Bernstein - Implementing Liability Matching and Risk Portfolios - in 2020

Prism

Audaxer said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

I agree with your comments on just going for a particular equity/bond mix.  In retirement you have specific needs which cannot be adequately covered by a simple 20/40/60/80 % bond allocation chosen on your psychological attitude to risk.

Linton, I know that a straightforward 60/40 globally diversified equity bond portfolio will not suit all retirees, but for someone looking for a withdrawal rate of around 3.5% per annum increasing with inflation, is there any reason why you would consider their specific needs would not be covered by such a portfolio?

For me I am not aware of a time that there has been such a situation where there has been such a high equity price at the same time there was such a low interest rate (and high bond price). The previous two bad periods for initial retirement, 1929 and 1969 started from a base of around 4-5% interest rates. I would be worried that a 60/40 might drop both equities and bonds if we encountered anything similar to 1969 again. Hopefully, a bit like in that period with Japanese equities rising another country would step in with a big growth story next time - maybe China or India.

I am hoping that by the time I retire the interest rates have recovered a least a bit.

[Deleted User]

Linton said:

The problem with Deleted_User said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

Neither do bonds except for Index Linked bonds for partial protection against well above currently normal inflation levels.  However "safe" bonds carry the risk of large reductions in capital value once interest rates rise.  The only viable inflation protection available at the moment sadly is equity or an inflation linked annuity. 

1. Returns on bond funds suffer in cases of unexpected jumps in inflation. Expected inflation has no impact. Conversely, shares suffer during deflation.
2. There are other tools which protect from inflation. For example, rate reset preferred shares. 

I have never come across rate reset preferred shares, they seem to be a US facility.

I hold Canadian prefs.  Wouldn’t normally but they were mispriced in March. Happens to niche products.   Brits call them “preferential”.  Not sure if there are any resets issued by the UK companies or if they are all perpetual. Its a tricky instrument. Described in the Intelligent Investor.  

Linton

Deleted_User said:

Linton said:

The problem with Deleted_User said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

Neither do bonds except for Index Linked bonds for partial protection against well above currently normal inflation levels.  However "safe" bonds carry the risk of large reductions in capital value once interest rates rise.  The only viable inflation protection available at the moment sadly is equity or an inflation linked annuity. 

1. Returns on bond funds suffer in cases of unexpected jumps in inflation. Expected inflation has no impact. Conversely, shares suffer during deflation.
2. There are other tools which protect from inflation. For example, rate reset preferred shares. 

I have never come across rate reset preferred shares, they seem to be a US facility.

I hold Canadian prefs.  Wouldn’t normally but they were mispriced in March. Happens to niche products.   Brits call them “preferential”.  Not sure if there are any resets issued by the UK companies or if they are all perpetual. Its a tricky instrument. Described in the Intelligent Investor.  

From my limited knowledge all preference shares in the UK are perpetual.  In general they were issued many years ago and are a very niche area - a quick Google says there are:  "only 100 still in existance" and that there are liquidity constraints so they arent really relevent to retirement finance 

[Deleted User]

Linton said:

Deleted_User said:

Linton said:

The problem with Deleted_User said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

Neither do bonds except for Index Linked bonds for partial protection against well above currently normal inflation levels.  However "safe" bonds carry the risk of large reductions in capital value once interest rates rise.  The only viable inflation protection available at the moment sadly is equity or an inflation linked annuity. 

1. Returns on bond funds suffer in cases of unexpected jumps in inflation. Expected inflation has no impact. Conversely, shares suffer during deflation.
2. There are other tools which protect from inflation. For example, rate reset preferred shares. 

I have never come across rate reset preferred shares, they seem to be a US facility.

I hold Canadian prefs.  Wouldn’t normally but they were mispriced in March. Happens to niche products.   Brits call them “preferential”.  Not sure if there are any resets issued by the UK companies or if they are all perpetual. Its a tricky instrument. Described in the Intelligent Investor.  

From my limited knowledge all preference shares in the UK are perpetual.  In general they were issued many years ago and are a very niche area - a quick Google says there are:  "only 100 still in existance" and that there are liquidity constraints so they arent really relevent to retirement finance 

Yes, there are liquidity constraints in Canada too.  Which is why they can be mispriced. Which could create a problem. Or an opportunity.  They require a lot of effort, not for the mass market but worked well for my retirement finance. 

Audaxer

Linton said:

Audaxer said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

I agree with your comments on just going for a particular equity/bond mix.  In retirement you have specific needs which cannot be adequately covered by a simple 20/40/60/80 % bond allocation chosen on your psychological attitude to risk.

