Firecalc says my portfolio reaching critical stage

[DblDoc]

Does anyone have a link to this interview? I'd like to see what he said in context. I think there is a difference between "cannot afford to lose another nickel they MUST sell " and "cannot stand to lose another nickel they MUST sell ". I'm curious to see which he meant. I'd tend to reluctantly agree with the former, but not with the latter. I googled around but couldn't find it.

This is the NYT article from the past weekend: http://www.nytimes.com/2008/10/26/business/26bogle.html?_r=2&pagewanted=1&ref=business&oref=slogin

It is cannot afford to lose...

DD

 

[harley]

Nevermind, I think I found it (CNBC, not CNN). Expert Advice: Bogle, Cohen, Doll, McCulley Weigh In - Financials * Europe * News * Story - CNBC.com If this is the same one, I think you misinterpreted his words. (My emphasis below)

"This is my tenth bear market and what's different about this one is that the problems of the financial markets are spreading over to the real economy," Bogle said.

"If you're a speculator, I have absolutely no advice to give you," Bogle said. "Get out of the market now."

But investors allocating their assets intelligently -- with diversified stocks and a reasonable position in bonds -- should "fight it out" and not change their strategy, he said.

 

[harley]

Ah, good. Mine was different. But still, what he said was

"So this isn’t the time to sell, he said, but he allows one big exception: “If you cannot afford to lose another penny, then you simply have no recourse but to get out of the stock market.”

I still don't see that as a recommendation to sell, unless you are in an absolutely critical situation. I didn't get the impression that was true for the OP.

 

[Bestwifeever]

This is the NYT article from the past weekend: http://www.nytimes.com/2008/10/26/business/26bogle.html?_r=2&pagewanted=1&ref=business&oref=slogin

It is cannot afford to lose...

DD

Thanks for tracking that article down, DD--it is interesting to read.

 

[52andout]

Cashflo, your last post said volumes. It looks like you can still have a great retirement you just need to set priorities. You note some places that you are spending more money, but this is discretionary spending. sometimes you need to take a look at what you are spending vs. what you can afford to spend.
Most of us do it all the time but don't even think about it. Putting off buying a new car for instance or cutting down eating out.
If your lifestyle includes going out for coffee and meals, driving more etc. there are probably lots of ways you can decrease your expenses and stack the odds in your favor to continue to live off your savings/pension/SS. Hard choices maybe but also necessary choices we all are facing.

 

[al_bundy]

i might rebalance into 100% cash today

there is a 30 day trading restriction on my 401k and the way it looks like the bottom won't be reached for about that long

 

[FIRE'd@51]

cashflo2u2,

Let me rephrase what I think has been alluded to in a number of the above posts.

Typically with FireCalc, we look at 30-year time periods, the start of which rolls through the years, beginning in 1871 and ending in 1978 (the starting point of the most recent 30-year interval). Some of those periods start near market highs, some near market lows, and some in between. The FireCalc result we all hang our hats upon is the 4% WR which was safe 95% of the time. If one could run FireCalc (and I don't know how to do this without actually going in and looking at the individual 30-year period's data), I am pretty one would find that starting the "retirement clock" after the market had dropped 40%, would lead to a SWR in those instances which is considerably higher than the 4% number (or whatever number one used for a SWR - even a 100% safe one). This is why I think restarting your clock now (even with a 21-year time horizon) could be very misleading, because any failures FireCalc reports are probably the ones that started right around those same market highs.

If you retired in 2001, it looks like the market was down somewhere between 10% and 20% from its 2000 high. By choosing a SWR that was 100% safe back then, it seems to me you have most likely been sufficiently conservative.

This is my understanding of the way FireCalc works - I'm sure others will chime in if my interpretation is incorrect.

 

[Gardnr]

cashflo2u2,

Let me rephrase what I think has been alluded to in a number of the above posts.

