Notifications
Clear all

Actuarial Spending Strategy vs. Bogleheads Variable Percentage Withdrawal (VPW) Strategy? It seems to give similar results but is not transparent. I am using the the 70%/30% (Equities/Fixed Income) column of the VPW table.

9 Posts
3 Users
0 Reactions
602 Views
(@gharbour)
Active Member
Joined: 8 months ago
Posts: 5
Topic starter  

Can I use the Actuarial Spending Strategy to approximate Bogleheads Variable Percentage Withdrawal (VPW) strategy?



   
ReplyQuote
(@gharbour)
Active Member
Joined: 8 months ago
Posts: 5
Topic starter  

The Actuarial Spending Strategy (ASS - sorry not my acronym!) seems to give similar results but it is not transparent. I am using the 70%/30% (Equities/Fixed Income) column of the VPW table. This starts my withdrawal percentage at 5.7% at age 70 and climbs every year (80 years=7.1%, 90 years= 11.9%, then I use a maximum of 20% per year from age 95 on as do many VPW proponents). I like the Actuarial Spending Strategy that lets me set a "legacy amount". But I can't see what percentages Actuarial is using for any given age and asset allocation. So it lacks transparency for me. I can't find the actual calculation method anywhere for Actuarial Spending to know how it is similar or different from VPN. I did do some calculations by manually calculating the withdrawal rate in the ASS vs VPN and show a few ages below).

Age. ASS VPN (70/30 table)

70. 5.7%. 5.7%

75. 6.5%. 6.2%

80. 6.7%. 7.1%

85. 8.5%. 8.7%

90. 11.7%. 11.9%

95. 21.9%. 21.6% (I have limited to 20% max)

98. 33.1% 51% (I have limited to 20% max)

I would love to use the ASS because it is as close as I can get to VPN. Benefits of both methods seem to be that they seem to consider (VPN does consider) your current portfolio amount, your asset allocation, and your age.

Where can I get more information about ASS and the calculation method behind it?



   
ReplyQuote
(@gharbour)
Active Member
Joined: 8 months ago
Posts: 5
Topic starter  

So in looking at the numbers I calculated for ASS it seems clear that it is using my age, my allocation, and my portfolio balance to calculate my withdraw at almost the precise levels as Bogleheads VPW does. So if Parlana is using VPW and Parlana seems to go out of their way to credit all the other withdrawal strategy to their source, why not call ASS by VPW and acknowledge that's where it comes from? This would let users benefit from all the wonderful information on VPW on the Boglehead site and allow transparency for Parlana users. Am I missing something?



   
ReplyQuote
(@ricke)
Reputable Member Customer
Joined: 5 years ago
Posts: 266
 

@gharbour

I don't know the details of the algorithms, but the concept of actuarial based spending predates anything from bogleheads. You are correct that the concept is to use actuarial tables (life expectancy) and your entered rate of return to find a spending rate that consumes your portfolio (or hits a legacy target). In real life, as the market moves up and down, the resulting spending will be variable, but with slow adjustments because it spreads things out over your remaining life expectancy. So both will be variable spending strategies with relatively slow adjustments.

As I recall, the VPW forward test that one of the boglehead members has been running for several years involves some additional smoothing of the spending in the short term (putting some money aside in months when the market is good and taking it back out when the market is down). The reason being that the big objection folks have to any variable spending strategy is that it is, well, variable. So you have to cut back spending when the market declines.



   
ReplyQuote
(@jkandell)
Reputable Member
Joined: 4 years ago
Posts: 258
 

The manual describes Pralana's method in detail: For each year, they amortize the portfolio on 12/31 of the previous year, based on your longevity and your portfolio rates of return. They deduct out essential expenses from this amount to give you your "variable spending" for the year. Note that unlike VPW pralana also has a ceiling (200%) and floor (50%) for essential expenses, to avoid large jumps year to year.

With regard to your suggestion, I caution you to remember that there are lots of different ways of using the actuarial/annuitization method, and we can't assume VPW is the standard.

For instance, the RMD method (that uses the IRS factor, multiplied by a user chosen constant e.g. 1.26), or TPAW (annuitizing what's left on ones total portfolio including future social security and pensions, after taking into account the npv of future projected essential expenses vs income streams). On the other hand, I am not sure of anyone who uses the Pralana method, though it seems as good as any. 🙂


This post was modified 4 months ago by Jonathan Kandell

   
ReplyQuote
(@gharbour)
Active Member
Joined: 8 months ago
Posts: 5
Topic starter  

Hey Jonathan, Thanks for your reply. I appreciate it and would like to hear more. But, I still don't know the actuarial table(s) that Parlana are using. For any of these actuarial methods the basic calculation is the same. There are some variations in how they handle side issues like; residual target portfolio values, social security, annuities, asset mix, floors/ceilings, exceptional income and expenses, essential and variable spending, etc. But the basic core calculation method is pretty much the same. It is what I am looking to understand as a starting point.

