What do you consider an acceptable success %

testtubes

Dryer sheet wannabe
Joined
Jan 24, 2007
Messages
19
I'm sure everyone has their own opinions. After running your numbers through Firecalc, what do you think is an acceptable chance for success? Do you think that maybe some of the scenarios that they look at in the history of the market might have safeguards built in now that may protect us from some of the worst ones?
Is there a way to have Firecalc run your numbers against say the last 40 years of market history?
 
I would consider going no lower than 95% is you want to rule out events like the Great Depression.
 
Depends on your situation. Can you cut expenses or generate income in a pinch (possibly by going back to work)? Can you relocate to a lower cost area if things get bad? Then maybe 75 or 80% is OK.

If not, I don't think I'd want to go below 95%.
 
Personally for me, 95%.

I don't know what scares me more, running out of money when I'm really old, or dying with a lot of money when I'm young.
 
testtubes said:
I'm sure everyone has their own opinions. After running your numbers through Firecalc, what do you think is an acceptable chance for success?
William Bernstein's "Calculator From Hell" series claims that anything over 80% is analysis paralysis...

http://www.efficientfrontier.com/ef/998/hell.htm
http://www.efficientfrontier.com/ef/101/hell101.htm
http://www.efficientfrontier.com/ef/901/hell3.htm
http://www.efficientfrontier.com/ef/103/hell4.htm
http://www.efficientfrontier.com/ef/403/hell5.htm

testtubes said:
Do you think that maybe some of the scenarios that they look at in the history of the market might have safeguards built in now that may protect us from some of the worst ones?
Yes, we're always ready to win the last war again. What concerns many people is the scenarios that no one really saw coming (or didn't expect to be so bad) like nuclear terrorism. Or the Federal Reserve no longer being able to control interest rates in a global economy.

testtubes said:
Is there a way to have Firecalc run your numbers against say the last 40 years of market history?
FIRECalc now discards partial-period runs, so the only way to look at the most recent 40 years is to plan for a retirement of 40 years or less. Unfortunately 40 years isn't much data, and anyway what we all want to analyze is the next 40 years...

You could perhaps run Monte Carlo on the last 40 years, but again there's a concern about garbage in, garbage out.
 
testtubes said:
Do you think that maybe some of the scenarios that they look at in the history of the market might have safeguards built in now that may protect us from some of the worst ones?

I don't think we're going to have a repeat of history, but that doesn't mean that we can't *easily* have a worse outcome than what we've experienced return-wise.

Consider just some of the topics discussed here almost daily:

1) What if your personal rate of inflation is higher than the CPI?

2) What if risk premiums have dropped and forward equity returns will be lower?

3) What if we don't have a giant bull market to quickly offset every bear market? (Think Japan.)

So, it boils down to "are you feeling lucky?" If you don't want to gamble on retirement failure, then go with 100% historical safety, and then add an extra margin in case we haven't seen the worst case yet.

Remember, if you think your retirement will last 40 years, there have only been 3 40-year historical epics to test against. The way I see it, that means we have a 25% chance of the next epic being the worst case epic.
 
I hope you are not using just FIREcalc for your calculations or viewing FIREcalc probabilities as future probabilities. Most experts (Bernstein, Swedroe, Bogle, etc.) agree that the equity risk premium for all equity asset classes has come way down, meaning that future equity returns should not be as high in the future as they were in the past. This will affect the optimal asset mix, total portfolio return, SWR, etc. FIREcalc should be just one of the techniques you use to evaluate your chance for success. I see almost a disturbing reliance on FIREcalc calculations on early retirement boards.
Do you think that maybe some of the scenarios that they look at in the history of the market might have safeguards built in now that may protect us from some of the worst ones?
You get paid for systematic risk. If the likelihood of bad things happening has decreased, then the equity risk premium decreases. There is no free lunch.

Kramer
 
I think something in the 85-90% range is worth considering. So much will change around so many parameters and I imagine that people who can round up a million or two are fairly good at adapting and improvising.

As far as safeguards...yeah sure, some bad things that have happened have resulted in some changes that might make those things a tough repeat...but I dont think theres much proactive fixing of what MIGHT happen.
 
For me, I feel most comfortable at 100%. I'd rather work an extra year or two than take the chance of running out of income. Once I quit, I don't intend to start working again. I will probably do volunteer work, but I do not want to be dependent upon work for income. I am also overestimating the income I need so that in the event of a bear market or a period of protracted low return I can easily make adjustments.
 
