It's unsafe now.

Financial Independence/Retire Early -- Learn How!
Post Reply
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

It's unsafe now.

Post by JWR1945 »

It's unsafe now.

KenM and Ben have asked some questions that I want to address on this thread. I will start by reposting a couple of my previous posts at the Motley Fool. They were on my It is safe now thread back in July of 2002. The S&P 500 had fallen to 900 and I asserted that it was OK to use the 4% rule on that date.

I will also repost what I consider to be the best single post on this topic. It was written by BenSolar and posted on our IndexFunds discussion board. Our IndexFunds people know a whole lot about asset allocations and they are worth listening to. Although BenSolar's post was written in terms of the accumulation phase of investing, I think that his comments on allocations are important during the distribution phase (retirement) as well. The 4% number is still a good number, but not if you apply it to the S&P 500 today. You can make other choices. They will recover the deficit.

I will answer Ben's question first. He asked (on the Future Proceedings on the SWR Matter thread):

That said; has anybody actually MADE the study of SWR including the 3 tier valuations at starting year? (low P/E below 10, medium P/E (around 10), high P/E (above 10) - or whatever levels found fair).

Is is hard to do?

The answer is yes. Guess who? You are right. It was BenSolar. And he had help. He even had help by some who deny his conclusions.

Have fun.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Here is the first post about this that I put up at the Motley Fool's discussion boards. Today's S&P 500 index value is greater than 900.

Subject: It is safe now. Date: 7/16/02 11:08 AM
Author: JWR1945 Number: 70457 of 108802

It is now safe to use the 4% safe withdrawal number.

Does that seem strange to hear? Consider this. The Retire Early Safe Withdrawal Rate study and its related calculators produce results that are accurate to the extent that the historical data can be used to predict the future. When the calculators state that a certain withdrawal percentage and investment mix produce a 100% safe withdrawal rate, that estimate is constrained by the underlying logic behind the calculation.

There is quite a bit of merit to assuming that the next three or four decades of investment results will fall within the bounds of history. What has bothered me, however, is that...as I have looked more and more deeply into the matter...I keep on finding things that really are new...or, more accurately, they really were new when they first happened. The events of both the depression era and the stagflation era were new when they wiped out retirement portfolios. It always seems to have been the unanticipated events that have caused disasters. Even real estate has been hit. At one time it was thought that the prices of homes would always rise (with a few obvious exceptions, of course). Then there was the disaster in the oil patch. So although having the safe withdrawal rate study fail has been likened to going through a nuclear war, I do think that it is worthwhile to consider such a possibility.

There has been much talk about the high valuations afforded stocks in the last decade. And although there is credible evidence that the Price to Earnings (P/E) ratio of the S&P 500 is a poor measure of valuation for predicting the safety of a withdrawal rate, I do use it here...because I don't have anything better.

What is different now...or, actually, what was different until recently...is that the S&P 500 P/E is higher than ever before and for an extended length of time. I refer to (Professor) Shiller's data that intercst has provided to us. (See post 67026 dated 5-20-02.) IIRC, Dr. Shiller averaged the earnings over ten years to smooth the data. Averaging also tends to reduce the overall P/E number. The P/E had never exceeded 30 except for two months in 1929 data prior to 1997. It stayed above that level for about 4 years. Today, of course, the P/E ratio of the S&P 500 is within normal bounds. (I have not been able to calculate this exactly via Dr. Shiller's method because I do not have recent earnings data. I have estimated the number in two different ways. The answers are both close. The simpler method is to take Dr. Shiller's December 2001 values of the S&P 500 and its P/E and scale by today's S&P 500 number...roughly speaking, this is (900/1140)*(30.27) = 23.90.) In addition, dividends are becoming popular again (at least for the many investors on this board that have bought REITS for retirement income and for others buying stocks inside IRAs and other tax favored accounts). I expect corporations to respond to this change. I expect dividends to increase.

Simply put, I think that things are normal again...at least in regards to the Retire Early Safe Withdrawal Study. I think that those who retire now and withdraw from their investments at rates according to the study really are safe and they really can feel secure. I also anticipate that they will be able to increase withdrawals via the Pay Out Period Reset (POPR) and/or the other standard approaches. Do realize that I am talking of a Safe Withdrawal Rate in a generally understood sense. This differs from the use of the phrase (along with its precise definition) as found in the study.

