than when you actually want to stop working.

On another side, the spreadsheet posted by michaels is a completely different proposal in that it seems to propose that you judge your SWR at the time you start your retirement journey, and then you courageously stick to it throughout on the basis that history shows this will work out ok in the end.

The argument against this as pointed out earlier is that most of these models look at about 130 years of history, and even within that some of the data for some categories is not perfect, so you are only looking at a few discrete clear sequences of retirement.  However their argument is that it's valid to look at overlappiong sequences for any start month within that time.

I suppose you have to keep an eye on whether the market starts to do something it has never done before.   

Indeed, but there is a much more important factor.  The SWR assumes the worst case. For most people most of the time actual investment returns will be higher than that.  So within say 5 years a recalculated SWR is likely to be larger because you have more money and there is less time to run out of money.

If you look at the SWR graphs you will see that if dont run out of money there is a high chance of ending up richer than when you started retirement.

So if you must use SWR I suggest y9ou replan on a regular basis.

Pat38493

Deleted_User said:

michaels said:

Back to the question of when to take a DB - lets assume a fully indexed DB to avoid questions around max indexation clouding the picture.

For me the question is how much the 'bridge' will cost.  If this period is say 12 years then you are in a no mans land situation where you are still at considerable risk of inflation but for inflation protection you need equities where 12 years leaves you with considerable volatility risk.  Thus your only 'safe' option is index linked govt bonds which generally means locking in 0% or less real growth over that period.

Thus if you take the DB early you reduce the amount of gap needed to be filled by 'short term safe' investments and increase the amount of your DC that can be invested longer term for (hopefully) above inflation returns thus generating on average a higher return on the DC component.  Is this incorrect?

There are different scenarios, depending on the proportion of your needs covered from DB vs DC.The general rule is that as long as you can afford delaying DB (or state pension), you should.  Thats assuming your real objective is to derisk retirement and withdraw more SAFELY vs to maximize potential returns.  

In your scenario you want to take a bet on your particular  mixture of stocks and bonds beating the guaranteed returns on inflation linked gilts.  That’s far from guaranteed. An acquaintance retired in the 90s being fully invested in the  latter and his returns so far beat any possible mixture of stocks and bonds. Crucially, you are taking on unnecessary risk.

If you have to wait for 10 years before getting unreduced DB, my approach would be to have 5 years’ worth of guaranteed income in cash-like investments, including bonds of appropriate duration.  And then the rest could be kept invested with withdrawals from the investment portfolio made annually to cover 12 month expenditure.  

Two comments / questions

- Isn't taking DB unreduced somewhat an arbitrary concept - you can also defer it to get even more in roughly the same proportions, so it's more a case, of, when is the best time to take it?

- Is the concept you describe for 5 years worth of cash what is known as a "fixed savings ladder" - I saw that mentioned elsewhere and I was wondering if this is what it means?

michaels

Deleted_User said:

michaels said:

Pat38493 said:

michaels said:

I find this tool useful as you can put in all sorts of other cashflows, lump sums etc and it tells you what your SWR (100%, 99%, 95% etc no failures) and annual DC withdrawals for a length of time you choose (I tend to use to age 95) would have been given your chosen asset split over the last 120 odd years.

An Updated Google Sheet DIY Withdrawal Rate Toolbox (SWR Series Part 28) – Early Retirement Now

I was looking at this sheet this evening.  On the first sheet what is the meaning of the sections called “safe cons amounts conditional on S&P 500 drawdown”, with various columns for different % levels.  Is this relating somehow to the market performance during the start year or suchlike?

I think it is suggesting that if you start your drawdown during a period when the S&P has fallen back from its peak level then historically you would have been able to withdraw slightly more  (logically it sounds reasonable and helps with the situation where if you retire in a year when markets are high your big pot gives a high annual amount at a given SWR whereas if you retire a year later when markets are say down 20% then your safe withdrawal amount based on a percentage alone is also 20% smaller - ie if markets are not as 'high' then they are likely to perform 'better' going forward - of course it contradicts efficient market theory)

The problem is that “peaks” or only ever identifiable after falls. One could be waiting for a 20% drawdown for 10 years. 

Looking at the last 12 months, 12 months ago the S&P was at an all time max so on the table you could only choose the lowest SWR rate.  12 months later it is down 20% so in theory you could be safe with the slightly higher SWR rate.  This higher rate goes some way to make up for the fact that your pot is now 20% (depending on asset mix) smaller than it was 12 months ago so your actual SW currency amount does not reduce by as much as a strict SWR rule would imply.

It is not about waiting for a period of market retenchment so you can risk a higher withdrawal rate but more suggesting that if the market has fallen the SWR is higher but this will not go as far as meaning the swr amount does not decline when markets fall.

Of course this is all about the 'traditional' definition of SWR - the highest (fixed, inflation adjusted) withdrawal rate that would not have resulted in an unacceptable number of failure scenarios based on historic actuals.

Albermarle

Isn't taking DB unreduced somewhat an arbitrary concept - you can also defer it to get even more in roughly the same proportions, so it's more a case, of, when is the best time to take it?

With some DB pensions you can defer them past the NRA and get an increase each year ( on top of any indexation increases) . However AFAIUI this is not the norm, and you usually lose out by deferring them past the NRA.

So for most the choice is taking at NRA, or a reduced pension before NRA.

[Deleted User]

michaels said:

Deleted_User said:

michaels said:

Pat38493 said:

michaels said:

I find this tool useful as you can put in all sorts of other cashflows, lump sums etc and it tells you what your SWR (100%, 99%, 95% etc no failures) and annual DC withdrawals for a length of time you choose (I tend to use to age 95) would have been given your chosen asset split over the last 120 odd years.

An Updated Google Sheet DIY Withdrawal Rate Toolbox (SWR Series Part 28) – Early Retirement Now

I was looking at this sheet this evening.  On the first sheet what is the meaning of the sections called “safe cons amounts conditional on S&P 500 drawdown”, with various columns for different % levels.  Is this relating somehow to the market performance during the start year or suchlike?

I think it is suggesting that if you start your drawdown during a period when the S&P has fallen back from its peak level then historically you would have been able to withdraw slightly more  (logically it sounds reasonable and helps with the situation where if you retire in a year when markets are high your big pot gives a high annual amount at a given SWR whereas if you retire a year later when markets are say down 20% then your safe withdrawal amount based on a percentage alone is also 20% smaller - ie if markets are not as 'high' then they are likely to perform 'better' going forward - of course it contradicts efficient market theory)

The problem is that “peaks” or only ever identifiable after falls. One could be waiting for a 20% drawdown for 10 years. 

Looking at the last 12 months, 12 months ago the S&P was at an all time max so on the table you could only choose the lowest SWR rate.  12 months later it is down 20% so in theory you could be safe with the slightly higher SWR rate.  This higher rate goes some way to make up for the fact that your pot is now 20% (depending on asset mix) smaller than it was 12 months ago so your actual SW currency amount does not reduce by as much as a strict SWR rule would imply.

It is not about waiting for a period of market retenchment so you can risk a higher withdrawal rate but more suggesting that if the market has fallen the SWR is higher but this will not go as far as meaning the swr amount does not decline when markets fall.

Of course this is all about the 'traditional' definition of SWR - the highest (fixed, inflation adjusted) withdrawal rate that would not have resulted in an unacceptable number of failure scenarios based on historic actuals.

End of Dec 2020 S&P500 traded at an all time high around 3800.  Today its trading around 4000, way off highs of about 4800 reached in December 2021.  Based on your approach a 60 year old starting retirement today can safely withdraw a lot more for life than his counterpart 2 years ago. Right?