Typically with FireCalc, we look at 30-year time periods, the start of which rolls through the years, beginning in 1871 and ending in 1978 (the starting point of the most recent 30-year interval). Some of those periods start near market highs, some near market lows, and some in between. The FireCalc result we all hang our hats upon is the 4% WR which was safe 95% of the time. If one could run FireCalc (and I don't know how to do this without actually going in and looking at the individual 30-year period's data), I am pretty one would find that starting the "retirement clock" after the market had dropped 40%, would lead to a SWR in those instances which is considerably higher than the 4% number (or whatever number one used for a SWR - even a 100% safe one). This is why I think restarting your clock now (even with a 21-year time horizon) could be very misleading, because any failures FireCalc reports are probably the ones that started right around those same market highs.

If you retired in 2001, it looks like the market was down somewhere between 10% and 20% from its 2000 high. By choosing a SWR that was 100% safe back then, it seems to me you have most likely been sufficiently conservative.

This is my understanding of the way FireCalc works - I'm sure others will chime in if my interpretation is incorrect.

Well said, I'll second that.

 

[jIMOh]

I greatly appreciate all the responses. It is very heartened. A few more specifics and answers to some questions.

• I retired 7 years ago at age 61.
• 100% vs. 90% success rate. Even though the 90/10 sounds like good odds, the 10% represents catastrophic failure to me so I weight it much more heavily in this circumstance, to the point that I cannot tolerate it and go for 100%.
• At the time I retired, never heard of SWDR or and WDR for that matter. At some point I discovered the SWDR concept and firecalc. I have done a “look back” calculation to see how far off I drifted. When I retired, my shortfall was $15,000 a year (SS, pension, rental income less expenses). I think I would have taken about 3 1/4% of my portfolio as an annual distribution. If I had followed that rule, my cumulative distribution to date, adjusted for CPI inflation, should have been $124,000. My actual withdrawals have been $162,000. My portfolio balance, after rising substantially, is back down close to the balance it was at retirement (unfortunately I fooled around with some penny stocks when things were good). After moderate cut back in expenses, my withdrawal rate is about 3% of my current portfolio. This is less than what my proforma inflation adjusted rate of distribution would have been from original retirement date, but of course I cannot make up for the over distribution I did of $38,000. Of course my life span is now 7 years less and I am using 21 years to lift off, a 3% inflation rate, portfolio is 55/45. I am not sure if I understand the nuances of not running firecalc after one retires, but if I “pretend” I am retiring 1/1/09, using the output that says what the beginning portfolio balance should be to get 100% success rate, it shows a balance needed of about 15% less than my current balance (the narrative shows one figure and the little chart below it shows another).
• This has been mentioned, but the firecalc shows a potential worse case drawdown where the balance gets to a point that I don’t think anyone with blood in their veins could tolerate.
• For anyone’s benefit, when I retired I had a fantasy of saving money from not going out to lunch everyday, coffee out at break twice a day, commuting expenses, clothes, etc. Yes, I was going to make my lunch at home and coffee, no commuting, no more suits, hang around the house, etc. Well, that lasted about 6 months. I go out to lunch often with my cronies, go out for coffee for excitement and now my retired wife goes with me, I drive around town a lot more, took up golf, gourmet cooking, drink twice as much wine, got a second dog, etc. etc. Plus, when I retired I just estimated my expenses using good old Kentucky windage. I now use Quicken and keep track of every penny.

I suppose I could use more specific numbers, but still would appreciate any additional comments.

If you have a 3% withdraw rate, something I would consider is trying to use dividend payout as the primary method of withdrawing.

Many threads and posters here suggest they get a 3.5%+ yield on their portfolio from dividends- from stocks and REITs. If you can set up portfolio this way, I would think success rates would be infinite- because you would never or rarely have to sell a share to generate income.

This is my preferred withdraw strategy if I can accumulate enough money to pull it off.

 

[bamsphd]

This is the way I look at using firecalc. It isn't "god" and capabable of telling you 100% safe or not. There is a flaw in its modeling capabilities...

