Betterment RetireGuide SWR

[AdrianC]

I was messing around in with Betterment's RetireGuide and got an interesting result. Now, it could be that it didn't like the inputs, but here's what I got:

Savings: 0 (it won't let me put anything in without linking accounts, which I will not do).
Retire: In six months
Social Security: 0
56% Stocks/44% Bonds at retirement
Annual Spending: $100K
Save/year: $3.8M

That's a retirement SWR of 2.6%

Yikes!

[matchewed]

Yes 100000/3800000=2.6%

I'm not seeing your point.

[AdrianC]

Yes 100000/3800000=2.6%

I'm not seeing your point.

2.6 <> 4

Betterment is saying I need $1.3 million more than the MMM generally accepted wisdom.

[matchewed]

Yes 100000/3800000=2.6%

I'm not seeing your point.

2.6 <> 4

Betterment is saying I need $1.3 million more than the MMM generally accepted wisdom.

Sure. You can see what their methodology for their RetireGuide is right here.

Their methodology is different than the MMM method. Primarily they are looking for a 99% success rate (defined as a balance >0) from your given age to 90. You may or may not be looking for a 100% success rate. Further information on their advice methodology here. That is separate from their RetireGuide methodology.

So yeah different calculation and assumptions means a different SWR.

I'm not seeing your point.

[matchewed]

Interestingly enough given that Betterment earns money off of a percentage of your portfolio it is in their best interests to be conservative about these values as that means your portfolio will (potentially) be larger with them. Which means more money for them.

[AdrianC]

I just found it interesting. Thought others might too.

Vanguard's retirement income calculator appears to use an SWR of 4%.

[abenke]

Hi there, I created both RetireGuide and Betterment's SWR advice.  I realize that many at MMM like to use 4%, but it's generally accepted that this rule of thumb can fail with a regularity that's too high for most people.  You'll notice that the Betterment SWR is heavily influenced by time horizon.  If you have a shorter horizon, you can take more than 2.6%.  But it's also critical to note that we assume *variable* withdrawals are taken - that you update how much you take each year as your portfolio changes in value.  In our research, this resulted in an *overall total withdrawal* through the retirement period of more than the 4% rule would suggest, but with a range from about 2% to about 6%, yielding a success rate of > 99%.

Alex

[AdrianC]

Thanks Alex.

A question: Under "How we calculated this" it says:

To calculate how much you should save each year, we first calculate how much you need in total at your selected retirement age in order to provide the level of spending you desire (adjusted for inflation). This includes a Social Security estimate and any other income specified by you in the assumptions. From this, we determine how much you'll need to save each month from now until retirement by assuming a 60% likelihood of the portfolio reaching its goal target.

Can you explain this last sentence? Is the calculator taking a 40% haircut off of the total savings amount?

[AdrianC]

An update: I left a comment for Betterment on their site and received a reply same day - very impressive - there is a place to enter external account balances under "Or, manually sync an account".

I was able to put those balances in, but the calculation comes out the same: 2.6% SWR for a 40 year retirement.

I still don't know what they mean by "assuming a 60% likelihood of the portfolio reaching its goal target". I've asked.

I just finished "A Random Walk Down Wall Street" and I see that Burton Malkiel also recommends a 4% withdrawl rate.

[thd7t]

That's a very high bond ratio.  It probably does a lot to account for the conservative WR.

ETA: Betterment probably also has to factor their fees in, which are higher than those advocated by MMM (low cost index funds).  This would also decrease the SWR.

[AdrianC]

I’ve done some more reading and some emailing back and forth with Betterment (very quick response - very refreshing*), and yes, what I considered to be a very low initial withdrawal rate (and therefore very high pre-retirement savings amount) is really Betterment’s advice. It’s not a mistake in my data inputs or the way I’m using the tool.

The Betterment RetireGuide is advising me to save enough so that the first year withdrawal rate would be 2.1% for a “Conservative chance” and 2.6% for a “Likely chance”.

Vanguard’s Monte Carlo tool gives these chances of portfolio survival (50/50 stocks/bonds, 40 year retirement):
100% chance at 2.1%
99% chance at 2.7%
95% chance at 3.3%
86% chance at 4.0%

https://retirementplans.vanguard.com/VGApp/pe/pubeducation/calculators/RetirementNestEggCalc.jsf

So, Betterment is not out of line with Vanguard. If I want a very likely chance of meeting my retirement goal the 4% rule isn’t going to do it.

Here’s Jon Stein, CEO of Betterment:
The 4% retirement rule is broken
http://www.cnbc.com/2014/11/03/

Here’s the official Betterment explanation:
Why the 4% Rule Is Broken
In today’s economy, you need a better, more dynamic approach to planning your retirement income.
https://www.betterment.com/resources/retirement/investment-income-retirement/why-the-4-percent-rule-is-broken/

(* Been dealing with Aetna and DirecTV recently...talk about bureaucratic nightmares...)

