Present value of Social Security

[ERD50]

I don't see how one would ever achieve 25x. At 4%, when you die, on average you're left with portfolio 2 times larger than the one you started with. With SS, you're left with a few hundred dollars for death benefit.

Well, that is where the whole 25x comes into play. Because of variation in the market and inflation, you need a nest egg ~ 25x to assure you don't run out of money under the worst case scenarios. The effect of that is that you will very likely end up with a much larger portfolio when you die.

If we assume SS is 'risk free', then I would say you need a high success rate portfolio to replace that income. And that gets you back to a 25x number. Right?

Averages don't really help the individual retiree. If they want to have a high probability of not running out of money, their portfolio needs to be ~ 25x their spend.

Run FIRECALC. Set to retire in 2007 and the defaults. Set the results to show portfolio required for 95% success.

I get $754,716 starting portfolio with zero SS.
I get $503,135 starting portfolio with $10,000 SS starting in 2007.

So that is a delta of $251,501 in starting portfolio. Divide by the $10,000 SS payment, and you get... 25.15x.

You could wind up with zero in your portfolio, or a bunch - but since we can't predict that, we can't really do much about it (other than buy a big annuity....)

-ERD50

 

[FIRE'd@51]

For someone age 62, wouldn't SS payouts beginning at age 62 be equivalent to a ~ 25x nest egg?

If you look at the actuarial table on the SS web site, it says the life expectancy for a 62-year old male is 18.3 years. Currently, on Bloomberg, the real rate of return on a 20-year TIPS is 2.3%. So if you discount 19 one dollar payments at 2.3% (with a future value of zero), you get 15.3. So it looks to me that a 15 multplier would be about right, for the average 62-year old with no spouse.

If the payments were to last for 30 years, the same calculation would give a present value (multiplier) of 21.5.

 

[Cut-Throat]

If you look at the actuarial table on the SS web site, it says the life expectancy for a 62-year old male is 18.3 years. Currently, on Bloomberg, the real rate of return on a 20-year TIPS is 2.3%. So if you discount 19 one dollar payments at 2.3% (with a future value of zero), you get 15.3. So it looks to me that a 15 multplier would be about right, for the average 62-year old with no spouse.

A 15 Mulitplier would be withdrawal rate of 6.66% - Would you be comfortable spending 6.66% of your Nest Egg at age 62?

 

[ERD50]

If you look at the actuarial table on the SS web site, it says the life expectancy for a 62-year old male is 18.3 years. Currently, on Bloomberg, the real rate of return on a 20-year TIPS is 2.3%. So if you discount 19 one dollar payments at 2.3% (with a future value of zero), you get 15.3. So it looks to me that a 15 multplier would be about right, for the average 62-year old with no spouse.

If the payments were to last for 30 years, the same calculation would give a present value (multiplier) of 21.5.

This is where the differences are coming from. You guys are talking *averages* - I am not. What you are doing is fine for calculating the *average* value of SS payments. That is useful for the government, or for a large group of people, they can average all those statistics into their risk pool. But I am not a pool, I am me!

By that same thinking, drop your homeowners insurance. On average, your house will not burn down.

I am asking, what do *I* need to do to account for how SS payments affect on my nest egg? I can't count on living an *average* life span, I want to account for some of the worst cases. With your approach, every male age 62 would type 18.3 years into FIRECALC and be happy with the result. Probably fine for 50% of the people

Bogle was speaking from the angle of the individual, so I still see 25x as the only reasonable assumption.

-ERD50

 

[FIRE'd@51]

This is where the differences are coming from. You guys are talking *averages* - I am not. What you are doing is fine for calculating the *average* value of SS payments. That is useful for the government, or for a large group of people, they can average all those statistics into their risk pool. But I am not a pool, I am me!

By that same thinking, drop your homeowners insurance. On average, your house will not burn down.

I am asking, what do *I* need to do to account for how SS payments affect on my nest egg? I can't count on living an *average* life span, I want to account for some of the worst cases. With your approach, every male age 62 would type 18.3 years into FIRECALC and be happy with the result. Probably fine for 50% of the people

Bogle was speaking from the angle of the individual, so I still see 25x as the only reasonable assumption.