If you re-run and adjust upwards when the market is good, even if you always goto 100% safe, you are slowly putting yourself into the position to experience the worst case scenario. (You will run out of money)

If you re-run and adjust downwards when the market is bad until you get 100% safe, you are slowly putting yourself into the "best case" scenario (you'll die very rich).

Well said.

Also consider that past returns are no guarantee of future results. Firecalc just back-tests your scenario against historical results. There are no guarantees that your retirement might not experience worse results than ever seen before in US history. Remember that the US stock market has been the global winner in the time period that Firecalc uses for its calculations. I recommend for your reading enjoyment this article from September 1998: Stocks? For the Long Run?

 

[youbet]

Many threads and posters here suggest they get a 3.5%+ yield on their portfolio from dividends- from stocks and REITs. If you can set up portfolio this way, I would think success rates would be infinite- because you would never or rarely have to sell a share to generate income.

.

Success with this method assumes divs and int grow along with your personal inflation rate. They may. They may not. Best to be sure you understand that.

 

[jIMOh]

Success with this method assumes divs and int grow along with your personal inflation rate. They may. They may not. Best to be sure you understand that.

Correct- as prior posts in this thread have suggested- even in this economy the payout of their S&P 500 funds increased relative to prior years.

Based on the past performance of S&P500 the dividend payout of the index has been higher than the government published inflation rates/ Personal inflation rates might vary.

 

[freebeer]

Are you running Firecalc with default 30 year payout period setting? If so it is rather extreme to say that "barely at 100%" is "critical". Since you say you are over 65, you would have less than 15 years of average life expectancy, and the chances of you being alive 30 years hence are probably only 2%. Thus even a 95% Firecalc result for 30 year payout doesn't mean you have a 5% chance of going broke, but more like 1%, since the vast majority of the times where you (would have) run out of money (most often late in the 30 year period) you would have "run out of life" first. And after all, the improbable outcome where you are still going strong at 95 but are broke would probably be a lot more attractive than the far more probable alternative.

I think a lot of us (myself included) fixate on achieving extremely high "safe" withdrawal rates, disconnected from reasonably expectations of longevity. Perhaps this is an attempt to deny our mortality. But it really doesn't make much sense.

 

[tipster]

Try another program

I suggest you test the projected success rate of your portfolio using some other free retirement calculators, and see what results you get. For example, Jim Richmond's Flexible Retirement Planner or my TIP$TER retirement calculator.

I'm curious how the results would compare for you. Both programs let you specify the expected average return and standard deviation. In Flexible Retirement Planner, you can specify the expected arithmetic return (after selecting "Investing Style:Custom"). In the TIP$TER spreadsheet, by contrast, you specify the expected annualized (geometric) return.

So, what annualized real rate of return do you expect, going forward, on your portfolio? The returns over the past 140 years may have been more generous than what you can realistically expect over the next 20-40 years, because we're (1) starting with a lower dividend yield, (2) the US and world population is not growing as rapidly (not as much of a "demographic bonus" going forward, and (3) services -- rather than goods -- command an increasing portion of our GDP, suggesting that productivity growth going forward will be slower.

One more comment: if you want a 100% success rate, maybe you should put all of your savings, or at least enough to meet your minimum withdrawal needs, in TIPS. The median expected returns from TIPS may be lower than the median expected returns from equities, but the success probability may be considerably higher. It's a trade off -- but it sounds like you very much value financial security.

 

[youbet]

Correct- as prior posts in this thread have suggested- even in this economy the payout of their S&P 500 funds increased relative to prior years.

Based on the past performance of S&P500 the dividend payout of the index has been higher than the government published inflation rates/ Personal inflation rates might vary.

I think we're agreeing, but....... just to be sure.....

The div + int themselves must = your budget (or the part of your budget not covered by other income). The yr to yr increase in div + int must equal your personal inflation rate.

The div + int itself =ing inflation isn't what you're looking for. You need the yr to yr increase in div + int to = your personal inflation rate.

I think we're saying the same thing?

 

[youbet]

I think a lot of us (myself included) fixate on achieving extremely high "safe" withdrawal rates, disconnected from reasonably expectations of longevity. Perhaps this is an attempt to deny our mortality. But it really doesn't make much sense.

