If you had enough in stocks to retire on TIPS would you sell?

[jaysee]

Suppose you had enough in stocks that if you invested it all in a "safe" asset such TIPS you would have enough to last until say 100 years old.

For example, suppose you're 40, your living expenses are $30,000 per year, and you have $1.8M in stocks (after capital gains taxes, fees, etc).

Would you sell it all and invest it in a TIPS ladder?

This is perhaps similar to the philosophy espoused by Zvi Bodie.

There seem to be pros and cons to this:

Pro: Safe from market crashes, at least in theory
Pro: Very simple, reliable calculation on how much you can spend

Con: Miss out on potential big upside if stocks do even moderately well
Con: Not very diversified; putting lots of faith in the US government
Con: Risk of personal inflation outpacing CPI, e.g. if you spend most of your income on housing

My own approach: As an Australian I use eTIBs (our equivalent of TIPS). I decided to go with a split approach. Most of my money is in index funds but there's a hefty chunk in an eTIBs ladder, enough to cover some essentials like food and pubic transport. If I somehow end up depending on my eTIBs my plan is to work part-time to make up the rest of my costs.

I realise it's a pretty catastrophic and unlikely scenario that stocks as a whole would lose real purchasing power, but it seems worth at least considering.

What do you think?

[GuitarStv]

Seems like it would be sub-optimal.  You have to both believe that stocks are going to behave in a way that they never have before in history, and that your issuing government will still be in a good enough spot to keep paying out when it does.  I think that the safety is probably not as safe as you think and comes at too high a price.

[Heckler]

“Safety” to me means diversify, be average across multiple asset classes.

[bacchi]

Like the OP, I haven't put it all into TIPS but I have built a ladder for some of my expenses.

Another con for the all-in approach is that expenses may not always be $30k/year. For example, if medical expenses are $15k in a given year, you'd need to pull from next year's bonds. Since the bonds aren't expiring, they could be sold at a loss. That would ruin the ladder for next year and all successive years.

[Radagast]

LOL no. Apart from those listed, there is also reinvestment risk. TIPS are only good for 30 years, and haven’t even existed that long, and longer durations haven’t always existed. If you live to 100 but your last TIPS expires at 70 years and new ones aren’t available or rates a severely negative, or the government has screwed with them somehow to make their own ends meet, what will you do?

[LD_TAndK]

Oof I'd never accept 0% real return over 60 years. That's a long enough time horizon and enough of a cushion to ride out any downturn. I'd probably diversify my 60x portfolio in index funds by adding real estate or some other asset class with real expected returns

[jaysee]

Oof I'd never accept 0% real return over 60 years. That's a long enough time horizon and enough of a cushion to ride out any downturn. I'd probably diversify my 60x portfolio in index funds by adding real estate or some other asset class with real expected returns

60 years with 0% return does seem a pretty dire scenario. If market returns are that bad, I'd like to think society will figure out some alternative way to grow wealth!

[green2bend]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

[jaysee]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

Likely but not guaranteed. Are you willing to take the ~5% risk of financial ruin?

[nereo]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

Likely but not guaranteed. Are you willing to take the ~5% risk of financial ruin?

I would definitely not.
Also, your claim of “financial ruin” seems sensationalist.   More accurate might be “a 5% chance you need to alter your plan a fair bit”.

[curious_george]

Idk....

My grandfather is turning 102 years old this month. If I only had enough tips to last me to 100 years old I might be worried about running out of money someday.

It's also really challenging for me to imagine a scenario where the stock market crashes unlike it has ever crashed in American history, in the modern age, and the government also remains solvent and actually pays out the TIPS plus inflation and the taxes don't destroy the portfolio.

Personally, if I were this concerned about a potential downturn in society, I would go buy a farm next to an Amish community, then buy a lot of guns and canning supplies and learn how to garden and develop ties with the local community.

I do, despite our current political environment and my general distrust of strangers, still have a certain amount of hope and faith in society. If I didn't have a certain level of faith in government and society and human beings to collectively do the right thing to ensure their own survival, and by extension the survival of this thing called society because we all depend on each other, I would have never bothered trying to retire early to begin with.

This just seems like a suboptimal way to try and counter an improbable outcome.

[GuitarStv]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

Likely but not guaranteed. Are you willing to take the ~5% risk of financial ruin?

I would definitely not.
Also, your claim of “financial ruin” seems sensationalist.   More accurate might be “a 5% chance you need to alter your plan a fair bit”.

But the risk isn't going to be 5%.  The risk is pretty darned close to zero, isn't it?

