Thinking of SS as a fixed income portion of portfolio

[rmark]

'COLA'd Pension' is what makes it equivalent (assuming a portfolio with adequate growth to offset inflation).

 

[Chuckanut]

'COLA'd Pension' is what makes it equivalent (assuming a portfolio with adequate growth to offset inflation).

Very few people have a fully COLA'd pension. Mine has a COLA cap. If we get into years of high-inflation it will be time for Plan B.

 

[Mulligan]

Very few people have a fully COLA'd pension. Mine has a COLA cap. If we get into years of high-inflation it will be time for Plan B.

Mine is capped at 80% of initial pension. If years of high inflation occur in next 10 years, plan B might mean going back to work. If its later on, I will be fine.

 

[photoguy]

On one hand, I say it does not matter. Consider this:

Person A: Plans to spend $40,000, inflation adjusted each year. He gets $20,000 from COLA'd Pension/SS. He draws an initial 3.5% from his portfolio (of ~ $570,000) for the other $20,000.

Person B: Plans to spend $40,000, inflation adjusted each year. He has no Pension/SS. He draws an initial 3.5% from his portfolio (of ~ $1,140,000) for the full $40,000.

A 3.5% WR from a portfolio is a 3.5% WR - period. The historical optimum AA has to be the same for each, how can it be different?

-ERD50

Even though the withdrawal rate is the same as a percentage of the portfolio, the consequences of failure is not the same in case A and B. Furthermore portfolio A is much less inflation sensitive (since half of it is COLA).

 

[ERD50]

Even though the withdrawal rate is the same as a percentage of the portfolio, the consequences of failure is not the same in case A and B. Furthermore portfolio A is much less inflation sensitive (since half of it is COLA).

Oh, they are clearly different. Other wise, I would have made them the same. But that makes for a lousy comparison.

But regardless, there is only one historically optimum portfolio AA for a given WR and time frame. The portfolio does not 'know' you have a COLA'd SS/pension!

For a wide range of scenarios, the optimum AA is often pretty flat from 50/50 to 90/10. Increasing Equities will likely increase volatility (and in some cases fail), increasing Fixed will likely not provide a high enough real return to survive the worst cases.

The question isn't whether Person A or B will be affected more/less by volatility or inflation - that will be different because they have different income sources. The question is which AA has historically survived over that time period. Neither want their portfolio to fail. And that AA is the same for each, because the factors are the same - only the amounts are different, and that does not matter.

-ERD50

 

[photoguy]

But regardless, there is only one historically optimum portfolio AA for a given WR and time frame. The portfolio does not 'know' you have a COLA'd SS/pension!

I think our disagreement is coming from the fact that we don't have the same definition of "optimum". If I had a COLA'd annuity that covered most of my expenses, my "optimum" portfolio would be one that had the greatest chance of being a moonshot and I would care much less about failure. On the other hand without any sort of SS/annuity/pension I would weight portfolio failure much higher in my definition of optimum (or objective function).

 

[ERD50]

I think our disagreement is coming from the fact that we don't have the same definition of "optimum". If I had a COLA'd annuity that covered most of my expenses, my "optimum" portfolio would be one that had the greatest chance of being a moonshot and I would care much less about failure. On the other hand without any sort of SS/annuity/pension I would weight portfolio failure much higher in my definition of optimum (or objective function).

Like I said, there are differences between the two. I don't think we are disagreeing on that.

Someone with a 1/2 their expenses covered by COLA'd SS/pension may decide they want to 'gamble' and hope for that moonshot and may not care so much about failure. Another person in that situation may decide that with such a nice buffer, why gamble with it at all?

That's a personal choice. But neither has any bearing on the historical optimum AA for a given WR and time frame. With 'optimum' defined in the common way - "not running out in my life time". If you want to optimize for average/peak end values within that success range, typically the higher EQ% will provide that.