-ERD50

I would say that "on average" the average life expectancy would be the best non-biased estimator, and the one to be used. If you choose the higher multiplier, aren't you potentially overestimating the size of your portfolio by using a 25 multiplier?

But I do understand your concern, and you could still do the calculation I suggested - just use more than 19 payments. However, as I mentioned above, even 30 years of payments only gives a multiplier of 21.5. To get a multiplier of 25, you would need 38 years of payments, which would put you way out in the right tail of the life-expectancy distribution.

With regard to FireCalc, as I understand it, it will just keep including the SS (and adjust them for inflation) for as long as you specify the FireCalc run (e.g. 30 years, 40 years, etc). So FireCalc will automatically take care of the problem you are raising. My comments refer to putting an equivalent bond valuation (e.g., for the purpose of asset allocation) on the SS payments. That is to say, what you could hypothetically "sell" your SS payments for.

 

[ERD50]

I would say that "on average" the average life expectancy would be the best non-biased estimator, and the one to be used. If you choose the higher multiplier, aren't you potentially overestimating the size of your portfolio by using a 25 multiplier?

But I do understand your concern, and you could still do the calculation I suggested - just use more than 19 payments. However, as I mentioned above, even 30 years of payments only gives a multiplier of 21.5. To get a multiplier of 25, you would need 38 years of payments, which would put you way out in the right tail of the life-expectancy distribution.

I'm 52, 38 years puts me at 90. That puts me to the right of 80.3, but not so far that I can choose to ignore it.

It still seems we are talking about two different things - 'averages', and 'what the individual must do to plan for themselves'. I don't disagree with your calcs at all as far as averages, I just don't think they apply to an individual.

Of course, on average, a 25x multiplier on my nest egg means I have a high chance of leaving a large estate behind. But that is what is required for an individual to protect themselves financially against a bad market and/or a long life. Isn't that pretty much the whole concept behind FIRCALC and the 4% SWR? Why is this discussion of personal valuation of SS payments based on averages?

So answer this: What do use for your own life expectancy for your projections?

Several of my relatives did not have good health in their early years, and they still made it well into their 90's. I don't think the chances of my wife or myself living to 100 is going to be that far out on the z-tail. At least not so far out that I don't want to try to prepare for it. What further advances in medicine will we have in 25 years? I don't want to be in the position where my 60 and 70 year old children need to support their 99 year old Mom or Dad for a few years. They will have their own problems.

If I have a huge nest egg at 85 or 95 and a better view of my life expectancy at that time, no problem. I will evaluate the needs of my children, any grand/great-grand children, other relatives and friends, and charities, and act appropriately. That is the least of my worries.

-ERD50

 

[FIRE'd@51]

ERD50,

I think we are pretty much in agreement. I added a paragraph about FireCalc while you were responding.

My present values would all be at the age of 62. To get them to 52 you would need to discount my multiplier by an additional 10 years (using the TIPS rate) but, of course, with no interim SS payments. This would have the effect of lowering the multipliers further.

My comment about your being in the far right tail of the distribution was also relative to age 62, so I was assuming you would live to 100, not 90.

Again, let me make it clear. I am trying to put a "bond-equivalent" value on a stream of SS payments. That is a straightforward calculation once the payment stream (payment amount and number of years) is specified.

 

[FIRE'd@51]

A 15 Mulitplier would be withdrawal rate of 6.66% - Would you be comfortable spending 6.66% of your Nest Egg at age 62?

What multiplier would you put on the average 62 year old's SS payment stream?

 

[ERD50]

I realized the 52 vs 62 thing ten minutes after I posted But for me personally, since I do want to plan out to 100, it is all pretty much the same. But that is just me.

OK, so you are calculating the bond equivalent of a stream of SS payments. But if the intention is to determine how it might impact an individual's asset allocation, I still think you need to look at what FireCalc would tell you about how it impacts the individual. That is still the thing, not the average value of that SS stream.


Let's go back to my earlier FireCalc example (defaults including 30 year cycle):

Run FIRECALC. Set to retire in 2007 and the defaults. Set the results to show portfolio required for 95% success.

I get $754,716 starting portfolio with zero SS.
I get $503,135 starting portfolio with $10,000 SS starting in 2007.

Further, let's assume each wanted an equivalent 75/25 split of Equities/Bonds. Let's round to $750K and $500K portfolios.

