Paritioned portfolios of Target Retirement 20XX for each decade of retirement

[chinaco]

I intend to retire @ 55. When I plan/devise/ponder a portfolio management strategy and expenditures, I tend to think in decades for example age 55 -to- 65 -to- 75 -to- 85 -to- 95.


Here is an approach I have been toying with. Divide the portfolio into 4 funding pools for 4 decades 55, 65, 75, 85.

50% of portfolio for decade 1 beginning @ age 55 invested in target retirement 2010
30% of portfolio for decade 2 beginning @ age 65 invested in target retirement 2020
15% of portfolio for decade 3 beginning @ age 75 invested in target retirement 2030
5% of portfolio for decade 4 beginning @ age 85 invested in target retirement 2035

Liquidate the house in the late years depending on health situation.


Let the fund auto-adjust the allocation. Most of these funds go to 20% stock about 5 years after the target date. So you can use the target date to adjust the risk profile of the portfolio to the decade.

This would be employed along with a cash (MM or CDs) account to be used as a buffer. The cash account would be seeded with enogh money to cover 2 to 3 years of [minimum] expenses beginning at 55 which is replenished oppotunistically during the good years all along the way. If we have to tighten the belt in one decade, we do not jeopardize the funding for the following decade. In otherwords, limit the expenditures in a decade to the pool of money allocated for the decade.

Please share your thoughts and criticisms.

[Telly]

Looks like you are creating your own Equity/Fixed allocation curve over time by paralleling different Target Retirements.

Just curious... did you load all the components you have selected into a spreadsheet, and using the "expected" allocation change vs. time from page 51 of the Prospectus for each of the components vs. your weighting factor, plot out a curve of the composite Equity/Income allocation versus your age years?

[cute fuzzy bunny]

I've thought about this, and it should work just fine, as described.

Only things are that the switch towards the higher bond % at the time of retirement is intended to dampen volatility at a time in your life when you presumably wont want very much, and increase income at a time when income is more important than growth.

If you dont expect those things to be relevant to you when you're 60-something, you can probably just go for a 'further out' target retirement fund and leave it at that...

[chinaco]

Originally Posted by Telly

Looks like you are creating your own Equity/Fixed allocation curve over time by paralleling different Target Retirements.

Just curious... did you load all the components you have selected into a spreadsheet, and using the "expected" allocation change vs. time from page 51 of the Prospectus for each of the components vs. your weighting factor, plot out a curve of the composite Equity/Income allocation versus your age years?

I have not actually worked the numbers yet. I just eyeballed the allocation adjustments by looking at the prospectus.

Portfolio pools further out in time are in growth mode allowing a smaller percent of the original portfolio to grow. For example the 5% pool (at start of retirement) has 30 years to grow before consumption. It is essentially like starting to save for retirement at 30 years old. The portfolio currently being consumed is very conservative to minimize volatility. The cash account is intended to let me skip drawing down the stock/bond fund if there is a bad year or two.

I have not worked the numbers closely... I want to determine the amount to seed each beginning portfolio and project expected growth. Some testing should help to validate the model.

One could do the same thing with other funds, but the target funds automatically adjust the portfolio allocation to more conservative as the target date approaches. It seems to me that makes it simple to manage (auto-pilot).

My goal is to maximize spending in the early years of retirement (55 - 75)

[Telly]

My thinking is that it would be worthwhile to spread-sheet it, and run graphs, then do some sensitivity analysis with respect to Vanguard's "expected" allocations.

An analogy is mixing up a custom color paint using four base colors. You want a specific color to result, but a color that changes a certain way as time progresses (hey, that would be cool paint!). Vanguard changes the hue of the base colors you are using over time. They have an "expected" hue change over time, but they are not required to fit that change exactly as shown in the prospectus. Nor is there any guidance given on how far off they may stray. Varying each of the "hues" in turn, and then varying all differently in a sensitivity analysis might show a surprise.

Like maybe you are measuring using the proverbial micrometer to get your intended result, but Vanguard will cut with an axe (four different axes, really). I don't know what it would look like without doing it.

[Rich_in_Tampa]

I'm not sure I see a whole lot of benefit over just going with 1 target fund. I see what you are driving at, but since at retirement you won't be draining your assets all at once, the long-term growth potential is preserved for the balance at any rate, self-balancing, etc.

Maybe select a target date later than your "real" one if you want to remain more aggressively allocated, but I'm not really seeing the value in what you propose.

[cute fuzzy bunny]

Rich...I think the perceived benefit is that this is like one of those 'many buckets' strategies where the more imminent withdrawal bucket is more conservative than the very last one you intend to tap. But your point (which I agree with) of simply choosing a single, or perhaps two, funds and adjusting the date to make the bucket more or less conservative also suits.

I do a little of this. For example, my most recent portfolio incarnation held TR 2025 in my taxable account, intending to use it first, and TR 2045 in my IRA, planning to use that last in life. Presumably i'll be in my 70's or 80's when I get to that, so having it mature in 40 years while remaining aggressive in that time period makes good sense to me as a risk taker. Using 2045 in my taxable and having to count on the income from that in my 60's is a little more volatility than I'd like.

[d]

why hasn't anyone pushed the numbers yet ? ... it would be interesting to see what it looks like over time.

[Olav23]

Wouldn't this just have an averaging effect? I'd imagine 50% in 2045 and 50% in 2015 would be (relatively) the same as 100% in 2030?

I could see utility if you needed to put it in different accounts, such as less aggressive in the taxable accounts to spend down first, and the more aggressive in the tax-free accounts, but if it is just to average the volatility of different funds, I don't really see the utility (logically anyway, but who knows how it plays out with actual numbers).

[citril]

This is almost exactly the plan that I described earlier, except you are grouping by decades, instead of by individual years. Make sure that you model out a worst case scenario to make sure that you can still cover your expenses in the later decades. 1966 is a good example of a worst case.

How did you come up with 50/30/15/5 split? These are very similar to what you would get discounting at 6%.

How do you plan to spend within each decade?

[cute fuzzy bunny]

Looked at overall, it does have an averaging effect. Looking at it in terms of withdrawing only from the 'nearest' bucket until its empty, it has a volatility lowering and income increasing effect on the near term money while maintaining higher volatility and potentially higher returns on the 'further out' money.

[chinaco]

Citril - I eyeballed each decades funding amount (%). It was mainly to illustrate the idea rather than a precise approach.

CFB sized it up correctly.