[Retire-Canada]

Advising people to save twice as much as 4% SWR would dictate as the solution to the issues identified with the 4% SWR is a poorly optimized solution except for two things:

1. it will take people so much longer to save/invest the required $$ [nearly $2M vs. $1M for $40K] that the chances of failure will decrease simply because people will be working away for so many more years their retirements will end up considerably shorter.

2. if you make money from a % based fee on the investor's accounts a much bigger account held for much longer makes you more money

For someone 25yrs old with a 40% savings rate trying to hit 4% WR that means working 25 - 46yrs old. To save/invest twice that much they'd work 25 - 61 yrs old. Based on MMM's chart: http://www.mrmoneymustache.com/2012/01/13/the-shockingly-simple-math-behind-early-retirement/

Sadly fear based analyses of FIRE is easy to sell and totally overlooks the opportunity costs of working so much longer in the prime of your life to assuage that fear.

The beauty of these fear based predictions aside from being an easy sell is that by the time you appreciate their true cost and wish you had done something else there will be no practical recourse against the folks that sent you down that path.

[jorjor]

So, Betterment is not out of line with Vanguard. If I want a very likely chance of meeting my retirement goal the 4% rule isn’t going to do it.

Depends on what you consider "very likely," or what you have for fallbacks if you get a few bad years of returns, especially early on in retirement.

The whole premise behind the Betterment study seems to be that bond yields are lower than when the 4% rule was developed, so you need a higher starting balance. Yet they use a 30% stock allocation (I assume the remaining 70% to bonds) to prove their point. That's way lower than I think most around here would suggest. Later in the Betterment article it says:

We found the expected failure rate of the 4% rule under current conditions to be 31% when applied to a Betterment portfolio with a 30% stock allocation. Though analysis marshaled in support of the 4% heuristic in the 1990’s tended to recommend a 50% stock allocation ... Notably, we found that applying the 4% rule even to a 50% stock Betterment portfolio is expected to fail 21% of the time under modern market conditions.

If the biggest fear is financial ruin due to running out of money in retirement, why use a 30%/70% stock/bond portfolio to prove your point when it fails 10 percentage points more often than a 50/50 portfolio? I'm not necessarily advocating any of these allocations, but what does their MC simulation look like at 70% stocks? 90%? 100%?

[AdrianC]

If the biggest fear is financial ruin due to running out of money in retirement, why use a 30%/70% stock/bond portfolio to prove your point when it fails 10 percentage points more often than a 50/50 portfolio? I'm not necessarily advocating any of these allocations, but what does their MC simulation look like at 70% stocks? 90%? 100%?

We can't recreate the Betterment MC simulation, but we can do the Vanguard one:
2.7% SWR 40 year retirement
Stocks/Bonds Probability that the portfolio survives 40 years
30/70 99%
50/50 99%
80/20 97%
100/0 95%

4.0% SWR 40 year retirement
Stocks/Bonds Probability that the portfolio survives 40 years
30/70 81%
50/50 86%
80/20 85%
100/0 83%

Vanguard's tool isn't much different to Betterment's advice.

If a person takes the 4% route they have to be sure the budget can allow for spending cut backs, just in case.

For me the 'stache is more-or-less complete. I'm trying to decide how ridiculously extravagant and philanthropic we can be for the next 40-50 years.

[Retire-Canada]

Just for a sanity check I ran the same ratios & period through cFIREsim [old site] using historical data:

2.7% SWR Stocks/Bonds - Probability portfolio survives 40yrs

30/70 - 100%
50/50 - 100%
80/20 - 100%
100/0 - 100%

4% SWR Stocks/Bonds - Probability portfolio survives 40yrs

30/70 - 41%
50/50 - 68%
80/20 - 86%
100/0 - 89%

[AdrianC]

That's interesting. Thanks.

If cFIREsim is a better predictor than the Vanguard tool (I have no opinion on that) it looks like those hopeful of a 4% SWR should be very aggressive in their allocation to stocks...

[matchewed]

That's interesting. Thanks.

If cFIREsim is a better predictor than the Vanguard tool (I have no opinion on that) it looks like those hopeful of a 4% SWR should be very aggressive in their allocation to stocks...

Well yeah... historically speaking stocks return more than bonds. So in order to sustain a higher SWR you need to have a higher percentage in stocks.

[AdrianC]

Well yeah... historically speaking stocks return more than bonds. So in order to sustain a higher SWR you need to have a higher percentage in stocks.