With zero SS, a 75/25 split would point to a $562K equity, $188K bond, portfolio. However, the person with the SS payment is thinking of these as bond equivalents. W/O SS, they would have had $562K in equities. But, they only have $500K total. So, it appears to me that they would keep their entire $500K in equities, no bonds.

I suppose another valid view would be Buffet's other statement, think of the SS payments as an average 5% bond dividend. Then the $10,000 SS payments represents $200K of 'phantom bond assets'. End up about the same,$750K minus $200K in bonds gets you to $550K equities - the whole thing.

I just don't think that the probable ending value of the portfolio fits into this calculation. We do not know that until the day the person dies, so we just plan on what we can. In this case, 30 years and a good chance of success based on the market history - maybe a big pile at the end, maybe not. On an individual basis, the size of the pile could end up at either extreme, we can't really 'plan' on either - it will be what it will be.

Is there a flaw in that logic? Or is that not the 'problem' you were trying to solve?

Oh, so my own answer to the question you posed to C-T is, yes, I would assign a 25x multiplier to the SS stream. Because that is how it impacts my personal portfolio, and that is my concern.

Edit/Add: One small point. This just became an iterative problem since we ran the defaults at 75/25, but the SS guy points to 100/0. So I re-ran FireCalc with 100% equities for the SS case. It wants a slightly higher starting portfolio of $528K, so it would still be all equities.

-ERD50

PS: I hope you don't think I'm just trying to be argumentative on this. I am trying to get a handle on this for my own needs. I'm really just trying to validate if my position is correct. If I am shown that I'm wrong, I will have learned something. That's a good thing. Or, it could be that people decide there are different, but still valid views to take. That is fine, too.

 

[ERD50]

Another thing. Maybe this is more of an asset allocation observation, but as long as I'm thinking of it, I'll spit it out now:

Your SS income stream clearly impacts how big of a portfolio you need for a given success rate. But, why would it change the asset allocation of that portfolio? At first, it does seem obvious that you can think of SS as a bond payment, so it represents bonds, adjust accordingly, etc, etc, etc.

But look at our two previous cases. Each requires a different size starting portfolio. But in each case, that portfolio is ~ 25x what needs to be drawn from it. So we really have two different *size* portfolios, but the *same* withdraw percentage rate from each. So if I run FireCalc and look for some optimum ratio of stocks/bonds, it will be the same. Whatever is good for a $500K portfolio with a 4% withdraw rate, will be good for a $750K portfolio with a 4% withdraw rate.

It appears that we should *not* change our asset allocation based on pension and SS? This has probably been discussed another thread, but - there you go.

-ERD50

 

[Cut-Throat]

What multiplier would you put on the average 62 year old's SS payment stream?

the average 62 year old does not interest me.

But for myself, I'd use a Multiplier of 25

 

[FIRE'd@51]

the average 62 year old does not interest me.

But for myself, I'd use a Multiplier of 25

I'm really surprised that you would say that. Are you aware that Vanguard will sell you the annuity we are talking about at a multiplier of 19? And they have all kinds of fees built into that product.

 

[FIRE'd@51]

Another thing. Maybe this is more of an asset allocation observation, but as long as I'm thinking of it, I'll spit it out now:

Your SS income stream clearly impacts how big of a portfolio you need for a given success rate. But, why would it change the asset allocation of that portfolio? At first, it does seem obvious that you can think of SS as a bond payment, so it represents bonds, adjust accordingly, etc, etc, etc.

But look at our two previous cases. Each requires a different size starting portfolio. But in each case, that portfolio is ~ 25x what needs to be drawn from it. So we really have two different *size* portfolios, but the *same* withdraw percentage rate from each. So if I run FireCalc and look for some optimum ratio of stocks/bonds, it will be the same. Whatever is good for a $500K portfolio with a 4% withdraw rate, will be good for a $750K portfolio with a 4% withdraw rate.

It appears that we should *not* change our asset allocation based on pension and SS? This has probably been discussed another thread, but - there you go.

