Fidelity RIP

[bUU]

IIRC, there's a link to a 20 (?) page PDF somewhere which outlined the methodology.

 

[Gone4Good]

IIRC, there's a link to a 20 (?) page PDF somewhere which outlined the methodology.

I saw the methodology. The "Returns" section pretty much says "we use a monte carlo simulation" and then spends a lot of words describing what monte carlo simulations are.

It's hard to get actual numbers out of the model.

So I set up a 100% equity portfolio and tried to run a retirement scenario where income matched expenses (Fidelity added some misc expenses anyway.)

Here's what nominal returns for Domestic Equity look like for the "down" scenario:

Year 1	-16%
Year 2	-1%
Year 3	-3%
Year 4	0%
Year 5	5%
Year 6	1%
Year 7	0%
Year 8	11%
Year 9	6%
Year 10	0%
Year 11	8%
Year 12	4%
Year 13	15%
Year 14	3%
Year 15	7%
Year 16	1%

That doesn't look like a terribly stressing stress test to me. Other people may come to other conclusions. But this is the kind of thing I need to know to make sense of the results from any retirement calculator.

Edit to add: Inflation looks like it is fixed at 2.5% annually - also not very stressing.

 

[braumeister]

Edit to add: Inflation looks like it is fixed at 2.5% annually - also not very stressing.

FWIW, Fidelity inflates health care expenses separately by 7%, the last time I looked.

 

[Gone4Good]

These are bond returns. Looks like they almost always exceed general inflation of 2.5%.

Year 1	-0.6%
Year 2	2.9%
Year 3	2.2%
Year 4	3.3%
Year 5	5.6%
Year 6	2.8%
Year 7	4.9%
Year 8	2.5%
Year 9	3.5%
Year 10	4.8%
Year 11	3.4%
Year 12	3.5%
Year 13	5.1%
Year 14	2.6%
Year 15	3.7%
Year 16	5.6%

 

[mathjak107]

I saw the methodology. The "Returns" section pretty much says "we use a monte carlo simulation" and then spends a lot of words describing what monte carlo simulations are.

It's hard to get actual numbers out of the model.

So I set up a 100% equity portfolio and tried to run a retirement scenario where income matched expenses (Fidelity added some misc expenses anyway.)

Here's what nominal returns for Domestic Equity look like for the "down" scenario:

Year 1	-16%
Year 2	-1%
Year 3	-3%
Year 4	0%
Year 5	5%
Year 6	1%
Year 7	0%
Year 8	11%
Year 9	6%
Year 10	0%
Year 11	8%
Year 12	4%
Year 13	15%
Year 14	3%
Year 15	7%
Year 16	1%

That doesn't look like a terribly stressing stress test to me. Other people may come to other conclusions. But this is the kind of thing I need to know to make sense of the results from any retirement calculator.

Edit to add: Inflation looks like it is fixed at 2.5% annually - also not very stressing.

those early years which determine the shape of the entire retirement look particularly stressing to me with those sequences . the first 5 years shape the retirement and by the 15th year the entire outcome of a 30 year retirement has been decided .

inflation is set to adjust in the calculator and is based on a few different parameters which currently point to 2.50% as the outlook

 

[Gone4Good]

those early years which determine the shape of the entire retirement look particularly stressing to me with those sequences . the first 5 years shape the retirement and by the 15th year the entire outcome of a 30 year retirement has been decided .

That's pretty much my entire point in a nutshell. We need to see what the model is doing so we can form some kind of opinion about what the results actually mean. It'd be nice if Fidelity could make it plain, rather than forcing us to cajole the model into giving up it's secrets.

Regarding Fidelity's one and only "underperforming market" scenario, I don't find a 16% stock market sell-off in year one, followed by a couple flattish years that taken together don't even break into bear market territory all that impressive. Perhaps I'm a bit jaded having already invested through two fiftyish percent market declines in the not so distant past. I don't see why my planning wouldn't include the possibility of a 3rd.

