Bernstein - Implementing Liability Matching and Risk Portfolios - in 2020

gm0

Just finishing Rational Expectations (William Bernstein) which was re-recommended in a recent Xmas gift book thread.  Very readable.

Thought experiment - reached the happy state of funds which "liability match" the retirement cashflow (income requirement after SP+ defined pensions) and a little over. 
But you still need to invest the liability matched portion for 40 years for a real return above zero (inflation matching) to die old and running out of pension pot at around the right time (say with a ~4 year 50% income buffer on a 95 death projection) - all that being so far out as to be risible in terms of projection accuracy.

Bernstein's book is very heavy on a message to lower your equity percentage in later life.  If you can - in this precise situation in fact.
He seems to be suggesting in practice that I take another ~30% equities out of the market. 
But where do i put it ?  In -ve yield world - how is this idea practically implementable to generate inflation returns + a little over for costs etc.
His stricture on not regarding corporate bonds as "riskless" assets would seem to kick a lot of bond fund options to the curb.

The Risk portfolio is easier.  He seems to suggest that the bit left over the Risk Portfolio can invested in equities to the benefit of heirs, charity, and taxes.
And that perhaps some cash like assets are also kept in the RP to be liquidated and invested into risk assets at major market events when opportunity arrives and the financial world is on fire.  A level of managed investible cash sits well with income cover for SORR in early retirement so I am more comfortable with this part - while acknowledging that there is money left on the table as well as risk.

Curious as to others' reactions to this book and practical solutions to the portfolio design challenge it presents in today's market for today's retirees.

Linton

After 15 years of retirement I am starting to take the same path,  without any help from Mr Bernstein as I havent read the book.  It seemed the logical thing to do since, assuming one can match inflation over the medium term, nothing more is required to last the rest of our days in comfort.
  
The first step is to cut back all the recent  increase in the Growth Porttfolio ("Risk Portfolio") .  The money is being added to the Wealth Preservation portfolio based on Troy Trojan, Capital Gearing Trust, and Jupiter Strategic Bonds and also increasing our cash reserves.  One doesnt need to go 100% riskless in the non-Risk/Growth portfolio, the significant cash holding can look after that requirement. And in my view 20%-40% equity is rather less risky over the medium term at the moment than 100% "safe" bonds.

I dont see any point in holding cash with the Risk Portfolio, much easier just to hold all cash centrally.  Keep the Risk Portfolio focussed on higher risk growth.

Thrugelmir

gm0 said:

His stricture on not regarding corporate bonds as "riskless" assets would seem to kick a lot of bond fund options to the curb.

Corporate bonds are highly correlated to shares. When a company goes "bust". You'll suffer the double whammy of a total write off in your investments. 

Linton

Thrugelmir said:

gm0 said:

His stricture on not regarding corporate bonds as "riskless" assets would seem to kick a lot of bond fund options to the curb.

Corporate bonds are highly correlated to shares. When a company goes "bust". You'll suffer the double whammy of a total write off in your investments. 

You can handle companies going bust simply through diversification - have a broad rajnge of corporate bonds and dont let corporate bonds as a whole have a dominant allocation in the portfolio..  More of an issue is large extended falls in equity value.  As long as the corporate bonds continue to pay interest  there is no great problem.

[Deleted User]

If a company was mismanaged and went bust then diversification between corporate bonds would mitigate your risk.  However in a major event if, say, insurance companies are going bust, then all corporate bonds would take a major hit.   And that would happen at the time when the equity is down. 

For various reasons we never quite got there in either 2009 or March 2020 but at one point it looked like we were moving in that direction.  The governments and central banks acted fast and in this case had enough resources.  

Someone in a withdrawal mode needs assets which are “safe “ and would offset some of the losses in this scenario. 

gm0

I bought it for the asset allocation education section. 

I was looking at the recommendations in the book and doing the thought experiment upon the basis provided - comparing my situation with his horror stories (essentially worst case deaccumulation SORR examples or behavioural errors around great depression, global financial crash etc.). 

This was interesting to contrast with the backtested drawdown mechanics and success probability, stress market MC simulation type of approach often to be found documented on ERN, Monevator, McClung book. i.e. hold lots of equities from 40%-50% to around 70-80% (not more due to isolated portfolio failures from SORR/volatility appearing in the data series for a given optimisation of WR), optionally have a cash buffer to suspend sales for 2-n stretch years (of a "normal" 7 year price recovery cycle for a medium size correction in early retirement - pre SP) this is more from tolerance/comfort in crisis than data driven.
Set WR low enough that the odds based on recorded history and stress testing look "OK".  And then try to forget about it until NRA and other guaranteed income cuts in.   And probably (for most cohorts - it will then be fine - and there might be an excess fund at the end (or income forgone).   But no guarantees.  

That sort of logic chain leads me to a much higher risk position with my pension portfolio (70-80%) than Bernstein is advocating.  He is essentially pushing to reduce materially below 50% - 20-30% or even derisked to ladders/pure FI and then for the equity risk portfolio portion he has views on a hierarchy of tilts to apply to a whole of global market approach (mainly a PoV on value and small but also some suggestions on amounts of PME and EM and REIT).