-ERD50

I look at retirement as matching assets to liabilities. My liabilities are my expenses. Just like SS, they end when I die. So I would use SS to offset that portion of my expenses, and use FIRECalc to handle the residual. Taking it to the extreme of the person who has a COLA'd pension (and SS) that meets the expenses necessary for him to achieve his desired lifestyle, I would argue he should hold 100% equities (OK, maybe he should hold a small amount of cash just for cases where he might need a quick cash infusion). It is in this sense, that I think about the bond-equivalence of SS (or a pension). But it is a strange type of bond since the payment stream disappears at death. It is also impossible to put a present value on such a stream (any more than it is impossible to put a present value on one's expenses), without making an assumption about the time-length of the stream, and the only way I can see to do this is to use average life-expectancy.

I would just say if I could randomly draw a 62 year-old's SS stream from the distribution of all 62 year-olds, and sell it to you and C-T for a 25 multiple, that is a trade I would do all day long in as much size as I could. Statistically I would make a tanker-load of money doing this trade.

 

[Cut-Throat]

I'm really surprised that you would say that. Are you aware that Vanguard will sell you the annuity we are talking about at a multiplier of 19? And they have all kinds of fees built into that product.

That's why I might buy an annuity at that time. - But if I were depending on my own Nest Egg, I'd still only take 4% at age 62. - Which is a multiplier of 25.

Vanguard plays the averages, individuals can not!

That is why I find the Cola Annuity idea attractive at that age.

 

[ERD50]

I'm really surprised that you would say that.

Vanguard plays the averages, individuals can not!

Exactly! As individuals trying to plan for for the case that we outlive the average, a 25x multiplier fits.

The annuity people can work with averages and standard devs. But I can't pay the rent and grocery bill by saying 'but, I should have had enough money on average!'.

And, our SS isn't something we can buy or trade, it just *is*. We can decide to delay or not delay though.

-ERD50

 

[Cut-Throat]

I would just say if I could randomly draw a 62 year-old's SS stream from the distribution of all 62 year-olds, and sell it to you and C-T for a 25 multiple, that is a trade I would do all day long in as much size as I could. Statistically I would make a tanker-load of money doing this trade.

OK - I see your point! and maybe S.S. should not be valued at 25, but I'm guessing not 15 either! - If you are married and your spouse would get your S.S. upon your death would be a valid Comparison.

- So what tell me what is the SWR of a COLA Annuity at age 62 with survior benefits? - I'm thinking it's a lot less than 6.66% which would be a mulitpler of 15 - Maybe around 5% - which would be a mulitplier of 20?

And then you have to decide which is a 'safer' bet the U.S. Government or an Insurance Company.

 

[FIRE'd@51]

That's why I might buy an annuity at that time. - But if I were depending on my own Nest Egg, I'd still only take 4% at age 62. - Which is a multiplier of 25.

Vanguard plays the averages, individuals can not!

That is why I find the Cola Annuity idea attractive at that age.

I don't see how you can put a multiplier of 25 on something that someone else (who is not the low-cost provider) will sell at a multiple of 19.

 

[FIRE'd@51]

If you are married and your spouse would get your S.S. upon your death would be a valid Comparison.

- So what tell me what is the SWR of a COLA Annuity at age 62 with survior benefits? - I'm thinking it's a lot less than 6.66% which would be a mulitpler of 15 - Maybe around 5% - which would be a mulitplier of 20?

Well, now you are changing the problem. We've been talking about a 62 year-old single male.

But again, the present value of a SS stream (or any annuity for that matter) can only be calculated if you know the lifetime of the stream, and this will require an assumption about joint-life expectancy. A multiplier of 20, using the TIPS rate as the discounter would imply a level payment stream that went on for 28 years, and this is neglecting the fact that the payment would drop from $1 to $0.67 after the first spouse died which would lower the present value (i.e., the multiplier).

 

[Cut-Throat]

Well, now you are changing the problem. We've been talking about a 62 year-old single male.

The problem was valuing S.S. - S.S. has spousal benefits. How am I changing the problem? - I never understood why you were talking about a single 62 year old male. - Weren't we trying to value S.S. benefits?

 

[FIRE'd@51]

The problem was valuing S.S. - S.S. has spousal benefits. How am I changing the problem? - I never understood why you were talking about a single 62 year old male. - Weren't we trying to value S.S. benefits?

That was the example I used (and stated clearly) to compute a present value in response to ERD50. But that's OK, read the second paragraph of my last post.