Now having dug through the numbers, I think Fidelity's "underperforming market" scenario is a reasonable enough basecase. I don't consider it a downside case, though. For that I'd actually like to see a severe bear market scenario and probably an inflationary scenario too.

But my point isn't to convince you about what is an ideal market forecast for retirement planning. It's that we need to see how the black box works in order to know whether we agree or disagree with the results its giving us.

 

[mathjak107]

the size of the drop does not mean much if it is short or happens after an up cycle , a modest drop that goes on for a few years does a lot more harm as well as burning principal up year one .

those who retired in 2008 are no different this many years in then most other average groups . on the other hand the y2k retiree is on track to mimic the 1929 retiree .

 

[dixonge]

The new name for the Retirement Income Planner is "The Planning and Guidance Center"

I'm not surprised they changed the name - the original seems like a bit of an inside joke...

 

[Gone4Good]

the size of the drop does not mean much if it is short or happens after an up cycle , a modest drop that goes on for a few years does a lot more harm as well as burning principal up year one .

But who says a big drop has to be followed by a big recovery?

And who says a 100% bond portfolio can sustain a 4% withdrawal for 32 years in their underperforming market scenario? Yup, that's Fidelity. Firecalc, meanwhile, says that same scenario fails 59% of the time. And even Firecalc's data doesn't include many periods where startging real yields were as low as they are now.

So do I think Fidelity is being too optimistic? Yup.

Are you completely free to think it's a rigorous stress test? Yup.

But at least now you have some actual information on which to make that assessment.

 

[mathjak107]

while firecalc typically uses actual historical data fidelity attempts to find even worse outcomes that are possible then historical so in that regard i find fielity even more conservative then firecalc .

we can all drum up visions in our head of devastating scenario's where nothing works . gold bugs have been doing that for decades .

they all have visions of being last man standing and only gold survives .

but reality is these things just don't play out or have a very very low chance of playing out .

in reality we have had nothing , not even 2000 or 2008 playing out where it even matched 1965/1966 .

could we ? sure but so far through the great depression , high inflation , world wars and the stagnant 2000's we haven't .

regardless of what the numbers look like , taken as a whole they are already based on some pretty nasty outcomes .

i just retired in july and right off the bat rip eliminated 15% of my principal day 1 in its calculations so i am pretty confident in the numbers.

personally i use bob clyatts dynamic system based on each year so this stuff pertaining to a rigid 4% does not concern me as much
as i am dynamic based on what is going on around me .

 

[Gone4Good]

while firecalc typically uses actual historical data fidelity attempts to find even worse outcomes that are possible then historical so in that regard i find fielity even more conservative then firecalc.

Except it doesn't actually do that, even though theoretically it could . . . and probably should given that current asset valuations are higher than those used in Firecalc's history.

So, no. Modeling bottom decile returns doesn't impress me when both equity and bond valuations are well within their top decile ranges. And when taken together, they're completely outside the historic data set. Based on valuations, one might expect bottom decile returns as a base case rather than a worst case.

So yes. I'd like at least the option of some more stressful stress tests. If you're comfortable that a 15% year one decline is as bad a scenario as we're likely to face, than the model works brilliantly for you. I, however, would like more options.

 

[mathjak107]

doesn't do what ? the rip planner reduced my balance by almost 15% day 1 when worst case was selected . .

 

[Gone4Good]

doesn't do what ? .

It doesn't do this . . .

while firecalc typically uses actual historical data fidelity attempts to find even worse outcomes that are possible then historical

It doesn't even claim to find worse outcomes than those in the historical record. It claims to model bottom decile results as it's downside case.

For laughs I tried to see what implied PE-10 equity valuations would be under it's average market return scenario. Using a 30-yr period and the sixty year growth rate of S&P 500 earnings of 5.5%, it looks like Fidelity's basecase scenario implies PE-10 values growing to about 46x.