Have fun and Fool on.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

This is the other post that I put on the Motley Fool's discussion boards about its then being safe to use the 4% rule (when the S&P 500 index had fallen below 900). If I had written this today, I would have been much more careful when I said: to the extent that historical results do predict the future accurately. That qualifier needs to be made much more restrictive.

Subject: Re: It is safe now. Date: 7/19/02 10:54 AM
Author: JWR1945 Number: 70783 of 108802

ariechert made an excellent observation when he wrote the following on this thread. It is a sanity check.

So if I started to take out 4% on March 10, 2000 I would be ok now? How about if I start to take out 4% of the March 10, 2000 value in March of 2003, even though my portfolio has lost 35% of it's 3/10/2000 value?
- Art

The underlying technical issue involves extrapolation. You can generally have a very high confidence if you make a small extrapolation. As you make larger extrapolations, however, your confidence level decreases.

There are credible reasons to consider the possibility that historical results will not predict the future. You can find many of these from the sources referenced in the link that intercst has provided in post 70676 dated 7-17-02 on "Future Stock Market Returns."￾ Of course, to the extent that historical results do predict the future accurately, it is by definition always true that an estimated 100% safe withdrawal rate number remains 100% safe. However, that information does not mitigate risk in case the extrapolation fails.

There are numerous questions that are worth answering. In all cases we want to mitigate risk in the event that the historical data fails to predict the future. Here are some:
1) How can you recognize that something is going wrong?
2) How soon will you know that you must take remedial action...such as working part time?
3) Do we have to take preemptive action...such as delaying retirement?
4) What changes should you make to your withdrawal approach (or tactics or strategy or policy) based on new information? (There have been several excellent approaches developed so far. Certainly, there can be many others...each appropriate for different individuals.)

It is important to understand that credible risk issues exist. It is much better to address them early than to ignore them. Opportunities abound for doing useful research and for providing useful information.

Having fun,

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

As promised, here is BenSolar's post on the 100% stocks?? thread on our IndexFunds discussion board.

Posted: Thu Apr 24, 2003 8:34 am Post subject: Re: 100% Stocks??

Trex wrote:
Right now, we hold 2 funds. VTSMX and CREF's S&P500 fund. We only have the money for 2 funds. I was going to immediately convert the Vanguard to a bond fund, but someone suggested that I stay in all stocks considering my horizon of 30+ years. Any thoughts on that?


Greetings, Trex

I'm sure you'll get lots of good thoughts, but it looks like I get to go first. I'm assuming that the assets are tax sheltered.

First, I think that with the 30+ year horizon you would come out perfectly fine (or better ) if you dollar cost average into these funds and add other diversification as you can. The thing to weigh in your mind is this: will you continue to DCA into these funds in the face of another 30-50% decline? If you are certain you will, then no problemo, you will come out great. But if you get rattled and bail in a steep drop, then you would damage your long term returns.

On the other hand, I've become convinced that a good valuation measure like PE-10 (Price/10 year rolling average earnings) or dividend yield does have some long term predictive power for future long term stock market returns. I can point you to studies if you are interested. Current valuations of US stocks remain high compared to average. So, I've chosen to overweight other asset classes like REITs, International, and I have more in bonds/cash than I really want because my 401k doesn't have great options. I intend to move that MMF into S&P 500 (of which I do keep some) when valuations improve. A dangerous game, I know.

So, with the valuation thing in mind, and with the possibility that more strong volatility to the downside would cause at least some people to rethink being 100% in US stocks, I might recommend a bit of diversification away from that asset class.

There are various options. The Vanguard Balanced Index would give you some exposure to bonds, as it is an automatically rebalanced 60/40 TSM/TBM index blend. That gives you the sell high/buy low benefit of rebalancing, and it will reduce volatility, while still offering most of the higher returns that stocks should offer over bonds. This could be a great choice.

Another option, is making a REIT index your other choice. Vanguard REIT Index: VGSIX offers a dividend yield of about 7%, lower volatility than the traditional equities, and low correlation to traditional equities. Growth should be about in line with inflation, or maybe 1% less according to raddr's study of their limited history. A real return of 6% or so over inflation is looking really good compared to the S&P 500's predicted real return of 3-5% using the Gordon equation. You lose the rebalancing bonus, though, since you won't be able to rebalance across funds until you can add another at Vanguard.