I just don't see myself as a "balls of steel" type who is going to take my 4.65% every year no matter how bad the market tanks. I will adjust my spending, at least to a certain degree, if I have a really bad year. I think it's just human nature. Considering I don't want to live my retirement on oatmeal and water, there will be fat in the budget year to year. So I see ( just for me, mind you ) that a high 80% low 90% success rate as sufficient with FireCalc.
 
Cute Fuzzy Bunny said:
I think something in the 85-90% range is worth considering. So much will change around so many parameters and I imagine that people who can round up a million or two are fairly good at adapting and improvising.

CFB, you're speaking my languague!

I look around myself, I don't find anything that gives more than 90% success rate. Marriage, career, raising sucessful kids, winning a tennis or billiard set, going on a rewarding vacation, adventure,... and yet I do all those thing. Heck, I would take a plunge at any opportunity that gives me a 75% advantage. 3 out of 4 is never a bad deal for me.
 
I would advise to be careful about confusing a FIREcalc success percentage with an actual probabilistic statement about the probability of success of an SWR over the next Y years. If FIREcalc says X% success, the actual future probability distribution is probably lower. IMO, this even applies if you are diversifying more than FIREcalc states and willing to lower your SWR in bad times, since the equity risk premium has dropped so far and the asset mix was not chosen ex ante.

I would certainly jump into retirement with a 90% probability of success. But IMO a FIREcalc number of 90% is not even close to a real probability of 90% success over the next 30 or 40 years.

I think rates of change of SWR are more trustworthy, however, since that is more of a relative number. Both FIREcalc and Monte Carlo analyses have shown that one should subtract about 0.5% from the SWR when going from 30 to 40 years. I think that is a good rule of thumb. So I would trust 30 year FIREcalc runs more than 40 year runs, and then subtract 0.5% for the extra decade.

On the plus side, you should have some social security. This helps the whole concept of SWR quite a bit.

Kramer
 
Alex said:
For me, I feel most comfortable at 100%. I'd rather work an extra year or two than take the chance of running out of income.

I agree with what Kramer said.

All you've really guarranteed yourself is an extra year or two of work. That 100% doesn't mean 100% chance of success for you. It's 100% for someone in the past with precisely that asset allocation. You want 100%--stick with death and taxes ;)

I am also overestimating the income I need so that in the event of a bear market or a period of protracted low return I can easily make adjustments.

the willingness/ability to make adjustments, imo, is far more of a safety net than a 100% firecalc run.
 
Sam said:
CFB, you're speaking my languague!

I look around myself, I don't find anything that gives more than 90% success rate. Marriage, career, raising sucessful kids, winning a tennis or billiard set, going on a rewarding vacation, adventure,... and yet I do all those thing. Heck, I would take a plunge at any opportunity that gives me a 75% advantage. 3 out of 4 is never a bad deal for me.

You probably realize that this is a very misleading analogy. 3 out of 4 at something that gives an even money payoff and can be bet in small increments is a gift form God. You will wind up with all the money in the game. But this isn't the same situation. This is more like a footbridge over a high canyon. You have some data that 3 out of 4 people make it over; otherwise it is a long and tiring detour.

3 out of 4 still good enough?

Ha
 
kramer said:
I would advise to be careful about confusing a FIREcalc success percentage with an actual probabilistic statement about the probability of success of an SWR over the next Y years. If FIREcalc says X% success, the actual future probability distribution is probably lower. IMO, this even applies if you are diversifying more than FIREcalc states and willing to lower your SWR in bad times, since the equity risk premium has dropped so far and the asset mix was not chosen ex ante.

I would certainly jump into retirement with a 90% probability of success. But IMO a FIREcalc number of 90% is not even close to a real probability of 90% success over the next 30 or 40 years.

I think rates of change of SWR are more trustworthy, however, since that is more of a relative number. Both FIREcalc and Monte Carlo analyses have shown that one should subtract about 0.5% from the SWR when going from 30 to 40 years. I think that is a good rule of thumb. So I would trust 30 year FIREcalc runs more than 40 year runs, and then subtract 0.5% for the extra decade.

On the plus side, you should have some social security. This helps the whole concept of SWR quite a bit.

Kramer

Very helpful analysis kramer- thanks!

Ha
 
in typical usage, FireCalc essentially provides a worse case scenario ... in that sense it seems foolish to ingore the worst cases ... for that reason i typically use 95%, and occassionally check-out 99%.
 
kramer said:
Most experts (Bernstein, Swedroe, Bogle, etc.) agree that the equity risk premium for all equity asset classes has come way down, meaning that future equity returns should not be as high in the future as they were in the past.