 

[youbet]

The problem was valuing S.S. - S.S. has spousal benefits. How am I changing the problem?

SS does not always have spousal benefits, even if you are married. I think you would need to define the problem to take this into account.

 

[ERD50]

I would just say if I could randomly draw a 62 year-old's SS stream from the distribution of all 62 year-olds, and sell it to you and C-T for a 25 multiple, that is a trade I would do all day long in as much size as I could. Statistically I would make a tanker-load of money doing this trade.

But aren't you trying to assign a 'market value' to something that is not 'marketable'?

Are you saying that since I value the SS at 25x, I should jump at the chance to buy that protection on the free market at 19, or that I would be willing to buy more, even at 25x? OK, but consider - I didn't 'buy' my SS, it was handed to me. Now that I know what it should be, I can place a value on it. That does not mean I wish to buy more, even at a lower cost. I can't fully trust Feds to deliver on their promise, I ccertainly can't fully trust a private annuity insurer to deliver 40 years from now. As pointed out in a recent post, there are survivor benefits to SS also (those were not factored into the FireCalc numbers though).

In a way, it is a bet against the Feds delivering, and the market delivering. That 25x number is based on the markets performing at their historic worst. Odds are, the markets will do much better. I'm not sure I want to bet any more of my future on the fed or a private insurer than I already have.

That is not to say I would not buy a private annuity. I doubt that I would, the numbers might not add up, but it might be a decent diversification strategy between Mr Market and Uncle Sam.

-ERD50

 

[FIRE'd@51]

But aren't you trying to assign a 'market value' to something that is not 'marketable'?

Well, the title of the thread is "PV of SS", so I was just trying to come up with a framework to do the valuation.

Are you saying that since I value the SS at 25x, I should jump at the chance to buy that protection on the free market at 19, or that I would be willing to buy more, even at 25x? OK, but consider - I didn't 'buy' my SS, it was handed to me. Now that I know what it should be, I can place a value on it. That does not mean I wish to buy more, even at a lower cost.

No, I'm not saying that at all. I was just using Vanguard's price as another possible indication of the value. If Vanguard/IAG prices an equivalent payment stream at 19 (and we know they are making a huge profit on these annuities), then I don't see how you can argue that it's price should be 25.

I am not trying to argue for or against an annuity here. I agree we have paid (overpaid) for our SS, so of course we are going to take it. But to value it properly one must compare it to other similar securites that exist in the market place.

 

[ERD50]

Well, the title of the thread is "PV of SS", so I was just trying to come up with a framework to do the valuation.

And I'm certain that you are technically correct in your calculations. What I've been saying is, those calculations are fine in a general sense, but are not that meaningful to an individual....

But to value it properly one must compare it to other similar securites that exist in the market place.

And I don't think you can compare being handed SS with other marketable items. This is probably a really lousy analogy, but...

The old rusty sink in my bathroom has no market value to anyone. Yet, I take care of it until I'm ready to remodel, because if it got damaged, I'd need to spend $$ and labor to replace it. So, it has value to *me*, I can even define that value in the price of a cheap sink to replace it until we remodel and can pick colors and style for a permanent replacement, and the labor to replace it, but it still has no market value. So, I claim that an item can have individual value separate from it's market value. Tangible, measurable value.

So, when I run FireCalc, or just think through C-T's analysis, I see the value of those SS payments to *me*, and I don't think it makes any difference what the market value of a similar product is. That does not change the equation for me. Unless I could sell my SS benefits and buy some other benefits - but I don't think that's possible. So it is a personal value thing, outside of market forces.

I think.

-ERD50

 

[Sam]

Run FIRECALC. Set to retire in 2007 and the defaults. Set the results to show portfolio required for 95% success.

I get $754,716 starting portfolio with zero SS.
I get $503,135 starting portfolio with $10,000 SS starting in 2007.

So that is a delta of $251,501 in starting portfolio. Divide by the $10,000 SS payment, and you get... 25.15x.

Huh? Why would I want to that? I'm only interested in the PRESENT value. Presently, I'm 49, so 2020 is the earliest I can collect SS.

That exercise might be applicable to you, if you're 62 now, and believe that you would live to 92 or beyond.