What I'm noticing is a little different. People want high success probabilities because they realize that "close calls" will drive them nuts during retirement. Even a plan destined to succeed (never run completely out of money) can dip and dive during the withdrawal period in ways guaranteed to test even the brave-hearted! Folks feel plans with high (100%) success probabilities, and probably some cushion on top of that, will be less likely to give them a scare when they're in their 70's and can't do anything about it.

Example: 2.4 yrs into retirement our net worth is down roughly 25%. Mother of mercy! DW has already extrapolated this to mean we'll be out of money before we end our first decade of retirement bliss. She wants me to cancel next springs fly-in fishing trip with my cronies!

What folks really need is less variability. Of course, plans with less variation generally have lower long term returns which requires more money, just like plans with high success probabilities...... Which leads me to the only truism I've discovered in retirement financing so far:

More money is better!

 

[REWahoo]

...the only truism I've discovered in retirement financing so far: More money is better!

This is how sig lines are born...

 

[jIMOh]

I think we're agreeing, but....... just to be sure.....

The div + int themselves must = your budget (or the part of your budget not covered by other income). The yr to yr increase in div + int must equal your personal inflation rate.

The div + int itself =ing inflation isn't what you're looking for. You need the yr to yr increase in div + int to = your personal inflation rate.

I think we're saying the same thing?

Correct

 

[FIRE'd@51]

Example: 2.4 yrs into retirement our net worth is down roughly 25%. Mother of mercy! DW has already extrapolated this to mean we'll be out of money before we end our first decade of retirement bliss. She wants me to cancel next springs fly-in fishing trip with my cronies!

Now that she's fixated on extrapolation, all you have to do is wait for the market to turn around and start going up. Then maybe she'll extrapolate that, and you will be able to go on two fly-fishing trips.

 

[Lusitan]

I would much rather settle for a 95% FIRECalc success rate than sell into this market. I'd be willing to go as low as an 80% success rate before letting myself get pushed into selling out now. Good luck.

 

[haha]

Success with this method assumes divs and int grow along with your personal inflation rate. They may. They may not. Best to be sure you understand that.

Overall I believe that this is one of the safest bets in the investment world. Choose wisely, and not for yield above all, and you are very likely to do well.

Also it helps to understand that for some companies the dividend is important to the shareholder base, and the company will mess with it only very reluctantly.

One more thing- there is a lot of work involved with really trying to stay on top of this stuff. At a miumum, keeping spreadsheets to track financial statement changes, listening to conference calls and presentations, watching for insider sales or buys, etc.

Still, with 20 year TIPS above 3%, this solves a lot of problems with much less need for ongoing work.

Ha

 

[Animorph]

It was clear to me that FC was going to give bogus answers using today's portfolio values (by adding any of the previous bear markets on top of the current bear market), so I have been avoiding running it.

However, it seems that maybe it is MOST valid for finding 100% safe SWR's only when the market is near a PEAK. The "near" part is only because the FC data is yearly and doesn't capture peaks and bottoms exactly. That is the only timing that makes sense when FC is going to take that entered portfolio value as a peak just before all the historic bear markets. Those are the cases that are going to fail.

I'm sure most other retirement simulations, including Monte Carlo, also are making no adjustment for the current bear market. Assuming equities are going to see 6% yearly gains (or any conservative market average) from a bear market bottom will not be any more reassuring than FC.

As for taking money out of the market, Buffet was certainly not saying less than 100% safety from FC meant you couldn't afford further losses and should take all your money out. We had a local TV interview with a guy who had all his college money in equities and now couldn't afford to go until the market recovered. He was the guy who should have gone to cash as soon as that became a problem, and shouldn't have been in equities this close to needing that money. That's what Buffet was talking about.

I'm still trying to find more cash and buy more equities at the lowest prices I can, without depleting my cash for living expenses for the next year or two. You should take the least amount out of equities as you can for now. Each dollar removed while the market is 40% down is worth $1.67 when the market recovers.

