Portfolio Income

[Admin/YBB]

#PortfolioIncome for Retirees. Older studies suggest 4% INFLATION ADJUSTED withdrawals from portfolios with allocations in a range from 30-70 to 70-30 [%stock-%bond]; the withdrawal period is assumed to be 30 years. Withdrawals are set at 4% of INITIAL portfolio values & then are INFLATION ADJUSTED ANNUALLY regardless of portfolio performance. Studies based on Monte Carlo simulations [computer-generated random data trials] also lead to similar results.

An alternate strategy with 5-yr portfolio reviews is to use 4-5% withdrawal from initial balance that is not adjusted for inflation. But EVERY 5 YEARS, if the portfolio balance is higher than that in the previous review, reset to 4-5% withdrawal of the higher balance; otherwise keep the withdrawal the same for another 5 years. This conditional method adjusts withdrawals up only when the portfolios can support it & keeps withdrawals level when the portfolio values are down to allow portfolios to recuperate. Use portfolio allocations similar to the above. In the current low interest rate environment, the bond portion can be split among bonds & alternatives [real estate, precious metals, commodities, etc]. Use EFFECTIVE EQUITY for final portfolio allocation adjustments [see an earlier post].

Portfolio income needs to cover only the INCOME GAP that is the difference between basic expenses & income from other streams such as pension, #SocialSecurity, etc. #FinancialIndependence , #PersonalFinance , 12/4/20.

[Admin/YBB]

Income Portfolio CAGR, TWRR, MWRR

Media and poster conversations are shifting from accumulation to decumulation [retirement income].

While there is agreement on how to look at returns during accumulation [TR w/reinvestment], there is not much agreement on how to analyze decumulation. Problem is how to account for amounts withdrawn and spent [what return for consumption?] or reinvested elsewhere [at what return?].

Here, I am suggesting to look at 2 stats that PV generates, CAGR and TWRR.

CAGR is the compound annual growth rate. In PV, additions are accumulated but withdrawals are gone/ignored/unaccounted. Then, the TR/CAGR is calculated from the beginning and ending balances. This treatment of additions/withdrawals is uneven. But CAGR tells the story of what is happening to the account balances in accounts under withdrawal. If CAGR is negative, that should be a cause for concern.

TWRR is the time weighted rate of return and is basically the geometric average of segments separated by the times for additions/withdrawals. It amounts to ignoring additions/withdrawals and is basically the same as point-to-point reinvested TR of the underlying investment.

PV also has MWRR that is formal but less practical. MWRR is money weighted rate of return and is the XIRR calculation accounting for additions/withdrawals and time value of money. It assumes that the same return could be obtained on withdrawals by investing them elsewhere as the MWRR here - not a realistic assumption.

Example: $10,000 initially invested in 3 popular income funds VWEHX, VWINX, PIMIX. Annual withdrawals were $400 [or 4%] initial that were subsequently adjusted for inflation/COLA; 11 full years of withdrawals totaled $4,843 [average 4.4% of initial amount]. PV ran from January 2008 - October 2019, limited by the history of PIMIX.

Link

                Final Bal      CAGR      TWRR

VWEHX     $14,030       2.90%     6.51%

VWINX      $14,572       3.23%     6.95%

PIMIX       $18,121       5.15%      8.28%


Comments

VWEHX and VWINX were comparable but PIMIX was far ahead.

TWRR - CAGR < Initial withdrawal rate in this example when the total period had positive returns. In several other examples that I ran over other time periods, I didn't find a consistent relation among CAGR, TWRR, initial withdrawal rate.

Edit/Add: See this link for analysis under a different scenario - when all distributions are taken in cash [i.e. not reinvested], community.morningstar.com/t5/forums/forumtopicprintpage/board-id/income/message-id/1308/print-single-message/true/page/1


Source
yogibearbull (Customer) asked a question.
November 19, 2019 at 1:04 PM
community.morningstar.com/s/question/0D53o00005E8YGqCAN/income-portfolio-cagr-twrr-mwrr

[Admin/YBB]

SOR - Sequence of Returns is a risk to be aware of.