The original example was 1.8 mil saved and pulling 30k a year.  You can make that work with TIPs.  But if you've got the money invested (let's say conservative 60/40 stocks/bonds), you're only doing a 1.7% withdrawal rate.  Plugging that into the rich, dead, or broke calculator (https://engaging-data.com/will-money-last-retire-early/), if you retire at 30 and live until you're 110 you have a 100% chance of dying before your money runs out.  If you manage to live to 75, you have a 98.6% chance of more than doubling your money and a 40% chance of more than quadrupling it.

If you have enough money to retire on TIPS, seems like you have more than enough money to weather a pretty bad economic downturn while invested.  The kind of catastrophe that would decimate all financial markets but leave your TIPS investment alone seems like a very, very, very unlikely occurance.

[Maximus28]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

Likely but not guaranteed. Are you willing to take the ~5% risk of financial ruin?

Using the numbers in your original post - $30k/year expenses and $1.8M invested - that results in a 1.67% withdrawal rate. Any withdrawal rate under 3.25%, invested in a more standard portfolio between 60% and 75% stock and 40% to 25% bonds, would have never failed in the last 100 years of stock market history.

It would be pretty difficult to mess up a retirement on a 1.67%WR with any portfolio - 100% stocks or 100% bonds. Knowing you have a 100% chance of success based on a century of stock market data, why would you want to significantly reduce the potential growth of the portfolio?

If you are 40 years old in the scenario, the amount of money left on the table would be tremendous if you chose 100% bonds.

[jaysee]

But the risk isn't going to be 5%.  The risk is pretty darned close to zero, isn't it?

That’s what I thought, but according to a recent paper:

“A 65-year-old couple willing to bear a 5% chance of financial ruin can withdraw just 2.26% per year…”
The Safe Withdrawal Rate: Evidence from a Broad Sample of Developed Markets
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4227132

[GuitarStv]

But the risk isn't going to be 5%.  The risk is pretty darned close to zero, isn't it?

That’s what I thought, but according to a recent paper:

“A 65-year-old couple willing to bear a 5% chance of financial ruin can withdraw just 2.26% per year…”
The Safe Withdrawal Rate: Evidence from a Broad Sample of Developed Markets
https://download.ssrn.com/23/07/24/ssrn_id4520188_code692959.pdf

I couldn't click on that link, but found this one (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4227132) which I think is what you're referencing.

In this study, we reevaluate the 4% rule for a US investor using a comprehensive new dataset of
real returns for domestic equity, international equity, government bonds, and government bills in
developed economies. The dataset is specifically constructed to combat the issues relating to small
samples, survivor bias, and easy data bias that impact the conclusions of prior studies. The data
cover approximately 2,500 years of asset-class returns in 38 developed countries over the period from
1890 to 2019. The long time series and broad cross-sectional coverage yield a wealth of statistical
information about potential investment outcomes and allow for a more complete characterization
of left-tail risk. Our fundamental premise is that US investors today can form better ex ante
expectations using a broad developed market sample free of survivor and easy data biases rather
than relying on the small, biased US sample.

Using this dataset, we simulate retirement outcomes associated with alternative withdrawal
rates. We model asset-class returns using a modification of the stationary block bootstrap of
Politis and Romano (1994) that maintains both cross-sectional and time-series properties of longhorizon asset returns. We incorporate longevity risk into the simulation design using mortality
tables from the Social Security Administration (SSA). Our base case simulation focuses on the
joint investment-longevity outcomes for a couple retiring in 2022 at age 65 who chooses a portfolio
strategy of 60% domestic stocks and 40% bonds.

The 4% rule proves woefully inadequate for current retirees. A retired couple faces a 17.4%
probability of depletion of financial wealth prior to death (henceforth, “financial ruin”) using the
4% rule, such that there is nearly a one-in-five chance that they must subsist — often for many
years — solely on social welfare programs. Given the poor performance of the 4% rule, we explore
alternative constant real withdrawal policies. Our findings suggest that most retirees (i.e., retirees
with relatively modest levels of wealth) cannot achieve a reasonable standard of living while maintaining a very low ruin probability. To achieve a 1% ruin probability, for example, retirees must
adopt a withdrawal rate of just 0.80% (i.e., $8,000 of withdrawals per year for $1 million in savings).
When we attempt to balance the desires to achieve a higher standard of living and to avoid financial
ruin, we find that a retired couple willing to bear a 5% ruin probability may withdraw 2.26% per
year. This value is considerably lower than those proposed in prior studies, and it is just over half
of the 4.22% rate implied by the post-1925 US data

They are using a weird data set that includes very heavy weighting to international stocks and bonds and don't count stuff like social security benefits in their calculations.  That leads to some weird statements to be made . . .