I did some runs in FIRECalc, but I can't really pick out volatility in the middle of all those squiggles. But for end points, using 40 years and 3.3% WR for 100% success, and testing a bit higher WR% for sensitivity - I found that 40/60 and 85/15 were both at the edges of failures. Neither ran out of money, and as expected, the peak and average ending portfolio was higher for 85/15.

But both pass, so (historically) either Person A or B could have chosen either AA and 'won'. And either could have gone for the moonshot (85/15), and neither would have had to cut any spending over the 40 years. So I guess we are really talking about whether the person with SS/pension would take on further volatility by going higher than 85/15? Hmmmm, OK, they could go to 100/0, have some failures (but still have their SS/pension), but increase their average and peak end points by about 1.35x.

But is going from 85/15 to even higher the kind of change one would typically consider? I don't hear too many voices here in the above 85/15 AA camp. I think it thins out pretty fast above 75%, even for those with a solid SS/pension.

-ERD50

 

[donheff]

That's a personal choice. But neither has any bearing on the historical optimum AA for a given WR and time frame. With 'optimum' defined in the common way - "not running out in my life time". If you want to optimize for average/peak end values within that success range, typically the higher EQ% will provide that.

I did some runs in FIRECalc, but I can't really pick out volatility in the middle of all those squiggles. But for end points, using 40 years and 3.3% WR for 100% success, and testing a bit higher WR% for sensitivity - I found that 40/60 and 85/15 were both at the edges of failures. Neither ran out of money, and as expected, the peak and average ending portfolio was higher for 85/15.

But both pass, so (historically) either Person A or B could have chosen either AA and 'won'. And either could have gone for the moonshot (85/15), and neither would have had to cut any spending over the 40 years. So I guess we are really talking about whether the person with SS/pension would take on further volatility by going higher than 85/15? Hmmmm, OK, they could go to 100/0, have some failures (but still have their SS/pension), but increase their average and peak end points by about 1.35x.

I suspect that few ERs would go for the shoot the moon scenario unless all of their essential expenses and then some were covered by the COLAd pension. To the extent that some necessary expenses, or even optional expenses, are dependent on the portfolio most people will aim for assuring their ability to keep up their life style rather than maximize growth. The larger the portion of expenses covered by a "guarantee" the more likely people might tilt a bit high on equities but I suspect most wouldn't stray far from the safe pack. I always thought I would but as I learned (and aged) I moved back into the herd.

 

[clifp]

Another vote for treating SS as reduction in your expenses. I have similar example with my PV system.

I just wrote a check for $12,800 to pay in advance for 20 years of electricity from the PV solar system that was installed on my house this summer. This is a power purchase agreement. Now when I looked at the economics of this. I said this PV is going to reduce my energy cost from ~$300/month (Electricity + electricity to charge the Tesla instead of buying gasoline) to about $20/month. So when I do my FIRECalc run next year, I'll reduce my spending levels by $280 and my portfolio by $12,800.

I suppose I could try and model this money as some type of ibond, or I could add the $12,800 to the value of the house. But it seems much easier just to reduce my expenses and not included it in my AA calculations.

 

[Chuckanut]

PV is going to reduce my energy cost from ~$300/month (Electricity + electricity to charge the Tesla instead of buying gasoline) to about $20/month. So when I do my FIRECalc run next year, I'll reduce my spending levels by $280 and my portfolio by $12,800.

I hope to see you in a Tesla commercial in 2033, driving your current Tesla with Elon in the passenger seat.

 

[Lsbcal]

After reviewing the comments, I tend to agree that SS is not a "real" asset, just an income stream until you die. I can see it to allow a bit more aggressiveness in a portfolio but not very much.

How aggressive to be in investing our portfolio's for later years? The 64k question. Your 75/25 is pretty aggressive but that's been a winning formula for the last 4+ years. Mine is 65/35 and may be dialed back somewhat in the next year towards 50/50. I don't wish to repeat a 2008 (or live through a 1929) scenario.