Step 1. In an actuarial table you look up your age and determine your withdrawal percentage. For example:

Age. Withdrawal%

70. 5.7%

75. 6.5%

80. 6.7%

85. 8.5%

90. 11.7%

95. 21.9%

98. 33.1%

(Footnote: in VPW this is where adjustments for asset mix and growth is made via a matrix of tables depending on fixed/equities split.)

Step 2. You then take this percentage and multiply it by your Portfolio Value on 12/31 of the previous year to get the withdrawal amount. For example:

If I'm 70. And have $5MM.

5.7% * $5M =$0.285M

But without knowing the actual Withdrawal percentages that Parlana uses, that is their actuarial table(s), (except from my own empirically derived values for Parlana see my message from ) I don't know their actual calculation. Unless, I'm just missing the information and it is disclosed somewhere? So, I wish Parlana would be more transparent. It looks from my empirically determined values for Parlana (see above) that they use a very similar actuarial table(s) to VPW (maybe with some divergence after age 95 and adjusting for asset mix in a way currently unknown to me).

BTW it is the same with TPAW, as far as I am aware, we don't know their actuarial table(s). From Sally's candy problem they present, I suspect it is a simple 1/nth of the expected "Total" portfolio value, where nth = number of remaining years and "Total" is the portfolio accounting for asset mix and growth. Since I don't have their formula or actuarial table(s) I don't know their basic calculation and have to just trust it, or not. But, I also can't do an easy comparison with other methods except based on output.

The RMD method at least reveals their secret sauce, they use the IRS actuarial table x 1.26. Which (IRS Tables) were the starting point for the folks at Boglehead that developed the VPW method. For various reasons they modified these table(s) and I buy into those reasons.

So, I know what VPW is actually doing, I feel comfortable with how they got to their actuarial table(s) and therefore with VPW. I am trying to get the same comfort with Parlana and that's why I want to know the actual actuarial table(s)/method that they use. I want to know their secret sauce. 😉



   
ReplyQuote
(@jkandell)
Reputable Member
Joined: 4 years ago
Posts: 258
 

Actuarial tables are just summaries based on the time value of money equations. (Like FV = PV x (1 + i) ^ n or the PMT amortization formula PMT(i, n, pv, fv).)

Charlie or Stewart can correct me, but to my understanding Pralana version of actuarial withdrawal is the following procedure (which you can see in year-by-year detail by looking at the "audit" tab under Analyze > Spending Strategies after clicking "Actuarial"):

  1. Pralana applies the PMT equation to the portfolio value as of 12/31st of the previous year, based on remaining years in plan and the expected return of the portfolio, to produce an annuitized total withdrawal for the year.
  2. Essential expenses are then subtracted out of that amount to leave you the Variable Spending amount for the year.
  3. This variable spending amount is subjected to a ceiling and floor to avoid wide swings: 50% of last year's variable spending as floor, 200% of last year's variable spending as a ceiling.
  4. Rinse and repeat the following year after applying the expected return to the remaining portfolio.

Step 1 is essentially the same as VPW. If you wanted to reduce it to a table like you cite, you could look at the audit results, and divide the fourth column ("actuarial expense") over the first column ("year end total savings"). (Note differences with VPW: Pralana uses portfolio expected return rather than historical averages; and imposes a ceiling and floor.)

So for me Pralana's just as transparent as VPW, at least once you find the 'audit' tab. VPW focuses on the "total" year annuity amount, whereas Pralana focuses on the discretionary amount ("variable spending") left over after essential expenses are taken out of the total yearly allotment.

My problem with both Pralana and VPW isn't lack of transparency, but that they treat each year's income and expenses discretely rather than looking at the total lifetime future income stream and total future lifetime expense stream-- but that goes beyond your question so I won't say more. If you want to see TPAW's equations, look at the old excel spreadsheet predating the online calculator--it shows the pmt equations of the model. It is the standard amortization pmt formula, which I think is what you're describing. TPAW's advantage is it diversifies risk over time, so e.g. if you have a huge pension and social security or low expenses in the future, you can take out more now compared to someone who doesn't, and vice versa.


This post was modified 4 months ago 29 times by Jonathan Kandell

   
ReplyQuote
(@gharbour)
Active Member
Joined: 8 months ago
Posts: 5
Topic starter  

Hi Jonathan, Thanks for your reply. I just skimmed it. It will take me a little while to work through it. I did a quick peak at the audit tab and will go through it. At first glance I'm not sure why the Audit "Actuarial Expense" is different from the Expense Statement tab?



   
ReplyQuote
(@jkandell)
Reputable Member
Joined: 4 years ago
Posts: 258
 

@gharbour Actually, the Spending Statement shows overall spending percentage, so you're right! But the audit shows the "logic" of how they calculate it each year, i.e. transparency.


This post was modified 4 months ago 2 times by Jonathan Kandell

   
ReplyQuote
Share: