Define "good"

[Andre1969]

Quote:

Originally Posted by REWahoo Wow. At the 4 year point (mid 2009) I was down about 25%.

I think it's more of a long, steady, year-after-year decline that you have to worry about, rather than one sharp drop. If you retired in mid 2005, was your net worth up by mid 2006, 2007, and possibly even 2008?

A few days ago, I ran a FireCalc scenario of retiring in 1999, with $1M, and a 3% withdrawal rate. As of 2013, the last year in the Excel spreadsheet, the ending balance would have been lowered to $876906. Adjusting for inflation, the 3% initial rate, which would have started you at $30K way back in 1999, would be up to $41,573 for 2013.

2014 was a fairly lackluster year for me. I think my return was about 5.6%. And as of this past Friday, I was only up around 4.2% for this year. So, while those percentages would be enough to offset the 3% SWR rate, the ending balance as of 2015 wouldn't be that much higher than 2013.

So, I'm guessing that the 1999 cycle is one of the few that might not pass the 3% SWR test, and at 4% it would probably be doomed to failure.

Still, considering that cycle contains two serious recessions (I lost about half of everything in 2000-2002 and again in "The Great Recession), it's probably a fairly rare occurence.

[REWahoo]

Quote:

Originally Posted by Andre1969 If you retired in mid 2005, was your net worth up by mid 2006, 2007, and possibly even 2008?

Up in 06 and 07, not in 08. By 2013 it did recover to my nominal starting balance.

[Andre1969]

Quote:

Originally Posted by REWahoo Up in 06 and 07, not in 08. By 2013 it did recover to my nominal starting balance.

Yeah, I figured 2008 would be an iffy one. Depending on what part of the year you were picking, it could be possible. For instance, in mid 2008, I was up compared to mid 2007. Sept/Oct/Nov of 2008 took care of that really quickly, though.

[Danmar]

I find this an interesting question. The most difficult issue for me is my starting point. Although I retired in 2006 at 56, I still had significant incentive compensation that paid out over the subsequent 5 years. During this period I spent a lot of this comp on real estate My pension started in 2012 when I hit 62 and that is really the beginning of only living off pension and portfolio.

So that is probably the best starting point. Since mid 2012, our portfolio is up about 27%. That makes me feel pretty good.

[Independent]

Quote:

Originally Posted by MrLoco Isn't this where a "healthy" bucket of cash comes in handy? If the first few years post-retirement result in negative returns: -- Reduce the withdrawal rate and supplement from cash accounts. -- Maintain the same withdrawal rate but do not adjust for inflation again supplementing from cash accounts if needed. For time periods such as end 2008......suspend withdrawals and take from cash accounts if possible. Of course this requires a large balance in cash/short term investment accounts.

I think the issue is:

"I don't want to reduce spending as soon as the market has a little drop. That's why I started with a 4% SWR, so I'd have some cushion to avoid immediate cuts like that. But, if things are bad enough, long enough, I figure I'll eventually have to bite the bullet. What is 'bad enough'? "

Or maybe, "Historically, if people had started with a 4% SWR, how often would they have found that they had to cut spending at some point?"

[tmm99]

Is there a calculator that mimics this scenario (mrket being down for the first 3-5 years of your retirement)?

Also, is there a calculator that mimics Nikkei Index? (Wait, forget I asked. I just looked at the historical chart, and most of us would not be able to survive the sustained drop they experienced for the 20 some years. I would have to reply heavily on SS at that point.)

Firecalc tells me it will be 100% successful with only 60% of my current asset for my expenses, but I still wonder how I would fare if the market stayed down for a prolonged period of time (lasting past my cash/short term reserves were all exhausted).

[jIMOh]

To me "good" would be the portfolio generated a gain equal to income needed. If I needed $35k, and portfolio generated $36k or more, that is "good", if it generated $34k or less, that is "bad". Meaning if portfolio in first X years has more than it had in year X-1, that is "good".

I have read that the market recovers in 3 years just about every time, so I think the right "length of time" is 3 years for X.

[youbet]

Quote:

Originally Posted by REWahoo This may help. Below is a chart from the front page of FIRECalc giving examples of a 750,000 starting portfolio withdrawing $35k per year. "Good"(green - retired in 75), "OK" (blue - retired in 74) and "Lousy"(red - retired in 73). You can check out how the market performed in the first few years following each retirement to see what "good" looks like.

I believe that the biggest influence, more than market performance, on the crashing red line was several consecutive years of high cpi. Both were factors, but cpi was extraordinarily high.

[walkinwood]

Quote:

Originally Posted by sengsational I agree that the models that are typically run don't adjust spending for down markets. This is a good model for people who's budget include zero discretionary spending, but that's not most of us here. When markets are down / there is an economic slow down, everyone, not just the retired, pull back, so it's not as hard to cut back on discretionary. The way the models usually work is you plan once, at the beginning of retirement, then blindly do what it says until the end (not realistic). The reality is that we run the model every year, if not more often, and if changes are called-for to stretch it out a bit, those changes can be made.

There are variable SWR models out there too like the 4/95 plan that Bob Clyatt laid out. We use a percentage of our portfolio at the beginning of the year as our max withdrawal for the year.

I liked the Guyton "decision rules" which allow both an increase and decrease in withdrawals based on portfolio performance and inflation. These have been discussed on the forum before

http://citeseerx.ist.psu.edu/viewdoc...=rep1&type=pdf

[NW-Bound]

After you have been in retirement for some time, say 5 or 10 years, how much should your portfolio be in order to ensure that you are on the right track? It occurs to me that it is easy to use FIRECalc to explore this. Here's what I did.

1) Set portfolio size to $1M. Leave the "Years" at 30. Portfolio mix to a 50/50 mix between total market and long interest rate, everything else in default value. Use "Investigate" tab to search for the max WR to support a 30-year retirement. Answer: $37124/yr.

2) Now go back to the "Start" page, set the spending to $37124. Set the "Years" to 25 instead of 30. Then, use "Investigate" tab to look for the portfolio size to support this spending for 25 years. Answer: $915K.

3) Repeat for 20 years, 15 years, etc...

Results:

30 years: $1M
25 years: $915K
20 years: $800K
15 years: $657K
10 years: $489K
5 years: $269K

Note that FIRECalc reports everything in inflation-adjusted dollars, so in nominal dollars you will need more than the amounts shown above.

It is interesting to see that the portfolio size is not proportional to the retirement duration. It takes a lot more money in the short-term than in the long run. Take the shortest duration of 5 years for example. At a WR of $37124, you only draw $37124 x 5 years = $185620 in 5 years. Yet, to die broke after 5 years you need to start out with $269K, or 45% higher. Yikes!

The above shows that in the short term, market crashes can really devastate a portfolio and leave you with no time to recover. Time is not on your side. If this bad short-term interval happens early in your retirement, then we have the bad sequence risk.

[NW-Bound]

OK, here's another way of looking at the above results. With the first result of $1M start value, $37124 withdrawal, 30-year period, you get to spend a total of $37124 x 30 = $1114K. That's only 11% above inflation for the worst case.

Moving on to 25 years, you get to spend $37124 x 25 = $928K, compared to start value of $915K. Keep going down the lines, and it gets very discouraging.

30 years, $1000K, $1114K
25 years, $915K, $928K
20 years, $800K, $742K
15 years, $657K, $557K
10 years, $489K, $371K
5 years, $269K, $186K

So, it looks like for 20 years or less, to protect against the worst case you are better off just looking for something that matches inflation.

Hey, Daylatedollarshort, are you there?

[walkinwood]

Nice analysis, NW-Bound.

Those are good numbers to check your portfolio values against.

[NW-Bound]

I made the above posts, then went upstairs to continue putting down some laminate floor boards. There was something bothering me, and I thought about it some more. In the process, with my thoughts diverted, I miscut a couple of boards.

I am taking a break now, and have figured out the fallacy in the above runs. You cannot break a 30-year retirement period into segments, then run FIRECalc on each segment. The market return is not a random walk! There's correlation between successive periods. After a sequence of bad years, the following years saw the market rebound with above average returns. By breaking a 30-year period into segments and looking for worst cases, what we see are successive bad segments one after another.