Timing is important. Retire at the end of 2000 with 100% in S&P500 at 4% SWR...your investments made 2.7% annual real return for 14 years and you could be in trouble.

[matchewed]

Well yeah... historically speaking stocks return more than bonds. So in order to sustain a higher SWR you need to have a higher percentage in stocks.

Timing is important. Retire at the end of 2000 with 100% in S&P500 at 4% SWR...your investments made 2.7% annual real return for 14 years and you could be in trouble.

Sure but you can't predict the future, so you need to remain flexible. If the market tanks right when you pull the FIRE trigger you're at risk. But you can always go back to work, you're still young. You can also tough it out and mitigate that risk. Nords has had great posts on how they weathered the 2008 crisis.

But this is all going to rehash an old old old discussion on these boards (maybe one less old, this board isn't that old but the topic is).

Read through this - http://forum.mrmoneymustache.com/investor-alley/stop-worrying-about-the-4-rule/

Post any concerns or questions in it. That seems to be the crux of your worries.

[AdrianC]

Well yeah... historically speaking stocks return more than bonds. So in order to sustain a higher SWR you need to have a higher percentage in stocks.

Timing is important. Retire at the end of 2000 with 100% in S&P500 at 4% SWR...your investments made 2.7% annual real return for 14 years and you could be in trouble.

Even worse, retire at the end of 1999 with 100% in S&P500 at 4% SWR...your investments made 1.95% annual real return for 15 years. So I wonder about the cFiresim numbers for a 2.7% SWR...

[matchewed]

Well yeah... historically speaking stocks return more than bonds. So in order to sustain a higher SWR you need to have a higher percentage in stocks.

Timing is important. Retire at the end of 2000 with 100% in S&P500 at 4% SWR...your investments made 2.7% annual real return for 14 years and you could be in trouble.

Even worse, retire at the end of 1999 with 100% in S&P500 at 4% SWR...your investments made 1.95% annual real return for 15 years. So I wonder about the cFiresim numbers for a 2.7% SWR...

Because it is the aggregate total of looking at all of its history not just one particular nightmare scenario year.

Now you seem to be fear mongering yourself into a 0% SWR. GL w/ that.

How many 1999's have you lived through? What are the odds that the year you FIRE will be a 1999? It's not likely. So that's why a higher SWR is sustainable. There are more positive years than massive huge negative ones.

Again read that link I provided.

[Retire-Canada]

Keep in mind where the advice is coming from. Do they make more money by you following the advice? Do they suffer any negative consequence if you work years longer than you need to and get to FIRE when you are much older?

[jorjor]

Well yeah... historically speaking stocks return more than bonds. So in order to sustain a higher SWR you need to have a higher percentage in stocks.

Timing is important. Retire at the end of 2000 with 100% in S&P500 at 4% SWR...your investments made 2.7% annual real return for 14 years and you could be in trouble.

Even worse, retire at the end of 1999 with 100% in S&P500 at 4% SWR...your investments made 1.95% annual real return for 15 years. So I wonder about the cFiresim numbers for a 2.7% SWR...

Okay fine then. Follow this to it's logical conclusion: Never retire. We could start having 0% or negative return. Is anything safe?

If you want to plan for the worst possible cases and then some, then you're fine to do that. If a 1999 happened, one could see said 1999 happening and return to work. You're likely able-bodied and able to work, after all.

So it all comes back to risk tolerance and your definition of "very likely." I would consider an 85-90% success rate for a 4.0% SWR with the added security of being an able-bodied person and  flexible withdrawals (rather than sticking exactly to 4% WR) to be "very likely"

[Retire-Canada]

Okay fine then. Follow this to it's logical conclusion: Never retire. We could start having 0% or negative return. Is anything safe?

That's why the doom and gloom pundits get traction. A person can only generate a limited amount of money to save/invest, but the amount of fear that can be generated is pretty much unlimited.

[AdrianC]

No worries!

I was just trying to understand why Betterment, a company recommended by this very website, has such different advice on how much to save for retirement. Then we found that Vanguard has similar advice. I’ve got a good handle on it now. Fiduciary responsibility and all that.

Thanks for everyone’s input, and good luck to all.

[matchewed]

No worries!

I was just trying to understand why Betterment, a company recommended by this very website, has such different advice on how much to save for retirement. Then we found that Vanguard has similar advice. I’ve got a good handle on it now. Fiduciary responsibility and all that.

Thanks for everyone’s input, and good luck to all.

Well MMM recommends it. Not the forum necessarily. This is after all just a collection of individuals yammering at each other over the intertubes. There are some core commonalities, Betterment is not one of them.

I still recommend that link as it has great knowledge on why the 4% SWR makes sense depending on your level of risk tolerance. Please read it and try to grok it.