# Safe Withdrawal Rates – ERN #5

Today’s post is our fifth visit to ERN’s mammoth series of articles on Safe Withdrawal Rates.

###### Low returns

We’re quickly catching up with ERN – the posts we will be looking at today are less than a year old.

- Article 16 is about retiring into a low growth environment.

It begins with a quote from Jack Bogle:

The dividend yield is 2 percent … the earnings growth has been a little over 5, that’s going to be a very tough target in the future so let’s call it 4. Four and 2 give you 6 percent, but then you have to take the valuations in the market.

Maybe knock off a couple of points for a slightly lower valuation over a decade and you’re talking about a 4 percent nominal return on stocks.

This is 4% pa nominal, so maybe 2% real – and it includes dividends.

- This implies that (US) stock prices will only rise with inflation (after paying out dividends), for 10 years (from 2017).

Historical US stock returns are 6.7% (since 1871), so Bogle’s prediction is 4.7% pa below that.

- With the CAPE at 30 at the time of writing (it’s higher now), ERN’s own prediction is 3.3% real (1/CAPE).
- That would give us around 3.0% real at the moment.

ERN also predicts that after 10 years, with the CAPE back in the low 20s (on Bogle’s projection), returns could go back to normal.

ERN also looks at sequence of returns risk (SoRR), using three scenarios:

- Boring Bogle
- 4% nominal each year for equities
- bond yields climb from 2.2% to 3.2% at 0.1% pa

- Bad Bogle
- equities: -30% in year 1, flat in year 2, then 9.8% pa for the remaining eight years
- bonds have a drop in yields (to match interest rates) followed by recovery to 3.2%

- Good Bogle
- the reverse of Bad Bogle for equities
- bond yields rise to 4.3% by 2025 then fall back to 3.2%

Note that since ERN wrote the article, 10-Year Treasuries have already breached 3.0% pa.

Note also that the Boring Bogle scenario produces poor bond returns due to the steady rise in interest rates.

- The other two scenarios produce strong returns from bonds just when you need them (when stocks are down).

For once, ERN’s simulation is annual rather than monthly.

He uses three withdrawal rates (3%, 3.5% and 4%) and seven allocation strategies:

- 50/50 stocks / bonds up to 100/0 in 10% intervals
- Equity glide paths starting at 60% and 80%, increasing to 100% in 5% annual steps.

ERN makes several points about the results:

- They are all bad – the portfolio value is down.
- Bonds do worse than stocks.
- A glidepath is the best hedge against the Bad Bogle scenario, but 100% stocks works best in Good Bogle.
- ERN prefers the 80% Equity glidepath overall.

- ERN thinks that the 3.0% SWR is too conservative – he would go for 3.5%.
- Across all the scenarios, I’m not so sure – I might go to my usual 3.25% and adjust as necessary.

- We can both agree that the 4% SWR is too aggressive.

###### Pensions

Article 17 has more on pensions (including the State Pension), and explains the 4% Rule is only a Rule of Thumb (a heuristic)

You need to adjust the 4% rule for two main factors:

- equity and bond valuations (which is why it’s more of a 3.25% / 3.5% pa rule at the moment)
- State Pensions and other cash flows.

So in this post, ERN looks at how to measure the impact of future cash flows on the SWR.

- Note that I usually do this by capitalising the future cash flows into a current “total portfolio” number.
- This approach encourages me to be more aggressive with the rest of my portfolio (hold more equities) as I can see the safe money that I have in pensions.

ERN provides a section on the underlying maths, but I’ll skip that.

- If you are interested, ERN’s post is here.

For the rest of us, here’s a diagram that’s a lot simpler:

So the task is to work out the height of the yellow rectangle from the height of the green one.

ERN provides a conversion table:

- It covers retirements of 40, 50 and 60 years.
- It uses pension starts dates from 0 to 40 years, in 5-year increments.
- It allows for 60%, 80% and 100% stock allocations.

I’m sure that this table will be tremendously useful to some, but I’ll probably stick with my existing approach:

- My retirement horizon is now down to around 30 years.
- I will struggle to reach a 60% stock allocation.
- My other half and I have five DB and state pensions to account for, across four starting dates.

ERN runs through some case studies if you want to see the conversion table in action.

He then repeats the exercise for future benefits that are not inflation linked.

Here’s the diagram for that scenario:

And here’s the conversion table:

###### Flexibility and CAPE rules

Article 18 in ERN’s series is about the flexibility in CAPE-based withdrawal rules.

- He begins by recapping the trade-off between protecting the final portfolio value and having to take lower withdrawals along the way.

Dynamic withdrawals don’t spare you from the pain of low returns, they just spread it out and lower your risk of running out of money entirely.