 

[Options]

while firecalc typically uses actual historical data fidelity attempts to find even worse outcomes that are possible then historical so in that regard i find fielity even more conservative then firecalc .

we can all drum up visions in our head of devastating scenario's where nothing works . gold bugs have been doing that for decades .

they all have visions of being last man standing and only gold survives .

but reality is these things just don't play out or have a very very low chance of playing out .

in reality we have had nothing , not even 2000 or 2008 playing out where it even matched 1965/1966 .

could we ? sure but so far through the great depression , high inflation , world wars and the stagnant 2000's we haven't .

regardless of what the numbers look like , taken as a whole they are already based on some pretty nasty outcomes .

i just retired in july and right off the bat rip eliminated 15% of my principal day 1 in its calculations so i am pretty confident in the numbers.

personally i use bob clyatts dynamic system based on each year so this stuff pertaining to a rigid 4% does not concern me as much
as i am dynamic based on what is going on around me .

This is the problem with any calculator or tool "for education purposes" (as Fidelity reminds us): the precision we're seeking just isn't there. There are too many variables, the biggest being you may well die before any of it matters. We humans seek security/certainty, but we will not find it in any calculator or spreadsheet (although don't tell that to many Bogleheads--they appear ready to defend their spreadsheets to the death in that forum). Calculators/tools are simply data points and nothing more.

I have come to the conclusion all we can do is have some sort of cushion, have a plan b and c, and remain flexible once we retire. I've noticed myself doing the same thing instinctively in retirement as I did before retirement: when I go over budget, I stop spending until it's corrected. It worked for the many decades before retirement and it continues to work now.

In reading Warren Buffet's May 2010 testimony to the FCIC, I was surprised he considered the possibility of financial armageddon in 08 had the government not intervened in the financial crisis. To me, this goes to show if everything goes to hell in a hand basket, all bets are off and it won't matter anyway.

 

[Gone4Good]

In reading Warren Buffet's May 2010 testimony to the FCIC, I was surprised he considered the possibility of financial armageddon at the time had the government not intervened in the 08 financial crisis.

It's actually quite hard to imagine where the bottom would have been had markets been left to freefall.

And it's not at all hard to imagine that the response to a similar financial panic would be much more restrained if it were to happen again today.

 

[Options]

It's actually quite hard to imagine where the bottom would have been had markets been left to freefall.

And it's not at all hard to imagine that the response to a similar financial panic would be much more restrained if it were to happen again today.

I'm not so sure about that. Warren Buffet has said elsewhere he believes (paraphrasing) the more severe the crisis, the stronger the response from the government would be. In his FCIC testimony he also discussed how derivatives haven't considered the possibility of a terrorist attack. Again, were some sort of attack (i.e., nuclear device detonation) to occur near one's hometown, all bets are off.

My point is the quest for the holy grail in financial retirement security fails to consider other possibilities making it all a moot point anyway. William Bernstein has discussed this. No one is going to die if their PF blows up. We do have a safety net in this country, such as it is. Will your quality of life be downgraded? Yes. Substantially? Possibly. OTOH, I just read a couple of articles yesterday describing places where it is still possible to retire on the average social security check alone.

I believe it is possible to over think this stuff, and it is in the best interest of most in the "financial services" (to include financial information) industry to have us do so. We're a highly lucrative marketing target and it's best to remember this when playing the cat and mouse game of chasing financial security.

 

[MasterBlaster]

we can all drum up visions in our head of devastating scenario's where nothing works . gold bugs have been doing that for decades .

how many times have you read on this forum something like - "FireCalc tells me I can take a 4% SWR but just to be safe I'll take 3%... or 2%... or even less" ?

At least the Fidelity Planner is based on actual measured data as opposed to pulling SWR rates out of thin air.

 

[Gone4Good]

Warren Buffet has said elsewhere he believes (paraphrasing) the more severe the crisis, the stronger the response from the government would be.