The book clearly acknowledges the aspect of financial risk capacity alongside liability matching + emotional tolerance.  If you have more guaranteed income, other sources of capital (cashflow) downsizing, inheritance then the options available change.  Freed by these options you could take more risk again - if you choose to "tolerate it". Or not. So the framework certainly supports a falling equity glidepath in retirement as unneeded risk is eliminated or a rising one (within increasing capacity to take it).

But the implication of the Bernstein book is that I should derisk my liability matched portfolio aggressively perhaps even materially below 40% equities.  Which is clearly at odds with the optimise drawdown writers.   I was playing the "what would I actually invest in game given -ve yields on gilts" in the scenario where I chose to follow this approach rather than a more aggressive one.

Thrugelmir

Linton said:

Thrugelmir said:

gm0 said:

His stricture on not regarding corporate bonds as "riskless" assets would seem to kick a lot of bond fund options to the curb.

Corporate bonds are highly correlated to shares. When a company goes "bust". You'll suffer the double whammy of a total write off in your investments. 

You can handle companies going bust simply through diversification - have a broad rajnge of corporate bonds and dont let corporate bonds as a whole have a dominant allocation in the portfolio..  More of an issue is large extended falls in equity value.  As long as the corporate bonds continue to pay interest  there is no great problem.

The result is an irreplacable loss of capital though. The pot becomes smaller. Dividend and interest returns both fall.  

Equities are a volatile investment. Always have been , always will be. An extended bull market has bred complancey into a generation of investors who know nothing different. 

[Deleted User]

But the implication of the Bernstein book is that I should derisk my liability matched portfolio aggressively perhaps even materially below 40% equities.  Which is clearly at odds with the optimise drawdown writers.   I was playing the "what would I actually invest in game given -ve yields on gilts" in the scenario where I chose to follow this approach rather than a more aggressive one.

Its been a while since I read it and all of his books are a bit mixed up in my head but I do not recall interpreting any of his suggestions in this manner.  

In fact, even his “low risk” or “cowards” portfolios have more equity than that. 

gm0

No - he really does Mordko - different chapters of different books for different purposes perhaps.  

Rational Expectations - Page 143.  He describes a Liability Matched Portfolio (made of TIPS. Vanilla Treasury's and CDs in US terms)  for the amount representing 25x RLE (Residual living expenses i.e. after guaranteed income) and then - *if* there are extra assets - a risk portfolio (this with equities in it alongside to your risk tolerance upon the assumption you have saved beyond LMP).  While I understand the "won the game already don't need to take any more dice rolls with the essential part logic" - it still seemed a somewhat extreme position to me for a book written in 2014.  He's not saying *everybody* should - just that some people are arguably forced by circumstances and others could make the choice because of the asymmetry in consequences vs actual objectives.

He defines RLE earlier in the book as the "needs" figure less state benefits and defined income (so more like "essential" income) in other writings.  So the "discretionary" vs "essential" gap is probably not available to shrink the LMP = 25x RLE value.  Though it clearly can make a big difference to the LMP number.  Depending upon assumptions used around longevity and contingencies thereon there could be risk asset boosting squeeze possible there as well while still cleaving to the LMP safe assets philosopy

There is a level of convergence with other writings such as discussions in other drawdown papers on buying or simulating some level of GI to a % of essential income level - be that 50% or 100% from SP, DB, tactical use of annuities or safe asset portfolios pretending to be annuities. All these techniques value putting a solid floor under variable income regardless of what transpires - at a considerable opportunity cost most of the time.

Overall the book section is a discussion about how hard it can be to save this amount in the first place.  And the whole piece is framed in an anecdote with tables about how the onset and aftermath of the Great Depression differently affected the 25/45/65 aged cohorts. 

Elsewhere in achieving required savings rates during accumulation he clearly makes the point that reliance on FI alone clearly won't cut it and some kind of risk must be taken.

itwasntme001

Linton said:

Thrugelmir said:

gm0 said:

His stricture on not regarding corporate bonds as "riskless" assets would seem to kick a lot of bond fund options to the curb.

Corporate bonds are highly correlated to shares. When a company goes "bust". You'll suffer the double whammy of a total write off in your investments. 

You can handle companies going bust simply through diversification - have a broad rajnge of corporate bonds and dont let corporate bonds as a whole have a dominant allocation in the portfolio..  More of an issue is large extended falls in equity value.  As long as the corporate bonds continue to pay interest  there is no great problem.

Are you really "that" diversified by owning a broad range of corporate bonds?  Don't corporate bond issuers generally come from a group of smaller number of sectors compared to an equity index?  Aren't the issuers usually the type of companies you may not really want to invest in due to the leverage (= risky)?  Is the correlation of returns for a "diversified" portfolio of corporate bonds really closer to 0 than it is to 1?

Has QE distorted the price of corporate bonds so much that they simply can not be included in any portfolio given poor risk reward?  The FED have been active buyers of corporate bonds.

itwasntme001

I haven't read this Berntein book.  I suspect that asset allocation strategies for retirees were based on historical returns?  How far back does this historical analysis go?