 

[tipster]

I'm sure most other retirement simulations, including Monte Carlo, also are making no adjustment for the current bear market. Assuming equities are going to see 6% yearly gains (or any conservative market average) from a bear market bottom will not be any more reassuring than FC.

I think that you are right. I think most MC simulators run "random walk" simulations (i.e., each randomly generated return is independent of all previous randomly generated returns), not "mean-reverting" simulations.

Over the long term, markets tend to revert.

And that's one of the benefits, I think, of Firecalc's use of historical returns -- it builds in that mean-reverting behavior.

But one can model mean reversion in a Monte Carlo simulator too. You can make the expected return, for any given simulation year, partially dependent on the degree to which the cumulative result of past simulated returns under- or over-shoots the original expected return. The way I do it is to treat the expected return on stocks for any given year like the expected return on a gov't bond that is going to mature in 30 years (or whatever length you choose). If you bought a 30-year gov't bond with a 5% coupon rate, and just a few weeks later it lost 25% of its value, the effective yield-to-maturity (YTM) on that bond -- at the depressed price -- would go up. Similarly, I can start a simulation with a 5% expected return, and if the random # sequence results in a 25% drop in the value, I calculate what annualized return it would take, over the next 30 years (or whatever length you chose) to get back to the originally expected return. And that's the expected value I use in generating the random return for the second year. And then I recalculate the expected return for the third year, and so on....

This type of simulation definitely improves the survival rates, as I illustrate in my tutorial. (And I do it without any significant change to the standard deviation of the randomly generated returns).

I've seen only one other free MC simulator on the net that tries to build in mean-reverting behavior. See Enhanced Monte Carlo Simulation. But he does it by filtering out about 97-99% of the simulations.

 

[freebeer]

youbet,

If "what folks really need is less variability" then they should buy inflation-adjusting annuities. My guess is that at 70 years old, as the OP may be, the payout on immediate annuity is going to be more than Firecalc's 100% success rate over default 30 year payout period, probably more than Firecalc's 95% success rate.

Why not do this? Because people really want many things, including the likelihood of passing on a legacy, and the ability to not work forever (this is after all the *Early Retirement* Forum) - so "more money is better" is not the sole answer either.

At the end of the day most of us probably prefer a plan that provides some comfort around all of these variables. But there's no free lunch or magic answer, it's all about balance. And at the end of the day this is where I come back to factoring in longevity. If the probability of your being broke is 10 times less than the probability of your being dead, you are probably focusing on the wrong thing if you are stressing that your Firecalc result is no longer 100%.

 

[youbet]

youbet,

If "what folks really need is less variability" then they should buy inflation-adjusting annuities.

Well, that would be your opinion. Mine would be that an inflation adjusted annuity would be an attempt to have zero (not just "less") variability. For myself, I'm working on trying to catch some of the current tempting fixed investment opportunities to improve my div + int cash flow to be near required spending needs. But I think I understand when an immediate annuity could be helpful.

Why not do this? Because people really want many things, including the likelihood of passing on a legacy, and the ability to not work forever (this is after all the *Early Retirement* Forum) - so "more money is better" is not the sole answer either.

I've run into very few retirement plans than were not improved, by my standard of measure , by having more money!

it's all about balance.

There ya go!

Your point regarding longevity is well taken freebeer. But some folks, sometimes myself, just can't develop a casual outlook about the possibility of being destitute for the last year or two or three of life and go to great lengths to protect against that. I assume living to 95 in my Firecalc runs, despite having no relatives that have made it that far.......

In my financial fantasies, DW and I live out a nice retirement with a pleasant standard of living, our portfolio remaining relatively constant in real terms, until we croak and our kids inherit a tidy sum enabling them to overcome some unexpected hardships (child with special needs). But nooooooo....... here comes Mr Recession early in our RE introducing dreaded downward variability and the question of future recovery, stability at this new, lower level, or even further drops. I tell ya freebeer, if it's not one thing, it's another.....