A lot has happened since 2008. One of the things that has happened is that attitudes about things like aggressive monetary policy, fiscal stimulus, bailouts, government intervention of any kind, etc. have all hardened in a way that would make a government response to a similar crisis much harder now than it was then. Even if the government had the same resources at it's disposal now, which it doesn't.

And it wasn't all that easy in 2008/09. The first attempt to pass TARP legislation failed.

 

[bUU]

how many times have you read on this forum something like - "FireCalc tells me I can take a 4% SWR but just to be safe I'll take 3%... or 2%... or even less" ?

Folks who are struggling to find ways to get anticipated SWR from 6% down to 4% don't have that luxury.

 

[Gone4Good]

how many times have you read on this forum something like - "FireCalc tells me I can take a 4% SWR but just to be safe I'll take 3%... or 2%... or even less" ?

4% is based on a 30 year retirement with something like a 95% success rate. Most of us are planning an "Early Retirement." In my case, that started at the age of 38. I wonder how that 5% failure probability expands as we move from a 30 year time period to a 60 year one.

I know. So silly to think about such things.

 

[MasterBlaster]

4% is based on a 30 year retirement with something like a 95% success rate. Most of us are planning an "Early Retirement." In my case, that started at the age of 38. I wonder how that 5% failure probability expands as we move from a 30 year time period to a 60 year one.

I know. So silly to think about such things.

Evidently my point is beyond you. FireCalc is already a very conservative model. And then many take that and make it much more so. My point was in reference to those that thought the Fidelity model was too conservative

FireCalc can and does model what you posted - ie long retirements.

 

[Gone4Good]

Evidently my point is beyond you. FireCalc is already a very conservative model. And then many take that and make it much more so. My point was in reference to those that thought the Fidelity model was too conservative

FireCalc can and does model what you posted - ie long retirements.

Deciding whether any model is conservative or not requires that you know how it actually works. And that was my initial point, which somehow seems to be "beyond" many folks here who would rather defend Fidelity's calculator as sufficiently conservative.

 

[Gone4Good]

Have you accounted for the probability and effects on your SWR of war and nuclear war/socialism and expropriation/Supernova/alien invasion/Ebola and Tika/Global Warming/population explosion ?

Perhaps you should spend nothing - just in case. Or maybe go back to work - Just in case. Or buy gold - lots of it - just in case.

Why do you seem so defensive about me posting the implied equity and bond returns embedded in Fidelity's calculator here?

Am I really not allowed to think that those scenarios aren't as rigorous as I'd like?

 

[Options]

Why do you seem so defensive about me posting the implied equity and bond returns embedded in Fidelity's calculator here?

Am I really not allowed to think that those scenarios aren't as rigorous as I'd like?

I'm curious: how do you think knowing Fidelity's assumptions will help you? So you could look under the hood, tinker with it, and then come up with an idea whether you "agree" with or "disagree" with them? This is my issue with all of these types of debates (if you want a real debate, head over to the BH forum where they have beaten the merits of various PF withdrawal methodologies to the grave and back--the VPW believers border on fanatic, IMO). It's this search for certainty that doesn't exist.

Even Fidelity states their calculator is an educational tool (aren't they all?). As I stated previously, any calculator is simply a data point. If you want to negate any calculator's results, attempt to account for the impact of the unknowable: SS/Medicare reform, tax reform, age of death, and low probability/high impacts. Bernstein has addressed this, Scott Burns has addressed it, and lately Dirk Cotton in his posts on bankruptcy.

 

[Options]

A lot has happened since 2008. One of the things that has happened is that attitudes about things like aggressive monetary policy, fiscal stimulus, bailouts, government intervention of any kind, etc. have all hardened in a way that would make a government response to a similar crisis much harder now than it was then. Even if the government had the same resources at it's disposal now, which it doesn't.

And it wasn't all that easy in 2008/09. The first attempt to pass TARP legislation failed.

True, but after that first attempt they did a lot of things. In fact, they did some things that had not been done before. As to future responses, I'm with Buffet on this.