Linton, I know that a straightforward 60/40 globally diversified equity bond portfolio will not suit all retirees, but for someone looking for a withdrawal rate of around 3.5% per annum increasing with inflation, is there any reason why you would consider their specific needs would not be covered by such a portfolio?

My comment was mainly about the use of packaged multi-asset funds for a retirement portfolio with the specific need in this case being risk managment. You need to be able to control the equity and the non-equity portions separately.

As I have said many times, I am nervous about the over-use of safe bond funds in the non-equity portion. For the past 40 years interest rates have been falling whch has led to a long term increase in bond fund capital values.  Now we are close to zero interest, what happens in the medium term future? Also, again as I said in a previous post, the non-equity needs to have an objective.  Merely being non-equity is not sufficient.

It makes sense to manage it jointly with the buffer and to devote at least as much  attention to diversifying its holdings as to the equity side especially as I see configuring non-equity as a more difficult task than equity where a single global tracker could be sufficient.  LMP seems to be operating in this area but I do not understand how what is being proposed there works in a close to zero interest world.

With a strong set of non-equity holdings securing the short term and to some extent the medium term you can look at taking more risk with the equity side than would otherwise be the case.

I use the terms equity and non equity, it would be clearer and more accurate if I had talked about "at risk" and "low risk", or something like that.

Thanks for your response. I agree that it could be beneficial to control the equity and non-equity sides of the portfolio separately, and you don't have that with a multi asset fund. However I also agree that configuring the non-equity is a much more difficult task, and that is where I think for the less experienced investors a multi asset fund can be simpler with less chance of getting the non-equity side wrong. I still think with a 60/40 multi asset fund, a 3.5% withdrawal rate along with a cash buffer would work in all but the very worse sequence of returns.

gm0

Thanks to JohnWinder for your expanded answer on LMP options.  And to everyone else for your interesting input and commentary on specific options in the non-equities area.

My own thinking has not moved far from dealing with our stockmarket risk profile by buffering early on for ability to pause/ or reduce income drawn for SORR during 55-67 and then a Guaranteed Income floor approach later.  No annuity - just 2xSP + 1x small DB spouse pension).

Calculating LMP though has its uses in quantifying about how you feel about your strategy and asset allocation - although there are clearly other ways to do that - ulcerindex etc.

[Deleted User]

Canada has its own versions of Bernstein. Dan Bortolotti and Justin Bender are finance professionals who have made massive amount of high quality information available to Canadian investors, including portfolio models and analysis. They publish blogs, videos and White Papers, analyze what’s available on the market and provide guidance on minimizing taxation when investing for retirement. 

Is there anyone like that in the UK? 

itwasntme001

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

Neither do bonds except for Index Linked bonds for partial protection against well above currently normal inflation levels.  However "safe" bonds carry the risk of large reductions in capital value once interest rates rise.  The only viable inflation protection available at the moment sadly is equity or an inflation linked annuity. The latter has the major disadvantage of being totally inflexible.  Of course there are physical assets like gold, fine art etc which should in the long term match inflation but their volatility would rule them out in my view.

My own strategy is I think similar to LMP, though I had never heard of LMP before this thread, For the inflation matching I rely on Wealth Preservation funds, hoping a manager with that objective can balance equity and other investments better than I can or have time to do.  This is coupled with a highly diversified equity/bond/others Income portfolio to provide a somewhat reliable income despite changing capital values.

I agree with your comments on just going for a particular equity/bond mix.  In retirement you have specific needs which cannot be adequately covered by a simple 20/40/60/80 % bond allocation chosen on your psychological attitude to risk.

I would not count on equities protecting against unexpected inflation.  Historically equities did well at inflation rates around 3-4% but above that equities suffered quite badly with the 1970s being an extreme example.

The problem now is that it is possible for equities to do badly at lower rates of inflation.  We saw this in March 2018.  When you have rates so low as they are now, any surprise up tick in inflation will cause bond yields to rise and therefore stocks fall, if there is not enough nominal growth (i.e. stagflation).  Stocks are long duration assets, particularly for markets with high PE multiples such as the US.  So stocks will be more sensitive to rate moves.

[Deleted User]

itwasntme001 said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

Neither do bonds except for Index Linked bonds for partial protection against well above currently normal inflation levels.  However "safe" bonds carry the risk of large reductions in capital value once interest rates rise.  The only viable inflation protection available at the moment sadly is equity or an inflation linked annuity. The latter has the major disadvantage of being totally inflexible.  Of course there are physical assets like gold, fine art etc which should in the long term match inflation but their volatility would rule them out in my view.