Finally Vanguard offers some other funds of funds that have broader diversification than the Balanced fund. I'm not familiar with them, but from what I've read they might be a good 1 stop shop for instant diversification. Maybe someone can chime in with more info on these?

Well, those are the thoughts from this muddled mind, Good Luck
Ben
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Now to address KenM's concern, although not adequately. Refer to the Alternative SWR Strategies-Part 2 thread (Posted: Tue May 27, 2003 7:40 am).

KenM
Our more immediate concern is to compensate for valuations well outside of the historical range.

I'm not really sure about that. Extreme valuations occur so infrequently that IMO the only way is to use logic/commonsense as a one-off correction to the model. One of my immediate concerns (may not be that of others) is how to set a starting withdrawal amount when, even in most years within the historical valuation range, the S&P500 varies by upwards of 20%. Should I take the average portfolio value for the year; or the value on the date I retire and apply the SWR or am I able to take the highest valuation. I know there are commonsense answers but it makes a big difference to the starting and subsequent annual $withdrawals. Such a decision would not be easy.

We are outside of the historical range today. I recommend that you start with a 4% initial withdrawal amount based roughly on today's prices. The key thing is to listen to what BenSolar and others are saying and have said about allocations. You can take your time. There is no need to rush.

I will add one more thing for you to keep in mind. It is volatility that wipes out retirement portfolios. Avoid selling heavily when stocks are down. Higher dividends (whether from stocks or REITS) help you do this. Investments with lower volatility can be safer than others during retirement even if their growth rates are somewhat lower. (Unfortunately, I cannot quantify what I mean by saying somewhat lower.)

Have fun.

John R.
[KenM]
*** Veteran
Posts: 133
Joined: Tue Mar 04, 2003 12:54 am

Post by [KenM] »

Now to address KenM's concern, although not adequately

John R
I think that's my point - there is no adequate answer.
Thinking aloud and forgetting exceptional valuations, in a typical year in January my portfolio may be worth $800,000; in July $900,000; I retire in August; at 4% SWR do I take $32,000 or $36,000?

$36,000 is 4% of $900,000 but 4.5% of $800,000
$32,000 is 4% of $800,000 but 3.5% of $900,000
That's almost 30% difference in SWR's.
.......beginning not to believe in SWR's - a few months ago I was quite excited to have found them :D
KenM
Never try to teach a pig to sing. It wastes your time and annoys the pig.
User avatar
BenSolar
*** Veteran
Posts: 242
Joined: Mon Nov 25, 2002 5:46 am
Location: Western NC

Post by BenSolar »

KenM wrote: I think that's my point - there is no adequate answer.
Thinking aloud and forgetting exceptional valuations, in a typical year in January my portfolio may be worth $800,000; in July $900,000; I retire in August; at 4% SWR do I take $32,000 or $36,000?

$36,000 is 4% of $900,000 but 4.5% of $800,000
$32,000 is 4% of $800,000 but 3.5% of $900,000
That's almost 30% difference in SWR's.


But it's only a difference of 11 or 12% in actual withdrawal ...
.......beginning not to believe in SWR's - a few months ago I was quite excited to have found them :D


Don't give up hope, Ken

One approach is to use a one, two or three year average value to determine your SWR. It's been noted elsewhere that using a multiyear average in historical analysis gives you a higher computed safe withdrawal rate from a lower value. But it helps weed out temporary spikes in valuation that might put you at risk. The 2000 peak was so big that it likely would have defeated even this approach. :(

But if you use a year long average value and you make certain that valuations are well with the historical range (i.e. below 20 as an arbitrary number), you should be pretty safe using a number that worked historically.

On another note, I think that when values are within a 'normal' range then the fact that stocks might have dropped and the supposedly safe 4% figure you started with is now 5% isn't such a dilema. Valuations are lower, dividend yields are higher. Even if valuations stay low, you've got the dividends working for you ... Now if you started when dividend yield was a pathetic 1% on the SPY, you've lost 1/3 your capital and the yield is still a pathetic 2%, then you might be in trouble!

Ben
"Do not spoil what you have by desiring what you have not; remember that what you now have was once among the things only hoped for." - Epicurus
Post Reply