While withdrawing retirement income from portfolio, a bear market early can hurt the portfolio so badly that it never recovers. On the other hand, a bull market early is good.

SOR is a risk during accumulation also - bear markets early and during are good, but bear market near the end is bad.

There are 2 ways to address SOR: 1) Use conservative- or moderate- allocation funds, 2) When portfolio value falls too much, make adjustments (temporarily reduce or stop withdrawals).

M* has the following articles on SOR. 

www.morningstar.com/articles/1048704/sequence-of-returns-what-it-means-and-how-to-deal

www.morningstar.com/articles/1048863/sequence-of-returns-part-ii

(highlight link text--Go to http://www....; or highlight link text--copy (don't copy link address)--paste in a new tab to open)

[Admin/YBB]

More on Sequence of Returns (SOR) Risks

M* Amy Arnott has a follow up interview on SOR. Basically, the issue is more serious during the withdrawal/income phase, not during the accumulation phase. A multi-asset diversified portfolio is a good defense; buckets or buffers may also be helpful.
www.morningstar.com/articles/1057939/one-of-the-most-underappreciated-dangers-of-investing
(highlight link text--Go to http://www....; or highlight link text--copy (don't copy link address)--paste in a new tab to open)

[Admin/YBB]

SOR and 529s

SOR is also relevant when one is saving for a goal and huge market swings near the date of the goal can have big impact, e.g. for college 529s.
www.morningstar.com/articles/1061437/what-bad-returns-at-the-wrong-time-can-mean-for-college
(highlight link text--Go to http://www....; or highlight link text--copy (don't copy link address)--paste in a new tab to open)

[Admin/YBB]

SOR During Accumulation or Withdrawal, Twitter LINK

YBB Personal Finance
@ybb_Finance
4m  (8/30/22)
Replying to
@morningstarinc

Key is that #SOR exists for fixed/variable dollar-amounts of contributions/withdrawals, but not for a set % of contributions/withdrawals (including 0% or none). The latter is because of #GeometricAveraging.

Also, SOR during withdrawal is more disastrous that during accumulation.

NOTE. In geometric-averaging, returns compound through multiplication and that is "cumulative", i.e. the order of multiplications doesn't matter.
Response to M* JR article, www.morningstar.com/articles/1111697/retirees-arent-the-only-ones-who-face-sequence-risk

[Admin/YBB]

Income vs Total Return

Some want reliable monthly income; others are fine with total return (TR) growth and selling as needed (that may be easier said than done). The TR formula shows that one cannot have high income and high growth:

%TR = %Dividend_yield + %Dividend_growth + %Change_in_P/D

For very long-term, the last term has a small contribution, so if dropped, we get the Gordon Equation approximation for multiple years (so, don’t create examples of annual exceptions):

%TR = %Dividend_yield + %Dividend_growth

I have seen this debate about income vs TR go on and on for years. My suggestion is for income-oriented posters to assume a fixed monthly/annual withdrawal rate, e.g. 5% annualized (not adjusted for COLA), and then compare the terminal balances using, say, Portfolio Visualizer (PV). In the example below, I will use utilities XLU, CEF PDI, REITs VNQ and VFIAX/SP500; the data will go back to 6/1/12 due to late-05/2012 inception for PDI; also, PV allows only 3 portfolios plus 1 benchmark at a time.

PV RUN

PV RUN (Inflation-Adjusted)

All 5% fixed withdrawal programs were successful in that the terminal values > initial values. The best for terminal values were, in order, SP500 (highest), PDI, XLU, VNQ (lowest).

Intrinsic portfolio income varied, but the withdrawals were full covered by PDI, only sometimes by XLU and VNQ; so often, the withdrawals consisted of intrinsic income plus some principal.

There were 3 equity funds and 1 leveraged CEF and their volatilities were high. The 3-yr volatility order was VNQ (highest), XLU, SP500, PDI (lowest).

Max drawdowns were for PDI (highest), VNQ, SP500, XLU (least). That this order was different from that for volatility was a bit surprising.

What happens to the withdrawn amount doesn’t matter. But it is most likely spent by the income-oriented investor. Some don’t need the income and reinvest it – but in that case, only a straight TR comparison (TWRR) matters, and the order then was SP500, PDI, XLU, VNQ.