My own strategy is I think similar to LMP, though I had never heard of LMP before this thread, For the inflation matching I rely on Wealth Preservation funds, hoping a manager with that objective can balance equity and other investments better than I can or have time to do.  This is coupled with a highly diversified equity/bond/others Income portfolio to provide a somewhat reliable income despite changing capital values.

I agree with your comments on just going for a particular equity/bond mix.  In retirement you have specific needs which cannot be adequately covered by a simple 20/40/60/80 % bond allocation chosen on your psychological attitude to risk.

I would not count on equities protecting against unexpected inflation.  Historically equities did well at inflation rates around 3-4% but above that equities suffered quite badly

Bernstein disagrees.

itwasntme001

Prism said:

Audaxer said:

Linton said:

JohnWinder said:

That approach of Bernstein's will be suitable for some financial circumstances and personalities. There are others pushing the same idea, but we hear of few people who put it into practice, so its appeal seems narrow, despite it's apparent good sense (as you'd expect from Bernstein).

Zwelcher includes it in his book (but regret I can't remember the book title). He talks about the topic in terms of an 'income floor' - no matter how badly the investment world turns into, your assets provide a 'guaranteed' income floor such as you'd get from a lifetime annuity. His view is that, along the lines of the '4% rule', if your SWR is less than 3.5% then building an income floor with a LMP is not so necessary. But if your assets aren't adequate for that, he suggests putting some of them at 'high risk' as a way of building up a 'safe floor' to an adequate level, even if you need to put 'stop loss' limits on those investments to protect your prospective floor from finishing up too low.

Lussier in his book 'Successful investing is a process' likes the LMP approach despite its low expected returns, but he 'tweaks' it by suggesting having some assets in LMP and some in an at-risk portfolio (the mix depending on risk aversion and how well covered the liabilities are).  He proposes a dynamic management of all the assets, moving from the at-risk to the LMP depending on value measures like PE10 and yield. He also likes to track asset class volatility and correlations and make AA adjustments in the at-risk portfolio accordingly. That seemed a bit complicated to me, so I wave it away.

Huxley in 'Asset Dedication' suggests a 'watered-down' LMP by buying a non-rolling bond ladder for the first 5 (or more) years  of retirement, and puts the rest in shares.  This avoids a bad SoR (as long as it lasts <5 yrs!). The ladder amount must anticipate inflation (or use an inflation linked product), and you extend it for the next 5 yrs either at the end of the first five or yearly, or after there’s a good year or two for shares. He says this approach is better than trying to match your risk aversion with a stock/bond mix, there being no quantitative way to do this; I think that's the first time I'd read that, and it had a ring of common sense to it. How do 'they' convert your risk questionnaire score of 'x out of 100' into a 60/40 mix, in any validated way?

I think Bernstein and others suggests the following assets for a LMP: a non-rolling bond ladder; a lifetime inflation linked annuity; and long term health care insurance. But he's American so you won't need the last one.

While interest rates are low, lifetime annuities seem expensive and there may not be many products on offer. But that's the nature of these times, so you just put up with it if you want a LMP.  Exactly the same consideration applies to a non-rolling ladder of inflation linked government bonds: yields are negative, so you'll pay more for them than you'll get back in the end; but that's the cost of having a LMP which is as near as possible to a bullet-proof investment as you can do. It doesn't seem quite so bad when the alternative is a 'risk portfolio' of stocks and bonds, both having high prices and low yields from quantitative easing.

Cash is probably not a good choice for a LMP as it has no secured inflation protection; and a twenty five year retirement needs inflation protection.

I agree with your comments on just going for a particular equity/bond mix.  In retirement you have specific needs which cannot be adequately covered by a simple 20/40/60/80 % bond allocation chosen on your psychological attitude to risk.

Linton, I know that a straightforward 60/40 globally diversified equity bond portfolio will not suit all retirees, but for someone looking for a withdrawal rate of around 3.5% per annum increasing with inflation, is there any reason why you would consider their specific needs would not be covered by such a portfolio?

For me I am not aware of a time that there has been such a situation where there has been such a high equity price at the same time there was such a low interest rate (and high bond price). The previous two bad periods for initial retirement, 1929 and 1969 started from a base of around 4-5% interest rates. I would be worried that a 60/40 might drop both equities and bonds if we encountered anything similar to 1969 again. Hopefully, a bit like in that period with Japanese equities rising another country would step in with a big growth story next time - maybe China or India.

I am hoping that by the time I retire the interest rates have recovered a least a bit.

This is why there is a stronger argument of moving assets to wealth preservation funds which are examples of all weather portfolio (although not exactly the same).

It is also an argument for having your equity allocation in global tracker funds as opposed to regional or managed.  Missing a region/sector/style over a rapidly changing macro environment can have a detrimental impact on your equity allocation's performance.