Using the Constant Percentage rule (which is dynamic, but crude), he looks at four unlucky cohorts:

ERN’s preferred dynamic withdrawal rule is a CAPE rule:

When the stock market falls, so does the CAPE (since the price falls more quickly than the 10-year earnings measure).

- This means that the CAEY (= 1/CAPE) will go up, cushioning the withdrawal rate.

The rule can be extended to include cash and bonds:

ERN looks at 8 dynamic withdrawal rates, most of them CAPE:

- CAPE 1/0.5: a=1% and b=0.5.
- This is the traditional CAPE-based rule used as the default at cFIREsim.
- With the CAPE at 30 (when ERN was writing) the SWR is 2.7%

- CAPE 1.5/0.5
- This adds 0.5% to the base, taking the SWR for CAPE=30 up to 3.2%.
- This was ERN’s preferred rule when we looked at CAPE withdrawals the first time around.

- CAPE 1.75/0.5, to give an SWR of 3.45% at the time of ERN’s writing.
- CAPE 2.08/0.4
- This drops the multiplier, but increases the base to produce the same SWR as rule 3.

- CAPE 1.42/0.6:
- This increases the multiplier but lowers the base, to give the same SWR as rules 3 and 4..

- “CAPE robust”
- ERN used Excel solver to maximize his August 2017 withdrawal rate whilst limiting historic drawdowns (in withdrawal rate) to 30%.
- I’ll ignore the details for now, but come back to them if this is the winning rule.

- “Best of 3” is a weighted average of the rules 4, 6 and 8 designed to produce the same August 2017 SWR as in rules 3, 4 and 5.
- The constant percentage rule at 4%, for comparison.

All of these are tested with monthly withdrawals from an 80/20 bond stocks portfolio.

Here are ERN’s comments:

- The constant 4% rule has the worst volatility and drawdown stats.
- The withdrawal drawdowns are routinely 50% or more.

- The CAPE-based rules have a withdrawal volatility significantly smaller than the portfolio volatility.
- Some of the dynamic rules now imply a withdrawal rate of 3.41% even with a CAPE at 30!

- The CAPE-rules handled the Great Depression extremely well.
- The CAPE was at 30+ in 1929 and then dropped to 5(!) in 1932.
- You would have withdrawn only 3% at the peak and over 10% p.a. at the bottom, so even after a precipitous drop, the withdrawal amount was not reduced that much.

- Between 1970 and 1982 we had four recessions, two of them major. Due to the inflation shock and rising bond yields, bonds got hammered and negated any diversifying benefit.
- Under the constant percentage rule, a $40K initial withdrawal would have been decimated to $16,000 in the early 80s.
- Withdrawals would have been below $25,000 for 11 straight years.
- Even with the CAPE-based rules, retirees had to tighten the belt by 20 to 36%.
- CAPE-rules also took even longer to recover than 28 years!

Most of the CAPE rules look OK to me.

- I will probably stick with CAPE 1.5/0.5, rather than dig into the complications of “CAPE robust” and “Best of three”.

ERN also comments on the other dynamic withdrawal rules that we have come across:

- Kitces ratcheting rule
- This is a stepped CAPE rule, to which ERN sees no advantage.

- The Bogleheads VPW (a variation on the US government’s minimum distributions from pensions).
- These rise with age and work well for those who don’t want to leave behind a final pot as a bequest.

- Guyton-Kinger guardrails
- These show the same steep withdrawal drawdowns as the constant percentage rule.

###### Conclusions

That’s it for today.

- We’ve covered another three of ERN’s articles, which takes us up to 18.

There are seven posts left in the series, so with a bit of luck we only have another couple of visits to go.

The first of today’s posts looked at SWRs in a low-return environment.

- Here, equity glide paths work well as a strategy.

As usual, SWRs above 3.25% pa look risky.

- I would have liked to have seen a comparison of fixed and dynamic (CAPE) withdrawal rules under this scenario.

The second post looked at how to account for future pensions (eg. the state pension).

- ERN supplies conversion tables to work out the (positive) impact on today’s SWR.

I prefer to capitalise the pension cash flows into a current value that I can add to my portfolio.

- This allows me to be more aggressive with (hold more equities in) the remainder of my portfolio.
- It also provides me with a current value against which to measure any potential transfer (cash-in) values.

The third article demonstrates that CAPE rules are superior to other dynamic withdrawal strategies, and the constant percentage rule in particular.

- Variable percentage / minimum distributions also work well if you don’t need to preserve value in the final pot in order to fund a bequest.

CAPE 1.5/0.5 still looks like the best rule to me.

Until next time.

*Share this with Twitter, Google+, Pinterest, LinkedIn, Tumblr, Reddit and StumbleUpon.*