If the equity risk premium is coming down, isn't the volatility reduced also ?

If so, maybe the impact of reduced equity volatility might help SWR sustainability more than lower equity premiums hurts ?
 
Delawaredave said:
If the equity risk premium is coming down, isn't the volatility reduced also ?

If so, maybe the impact of reduced equity volatility might help SWR sustainability more than lower equity premiums hurts ?
That's a good question, and I don't really have a good, precise answer. Maybe others can chime in.

It is almost certain, however, that an optimal, ex-ante mix of assets today, based on what we know about risk premiums, would contain more bonds than FIREcalc would show was successful for a given SWR in the past, probably significantly more. I quote William Bernstein on this below.

For instance, if you can get long term TIPs at 2.4% real (about the current rate), the SWR for 45 years is about 3.57% (40 years is 3.83%). This becomes more impressive when you consider that most folks will also have social security. Of course, there will be reinvestment risk and CPI risk and probably tax issues, but that is still impressive as an extreme example.

I think it is important to be flexible and think out of the box, too. I am actually semi-FIREing in the next few months, sooner than I would like, but I am totally burned out on my job and can't take it anymore. My initial SWR will be in the low 3's if I take on no employment. But there is a lot of uncertainty not just in future returns, but in regards to my future expenditures -- health insurance inflation, what if you get married (or divorced or reproduce), really sick, want to keep up with technological advancements and their cost, find a more expensive hobby, need to help out a close relative, etc. That is why I am calling it semi-FIRE and have plans for some interesting work in the early retirement years. I might dial this up or down depending on overall portfolio size (but not the previous year's portfolio performance, big difference).

Kramer

http://www.efficientfrontier.com/t4poi/Ch1.htm

n short, bondholders in the twentieth century were blindsided by what financial economists call a "thousand year flood": in this case the disappearance of constant-value gold-backed money. Before the twentieth century, nations had temporarily gone off the hard money standard, usually during wartime, but its permanent global abandonment was never contemplated until a decade before it finally occurred after World War I.

The shift in the investment landscape was cataclysmic, and the resulting financial damage done to bonds was of the sort previously seen only as the result of revolution and military disaster. Even in the United States, which suffered no challenge to its government or territory in the 1900s, bond losses were severe.

Consider that in 1925, the U.S. stockholder and bondholder both received a 5% yield. The bondholder could reasonably expect that this 5% yield was a real one—that is, that its fixed value would not decrease over time. The stockholder, on the other hand, balanced the prospect of modest dividend growth versus the much higher risk of stocks. The abandonment of hard money turned all that upside down—suddenly, the future value of the bondholder’s income stream was radically devalued by higher inflation, whereas that of the stockholder was enhanced by the ability of corporations to increase their earnings and dividends with inflation. It took investors more than a generation to realize this, in the process dramatically raising the prices of stocks and lowering that of bonds.


[Kramer NOTE: Bernstein goes on to explain why moving off gold standard was a good thing, despite the one-time cataclysm]


Although it is difficult to predict the future, it is unlikely that we shall soon see a repeat of the poor bond returns of the twentieth century. For starters, our survey of bond returns suggests that prior to the twentieth century they were generous.

Second, it is now possible to eliminate inflation risk with the purchase of inflation-adjusted bonds. The U.S. Treasury version, the 30-year "Treasury Inflation Protected Security," or TIPS, currently yields 3.45%. So no matter how badly inflation rages, the interest payments of these bonds will be 3.45% of the face amount in real purchasing power, and the principal will also be repaid in inflation-adjusted dollars. (These are the equivalent of the gold-backed bonds of the last century.)

Third, inflation is a painful, searing experience for the bondholder and is not soon forgotten. During the German hyperinflation of the 1920s, bonds lost 100% of their value within a few months. German investors said, "Never again," and for the past 80 years, German central banks have carefully controlled inflation by reining in their money supply. American investors, too, were traumatized by the Great Inflation of 1965-1985 and began demanding an "inflation premium" when purchasing long-term bonds. For example, currently, long-term corporate bonds yield over 7%, nearly 5% above the inflation rate.

Lastly, and I’ll admit this is weak reed, it is possible that the world’s central banks have finally learned how to tame the inflationary beast.

But the key point is this: bond returns in the last century should not be used to predict future bond returns.
 
Newbie here, but my take is once retirement commences, we need to be in constant re-evaluate mode. No different that I do today in my so called "accumulation phase". On a regular basis I adjust my spending based on budgets (1 yr or less) and projections of wealth growth (1-20 yrs). Why would this be any different in retirement? If I retire at 95% probability of success, I would not be suprised in the next years review if it dropped to 90%. I would make corrections in the firecal tool to understand what changed and what adjustments need to be made in real life to get back to goal level of 95%.

