Capital Market Expectations and SWR

[REattempt]

Wade Pfau has an article coming out this month in the Journal for Financial Planning:

Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates

He posits that using capital market expectations (returns, deviations and correlations) along with asset allocation to optimize SWR....AND predict some level of reasonable SWR based on longevity etc.

Nice read.

He also has a follow-up in his blog with more information...

Pensions, Retirement Planning, and Economics Blog: Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates: Additional Results

The two most interesting parts are:

1) the impact of fees (see figure 9 in the blog post) is big
2) that optimization occurs at relatively low stock allocations

THE most interesting personal observation is that the SWR for a 40 year retirement and 1% failure rate is 2.8%.

2.8% is the WR I established my budget against and put in my Withdrawal Policy Statement!!!

[Lsbcal]

From the conclusion of the paper:

Quote:

Planners can use this framework to translate their own forecasts for capital markets into withdrawal rate and asset allocation recommendations for clients.

Forecasting capital markets can be done by anyone. Doing it successfully ... not practical.

[REWahoo]

Planners can use this framework to translate their own forecasts for capital markets into withdrawal rate and asset allocation recommendations for clients.

Quote:

Originally Posted by Lsbcal From the conclusion of the paper: Quote: Planners can use this framework to translate their own forecasts for capital markets into withdrawal rate and asset allocation recommendations for clients. Forecasting capital markets can be done by anyone. Doing it successfully ... not practical.

Yep. I cannot imagine basing my "Safe" WR on what a planner "thinks" the capital markets will do in the future. Talk about rolling the dice...

[Lsbcal]

All we have to do is look at the physical world we live in. Most good physicists nowadays believe the Heisenberg Uncertainty principle to be valid. There is no way to predict the future because we live in a fundamentally uncertain universe.

So you don't have to worry about (1) fate, (2) how your evil thoughts will influence the direction of the world or tomorrow's weather, (3) what the capital markets will do 1 millisecond from now.

The truth will set you free!

[REWahoo]

Quote:

Originally Posted by Lsbcal The truth will set you free!

The parallel to not wearing pants is almost eerie...

[pb4uski]

Quote:

Originally Posted by REattempt THE most interesting personal observation is that the SWR for a 40 year retirement and 1% failure rate is 2.8%. ....

That sounds about right. If you accept that 4% is a SWR for a 30 year retirement, that equates to a minimum 4.2% annual return on your nestegg assuming 3% inflation. If you accept the 4.2% investment return and then extend the projection from 30 years to 40 years and solve for the withdrawal rate it would go down to 3.2% which is close to 2.8%.

[REattempt]

Very interesting responses...I thought I would get skepticism (I love a reactions to heresy [is there a better word?]...you guys never disappoint!

Lsbcal/ReWahoo...I get that predicting the capital markets is a crap shoot. The article used the SSBI historical data. What I found interesting is that the resulting rate was lower than OTAR and firecalc. I had already reduced "my" SWR 40bps from the most conservative 3.2% (OTAR). I would love to hear what you think is a SWR ReWahoo and what you base it on...?

[brewer12345]

Quote:

Originally Posted by REattempt Very interesting responses...I thought I would get skepticism (I love a reactions to heresy [is there a better word?]...you guys never disappoint! Lsbcal/ReWahoo...I get that predicting the capital markets is a crap shoot. The article used the SSBI historical data. What I found interesting is that the resulting rate was lower than OTAR and firecalc. I had already reduced "my" SWR 40bps from the most conservative 3.2% (OTAR). I would love to hear what you think is a SWR ReWahoo and what you base it on...?

The safest withdrawal rate is 0%. Good luck.

[REWahoo]

Quote:

Originally Posted by REattempt I would love to hear what you think is a SWR ReWahoo and what you base it on...?

A truly safe WR is the maximum you can withdraw to pay everyone but the undertaker.

As to what that amount/percentage is, I think we would all be equally served debating how many angels can fit on the head of a pin - it is an unknowable number. There are far too many unknowns and variables - thus my issue with the article.

Certainly the lower the percentage the less chance you'll run out of money before you run out of you. Your 3.2% is as good a guess as any.

[clifp]

I think Wade is smart guy and enjoy his blog. However, I fundamentally disagree with him using Monto Carlo simulations even if he has correctly modeled past market performance.

The performance of the stock market year to year is not independent of the previous years performance. Now I know their are some who disagree with me on this.

However, I feel very strongly that there was no way in hell that after the crazy bull market in 98-2000 that the market performance 2001 or 2002 were equally like to have a good year as "normal" years. Likewise it was clear to me and many others on the board that 2009, or 2010 were almost certain to be good years for the market after the collapse of 2008.

I believe the consequence is that his model will show worse case scenario bears market which are deeper than we can historically expect which exaggerate the impact. Wade's SWR are consistently lower than other studies I think it is the Monto Carlo simulation which is causing it.

[NW-Bound]

Quote:

Originally Posted by clifp The performance of the stock market year to year is not independent of the previous years performance. Now I know there are some who disagree with me on this.

Not I! I am no economist, but not afraid to share my layman's observation as follows.

Apologies to Mr. Malkiel who authored "A Random Walk down Wall Street" which I truly enjoyed, but if the market were truly a random walk, we would not be talking about the phenomenon of "reversion to the mean".

The head/tail sequence of a tossed coin is a true random walk; the coin does not remember its history. The market however does. For example, several years of a bull run, if not accompanied by a corresponding growth of the economy (which depends on demographics, technology advances, population growth, etc...), would cause the P/E ratio to grow beyond the norm. This then means that the risk of a reversion-to-the-mean movement will grow with time, until a market crash (or a Wh** announcement ) occurs to take away the excess.

It is difficult to predict the exact moment when the scale would start tipping. However, the probability of a correction would grow with the increase of the unbalance.

Economists have been working on ways to repeal the market or business cycle, but so far I have seen no real convincing evidence that they have achieved much.

PS. My comments were simply meant to show agreement with Clifp's statement above. I have not read the OP's cited article by Wade Pfau to know what the latter actually said.

[Lsbcal]

Quote:

Originally Posted by REattempt Very interesting responses...I thought I would get skepticism (I love a reactions to heresy [is there a better word?]...you guys never disappoint! Lsbcal/ReWahoo...I get that predicting the capital markets is a crap shoot. The article used the SSBI historical data. What I found interesting is that the resulting rate was lower than OTAR and firecalc. I had already reduced "my" SWR 40bps from the most conservative 3.2% (OTAR). I would love to hear what you think is a SWR ReWahoo and what you base it on...?

Hi REattempt, I did not mean to be overly critical about the article. Sorry if I came across that way to you.

Luckily I don't have to solve the general case or come up with a set of SWR choices based on a lot of variables. I think all the studies that are well done are helpful. For me, being in the drivers seat and actually retired is a whole lot different from theoretical calculations of SWR. These studies are at least a good starting point, especially if you are not yet retired and trying to sort things out.

[Midpack]

Interesting read, thanks!