The period under consideration included the Covid pandemic and Russia-Ukraine war.

As expected, there was no clear conclusion. It boils down to the current needs and volatility tolerance (comfort level) of the investors

It may be useful to add conservative/moderate-allocation funds into this mix but that is for another time.

[Admin/YBB]

Example - VWINX, VWELX, PRWCX, Benchmark VFINX (SP500)

Terminal Balances  PRWCX (highest), SP500, VWELX, VWINX (lowest)

Intrinsic Income (including CGs) covered the withdrawals.

3-yr SDs  SP500 (highest), PRWCX, VWELX, VWINX (lowest)

Max Drawdown  SP500 (highest), PRWCX, VWELX, VWINX

Total Return (TWRR)  PRWCX, SP500, VWELX, VWINX

These results were in line with expectations. Note that while PRWCX is often in the moderate-allocation category, it is a capital appreciation fund with the goal to beat the SP500 with less volatility, and it has been successful in that.

PV run is from 01/1987 (without PRWCX, it could run from 01/1985)  PV RUN2    PV RUN2 (Inflation-adjusted)

[Admin/YBB]

Final Values With X% Withdrawals (FVXW Ratios)

Define

FVXW = Final Value with monthly withdrawals at X% annualized
FV0W = Standard FV with reinvestments (i.e. with 0% withdrawals)
FV5W = Final Value with monthly withdrawals at 5% annualized

These final values can be stated as ratios that are for per unit of initial values. However, calculations in Portfolio Visualizer (PV) are more accurate with larger initial amounts, so the accuracy is improved for the initial amount of 100,000 (vs 10,000 or 1,000). One obvious reason is that PV allows only integer (whole) withdrawal amounts xyz or xyz.00; likewise, the initial and final amounts are also stated as integer (whole) amounts. This technical detail is only for PV, but not a limitation of the concept.

Using the previous examples:

XLU, PDI, VNQ and benchmark VFINX, 6/1/12-5/31/23, PV RUN, for FV5W and TWRR,
VWINX, VWELX, PRWCX, and benchmark VFINX, 1/1/87-5/31/23, PV RUN2, for FV5W and TWRR,

and noting that point-to-point TR = TWRR with withdrawals (CAGR is less meaningful in this case and TWRR > CAGR), or CAGR without withdrawals (TWRR isn’t shown as it would now be same as CAGR), the following table can be constructed. The last column for FV0W was obtained from new PV runs without any withdrawals, or it could also be calculated manually from the TR.

Name               FV5W          TR             FV0W

6/1/12-5/31/23

XLU                  1.7759         9.20%       2.6338
PDI                  1.9324         9.32%       2.6659
VNQ                 1.2927         6.38%       1.9746
VFINX              2.8331        13.14%      3.8894

1/1/87-5/31/23

VWINX             7.9554         8.02%     16.6125
VWELX           13.7740         9.34%     25.8594
PRWCX           26.5726       11.06%     45.5363
VFINX             21.1596       10.37%     36.3565

Note that, these final value ratios are very sensitive to the returns realized (TWRR or TR, CAGR, MWRR – not shown here but is the standard IRR seen in the PV runs).

Although Vanguard has eliminated the older Investor class VFINX, and that ticker is not even recognized at Vanguard website, other data sources, including PV, continue to provide up to date data for VFINX. Replacement Admiral class VFIAX doesn’t have as much history (only since 12/1/00 at PV). Vanguard must still be providing the current data for VFINX to data vendors.

[Admin/YBB]

Withdrawal Rates With Final Value 0 or Initial Principal (Inflation-Adjusted)

Portfolio Visualizer (PV) SWR (Safe Withdrawal Rate) is the percentage of the original portfolio balance that can be withdrawn at the end of each year with inflation adjustment without the portfolio running out of money (dollar amount withdrawal).

$AFI is the initial principal amount needed to support $1/month initial withdrawal that is adjusted annually; $AFI will be completely exhausted after the term. It is found from PV SWR as follows,

AFI = 100*12/SWR .