Like Kramer just said, this needs to be a flexible existance. Thinking one can lock in to a static existence is a little presumtuous. I would plan for monitoring and updating budgets/expecation regularly if 100% is really the confidence goal.

Dave
 
Scarab Dave said:
Newbie here, but my take is once retirement commences, we need to be in constant re-evaluate mode.

Bingo. We adapt our lifestyle to our income, and ideally leave enough headroom for a long rainy season.

I always enjoy all the "who needs bonds?" threads that appear after a 4-year-long bull market run. I look forward to the "uh oh -- now what?" era. Not so much for the schadenfreude, but for a nice new mix of topics. :)
 
Scarab Dave said:
Like Kramer just said, this needs to be a flexible existance. Thinking one can lock in to a static existence is a little presumtuous. I would plan for monitoring and updating budgets/expecation regularly if 100% is really the confidence goal.

Dave

The difficulty is that except for those who are way over funded, there isn't much one can do other than find some kind of employment. Depending on what you did, how long ago, the state of the economy (bad, I would expect if a retirement is messed up) this may not be very easy to do.

Ha
 
That would only be true HAHA if you planned to exist on cat food. Otherwise there are levers to be pulled to maintain the essentials. Cancel the golf outings, sell the mobile home in Brownsville, or for me, let the scarab sit a bit more than I had hoped. Regular adjustments are imperative to a successful retirement, in my opinion. To blindly follow a plan for 30 years would be unsuccessful unless you are very rich or quite lucky.

Dave
 
A number of posts above compare "expected 21st century" returns to the "actual 20th century" returns in FIRECALC. I think they have valuable insights, and I don't have much to add.

However, I'd like to add to this comment:

brewer12345 said:
Depends on your situation. Can you cut expenses or generate income in a pinch (possibly by going back to work)? Can you relocate to a lower cost area if things get bad? Then maybe 75 or 80% is OK.

Your required probability of success certainly depends on the split of "needs vs. wants" in your spending. If you think your spending is mostly on "needs" that you couldn't practically cut, then you have to look for a high probability. If your spending has plenty of "wants" that you'd be willing to do without, then a low probablity makes sense.

Here's an approach in FIRECALC: Determine your needs (as in "I need to have $35k per year just to pay for the basics), then run FIRECALC, solving for the assets required to support that spending at a 99% or so level. Intuitively, this much asset goes in one bucket, and you withdraw at the FIRECALC assumed rate from that bucket.

The rest of your assets can be spent on "wants". If you like, you can blow them all on the around-the-world cruise in the first year and know that your "needs" bucket will still sustain you for the rest of your life.

Or, you can just put them in a "wants" bucket, and spend from it in any pattern you like.

Or, you can put those assets in FIRECALC and solve for the annual income that gives you a modest probability of success - maybe 50%. Then you start your retirement by spending at that rate. Periodically you can re-run the "wants" bucket, maybe after the market makes a big up or down move, and adjust your "wants" spending depending on what you find.
 
Independent said:
Your required probability of success certainly depends on the split of "needs vs. wants" in your spending. If you think your spending is mostly on "needs" that you couldn't practically cut, then you have to look for a high probability. If your spending has plenty of "wants" that you'd be willing to do without, then a low probablity makes sense.

Here's an approach in FIRECALC: Determine your needs (as in "I need to have $35k per year just to pay for the basics), then run FIRECALC, solving for the assets required to support that spending at a 99% or so level. Intuitively, this much asset goes in one bucket, and you withdraw at the FIRECALC assumed rate from that bucket.

The rest of your assets can be spent on "wants". If you like, you can blow them all on the around-the-world cruise in the first year and know that your "needs" bucket will still sustain you for the rest of your life.

Here is a post from Dory36 going back 3 years that carries the same basic concept. It's helped me come to terms with the "do I have enough" question and stop worrying about moving in with the kids:

http://early-retirement.org/forums/index.php?topic=496.msg6096#msg6096
 
wab said:
Bingo. We adapt our lifestyle to our income, and ideally leave enough headroom for a long rainy season.

I always enjoy all the "who needs bonds?" threads that appear after a 4-year-long bull market run. I look forward to the "uh oh -- now what?" era. Not so much for the schadenfreude, but for a nice new mix of topics. :)

Hike over here son, grab a knee, and I'll tell you about the 60's and 70's. ;)
 
Back
Top Bottom