Here's something to help. At the beginning of retirement, the SWR for a 30-year retirement is $37,124 on a $1M 50/50 portfolio. This is 3.7% WR.

At the mid-point, when we look at what it takes for a 15-year retirement and drawing $37,124, FIRECalc tells us we need $657K. But $37K withdrawal on $657K is 5.6%! Going down to 5 years, FIRECalc says we need $269K, but the WR of $37K has gone up to 13.8%.

What happens to our original 3.7% WR? Confusing, is it not?

The values in post #31 still stand, but they are worst-case values. If you start out at 3.7%WR, and at year 15, you find yourself drawing 5.6%, that means your first 15 years were lousy, and chances are that the next 15 years will not be as bad.

By rerunning FIRECalc with the shorter period and with a higher WR, you are subjecting yourself to another bad 15 years, and historically that has not happened. Not in the US so far anyway. Keep our finger crossed.

[RetireAge50]

I think there are ways to protect against the worse case while still allowing for the average and upside to occur. Classifying the first few year's returns as good, bad, or otherwise mostly only applies if you use a predefined/non-dynamic withdrawal strategy.

[kcowan]

If you truly believe in LBYM, a couple of down years forces you to cut back. Then when the market corrects, live is good. We retired in 2003 so we knew about the 2000 market meltdown and we had to adapt to that new reality. 2008 was a bonus because it was not permanent!

And our budget run was great from 2003 to 2008! As luck would have it, in 2007 we decided to buy a place in Mexico for 6 months snowbirding and our budget declined by 40% without sacrifice so we concluded that we can live forever!

It is ironic that the decision to enjoy life in the winter solved the budget problems. Maybe unique to Canadians because the COL is relatively high. And then the property we bought was at C$1.065 and now the loonie is at $0.73x.

[NW-Bound]

It occurred to me that to look at the vagaries of the market, there is a much simpler way to use FIRECalc. One simply sets the WR to 0, and see how a portfolio shrinks in the worst case after 5 years, 10 years, 15 years, etc... I have done that before, but old age has reduced the superiority of my memory (and I am not quite 60 even).

You can do the above and see the results are really bad in the worst case. And it does not matter whether you are 100% stock, 100% bond, or balanced in that doomsday scenario. See for yourself.

[sherrywilliam]

The market has not been good to us this year so far we are down less than what we started with. And it is first year of distributions.

[W2R]

Quote:

Originally Posted by sherrywilliam The market has not been good to us this year so far we are down less than what we started with. And it is first year of distributions.

I didn't notice that but you're right! Share prices for the funds I have are down, too. But don't forget, December fund distributions are coming up and for some funds they can be substantial. So, that may help.

My portfolio is down about 4% YTD. I can accept that, especially because I sold funds to pay for buying my dream house in cash, fixing it up, moving into it, and selling my former home.

[petershk]

I consider "bad" when I am unable to adjust my spending to match the economic conditions.

Before ER that means employment income - expenses > 0 every year (usually saving around 50%). That 50% saving target allows for all kinds of terrible things to happen.

Post ER it's total return - expenses > 0 (inflation adjusted). Of course that's ideal and market fluctuations tend to be more turbulent than job fluctuations.

That said... I try to look at my target SWR (3%) and then see what 3% looks like in terms of historic income changes. For example if I have a 90/10 stock/bond mix... My income fluctuates about 50% in 08/09. That's pretr brutal. Could I cut my ecpenses in half? Not easily.

During stagflation real income drops a couple % a year... But over a long period. That also would have been tough to survive.

But if I have fairly broad flexibility I can at least prepare in advance what I will do if there's a big swing or a long, protracted dead period.

Sent from my HTC One_M8 using Early Retirement Forum mobile app

[audreyh1]

The last few years we've spent less than our withdrawal, and built up a significant surplus in short-term accounts. So I guess that means we're already ready for a "bad" year or two when we draw less from the portfolio.

We'll be drawing a little less next year as our portfolio is more or less flat and thus hasn't recovered the Jan withdrawal. Unless there is a heck of a Santa Claus rally!