$AAFI is the initial principal needed to support $1/month initial withdrawal that is adjusted annually, and at the end of the term, the leftover amount is inflation-adjusted P,

AAFI = AFI*CFI/(CFI - 1) ,

where, CFI = Final Balance (inflation-adjusted)/Initial Balance. It is available from PV run.

Lower withdrawal rate SWRM (SWR-Modified) corresponding to AAFI is then,

SWRM = 1,200/AAFI = SWR*(CFI - 1)/CFI .

Another less common approach is available from PV runs that is based on %withdrawals at yearend that will leave the final value as inflation-adjusted initial principal. 

Fixed Annual Percentage PWR – This is the fixed annual % from annual balances that will leave the final value FV that is at least equal to the inflation-adjusted initial lump-sum. The annual withdrawals will be variable depending on portfolio performance.

These summaries are from other post on THIS site where more details are available.
ybbpersonalfinance.proboards.com/post/1086/thread
ybbpersonalfinance.proboards.com/post/1249/thread

[Admin/YBB]

MFO Post www.mutualfundobserver.com/discuss/discussion/comment/169522/#Comment_169522

PV - SWR, PWR (& SWRM)
yogibearbull
(11/16/23)

An update.

PV PWR is less commonly used, so I have stopped using it in more recent data.

PV SWR is the percentage of the original portfolio balance that can be withdrawn at the end of each year with inflation adjustment without the portfolio running out of money (dollar amount withdrawals).

New SWRM is the percentage of the original portfolio balance that can be withdrawn at the end of each year with inflation adjustment but at the end of the term, the leftover amount is inflation-adjusted original principal (dollar amount withdrawals).

Three examples below use the start dates of 01/2007 (Pre-GFC), 01/1987 (Year of 1987 Crash), 01/2000 (Year of Dot.com bubble burst). The idea is to select "bad" starts of withdrawal programs in recent history and see how various hybrid funds did; VFINX/SP500 is included for benchmark comparison. The end date is common 10/2023 (so, these data are not comparable to those in the examples in the June 2023 Post).

Example 1 - Period 01/2007 - 10/2023 that included GFC 2008-09 and pandemic 2020.

FUND    SWR      SWRM    SWRM (corrected)
VWELX  8.63%   4.35%    4.37%
FBALX   8.17%   4.28%    4.09%
ABALX  8.27%   3.92%    3.99%
VFINX   8.05%   5.02%    4.82%

Example 2 - Period 01/1987 - 10/2023. Limited by 11/1986 inception for FBALX, although the (free) PV data goes back to 01/1985.

FUND    SWR      SWRM    SWRM (corrected)

VWELX  7.89%   7.03%    7.035%
FBALX   7.43%   6.58%    6.545%
ABALX  7.63%   6.69%     6.70%
VFINX   8.82%   8.15%    8.106%

Example 3 - Period 01/2000 - 10/2023 captured the dot.com bubble burst and was possibly the worst start for withdrawal programs in the recent history.

FUND    SWR      SWRM   SWRM (corrected)
VWELX  7.25%   4.80%   4.82%
FBALX   6.83%   4.58%   4.47%
ABALX   7.40%   4.83%   4.865%
VFINX   4.34%   2.52%   2.40%

If higher withdrawal rates than indicated are used, then the goals would be missed. But if lower withdrawal rates are used, then the balances at the end would be higher than the goals. The goal for the SWR is $0 balance; goal for the SWRM is the inflation-adjusted original principal.

It is interesting that all-stock SP500 did the best in Examples 1 and 2, but the worst in Example 3. That is the danger in using all-stock portfolios in withdrawal programs - they may work much of the time, but may bite occasionally.

Another point is that the "bad" times are more recent, and don't go back to 1920s or 1930s that included even worse times. And the withdrawal rates these funds sustained in the recent history are higher than the typical Bengen Rule of 4% inflation-adjusted withdrawal.

ybbpersonalfinance.proboards.com/post/1249/thread

Edit/Add, 11/19/23. SWRM (corrected)

For data consistency, the PV runs should be to 2022 (recent full year), not 2023 (partial year in 11/2023). While SWR is unaffected, SWRM is affected and is corrected.