So yes, SS is an income stream that let's us manage the remaining income stream (4% of portfolio or whatever) in a slightly different way ... if we wish to ... or not. It's all a bit of a gamble as is life.

 

[photoguy]

That's a personal choice. But neither has any bearing on the historical optimum AA for a given WR and time frame. With 'optimum' defined in the common way - "not running out in my life time". If you want to optimize for average/peak end values within that success range, typically the higher EQ% will provide that.

I did some runs in FIRECalc, but I can't really pick out volatility in the middle of all those squiggles. But for end points, using 40 years and 3.3% WR for 100% success, and testing a bit higher WR% for sensitivity - I found that 40/60 and 85/15 were both at the edges of failures. Neither ran out of money, and as expected, the peak and average ending portfolio was higher for 85/15.

If you limit there to being only one definition of "optimum" then yes there's only one optimum historical portfolio (actually even in this case there could be multiple optima if several different portfolios come out to exactly the same probably of failure. And at 0% historical failure rate there are an infinite number of "optimum" portfolios)

But clearly some people prefer 40/60 and other 85/15 (and Obgyn perfers 0/100). What's driving these different portfolio choices is other concerns beyond strict historical probability portfolio value as calculated by FIRECALC <= $0 at 30 years.

 

[jclarksnakes]

DW and I have pensions that can cover our expenses. When we finally get around to collecting SS that will be gravy. Except for a little of DW's retirement funds that she put into annuities we are 100% in equity funds. The way I see our portfolio is that I can use it to buy a couple of new Lamborghinis or I can keep it in equity funds. Equity funds have done great for me for decades so I will probably stay 100% in equities till I die. It seems to me that whether someone has SS and or pensions makes a giant difference in how they should invest their portfolio. If we had no pensions or SS we would need to be very conservative in our AA. Maybe 80/20.

I understand how accountants would not count SS as an asset on a ledger. My DD is an accountant but she does speak English. If I was to tell her she is an asset to the family she would understand. I consider my SS to be an asset. I do not keep a ledger or worry about my AA. If I did keep a ledger I would not put my SS on it but I would still not worry about my AA.

We are all different. It is all in how you look at things.

 

[Independent]

I won't opine on whether people should label SS an "asset". To me, that's a fuzzy enough word that there are lots of definitions. SS would qualify under some but not others.

The important question is how to use SS in making decisions about investments and withdrawals.

For that question, I agree with earlier posters who say they use it to offset spending.

 

[Nords]

This is a distinction without a difference. It's right up there with paying off the mortgage or including your home equity in your net worth.

However it's giving Bogle & Merriman plenty of publicity and column inches, for which they're undoubtedly eager to keep contributing their opinions. It'd be entertaining to bring it up in a hotel bar during a CPA or CFP convention and watch the fistfights break out.

Regardless of what accounting category we assign to it, the income stream reduces your retirement's unfunded expenses. Pretty straightforward.

If your income stream covers a large percentage of your retirement expenses, and if you think that it's reliable, then you could pretend that it's the income from a portfolio of bonds or CDs. You could load up the rest of your portfolio with equities to "balance" the portfolio's sources of your total income. This would be especially important if your pension has no COLA, because the rest of your assets should ideally provide some sort of inflation hedge.

Here's another reason to pretend that your [-]pension[/-] income stream is an asset: you can slap a number on it. Most military servicemembers heavily discount their pension potential because it's "only" $xxxx per month. (The concept of a "COLA" is even more susceptible to parabolic discounting.) When you convert that monthly inflation-protected income to the amount of I bonds you'd have to hold in your portfolio, it causes people's behavioral psychology to view the pension in a completely different light.

We can always bicker about the issue while waiting for the Investment Police to enforce the GAAP dictionary definitions, but it would be shortsighted to dismiss the advantages of counting an income stream as an asset class.

Asset allocation considerations for a military pension (part 3 of 3)