Given the inadequacy of popular rules for facilitating a safe retirement, we explore alternative
withdrawal rates. We find that there is no withdrawal rate that both supports a reasonable standard
of living for most retirees and nearly assures they will not outlive their wealth. Allowing for a 1%
ruin probability, for example, implies a withdrawal rate of just 0.80%, which provides only $667
per month of income for each $1 million in savings.

and

Even if a couple is willing to bear a 5% ruin probability, the [safe] withdrawal rate is just 2.26%.

This seems pessimistic to me and doesn't even match a simple smell test.  They're saying that a 40k withdrawal rate requires 5 million dollars invested for 99% certainty that you'll be OK in retirement.  But if you 'invest' that in a mattress and assume 2% inflation and 60 years retirement before death with regular monthly withdrawal from your mattress you would die with about 300,000 left in your 'savings'.  Their argument seems to be that investing money returns nothing to the investor?

[Radagast]

It's accurate, they're just including countries that lost wars or revolutions in their dataset, in the view that US returns are survivor bias. Add in Russia (commie revolution) France (Nazi occupation) Germany (lost two world wars and three currencies) Japan (lost one world war, then 1989) Austria (empire collapsed, still negative returns ~1900 to present) Spain (civil war with fascist takeover for two generations) Argentina (what even happened?) and the like and things look less rosy than the single best country in the world in terms of stock returns and top 3 or so in bonds.

[MustacheAndaHalf]

But the risk isn't going to be 5%.  The risk is pretty darned close to zero, isn't it?

That’s what I thought, but according to a recent paper:

“A 65-year-old couple willing to bear a 5% chance of financial ruin can withdraw just 2.26% per year…”
The Safe Withdrawal Rate: Evidence from a Broad Sample of Developed Markets
https://download.ssrn.com/23/07/24/ssrn_id4520188_code692959.pdf

Their simulation had results five times worse than reality for U.S. retirees.  After mentioning this glaring flaw, they offered no explanation or justification.  An unrealistic simulation generated unrealistic results.

"The simulation results suggest that Bengen’s (1994) original 4% rule exposes investors to considerable risk of outliving their wealth in retirement. The 4% rule leads to financial ruin in 17.4% of simulation runs."  (see bottom of page 11 in their paper)

"In contrast, a model based on the historical US experience leads to a more encouraging view of the popular withdrawal plans. The failure rate for the 4% rule is 3.5%"  (see top of page 12 in their paper)

[MustacheAndaHalf]

It's accurate, they're just including countries that lost wars or revolutions in their dataset, in the view that US returns are survivor bias. Add in Russia (commie revolution) France (Nazi occupation) Germany (lost two world wars and three currencies) Japan (lost one world war, then 1989) Austria (empire collapsed, still negative returns ~1900 to present) Spain (civil war with fascist takeover for two generations) Argentina (what even happened?) and the like and things look less rosy than the single best country in the world in terms of stock returns and top 3 or so in bonds.

This was an important point of the paper - that other developed countries should be included in simulations of retirement.  The paper's 17.4% failure rate vs 3.5% historical (4% rule) made me question their accuracy.  It would be interesting to see a simulation that both accurately models the U.S. historical experience, and includes data from other developed countries.

[jaysee]

I couldn't click on that link, but found this one (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4227132) which I think is what you're referencing.

Hi, yes that's the one. I've fixed the link now. Apologies, I typed the post up on my phone and the link didn't paste properly.

They are using a weird data set that includes very heavy weighting to international stocks and bonds

Yes it's a "weird" method – the so-called "block bootstrap" sampling. Not what you typically see in the literature.

But I think their method provides useful insight. As they write, this approach is intended to "combat the issues relating to small samples, survivor bias, and easy data bias that impact the conclusions of prior studies". Basically they recognise how unusually good the returns were in the recent past and want to account for that in their models. This seems sensible to me.

I also want to note that in their findings, international stocks offer better long-term safety than US-only or domestic stocks and than bonds. They're not advocating use of bonds - international or domestic. (My opinion is somewhat more nuanced ... see below.)

  That leads to some weird statements to be made . . .

Given the inadequacy of popular rules for facilitating a safe retirement, we explore alternative
withdrawal rates. We find that there is no withdrawal rate that both supports a reasonable standard
of living for most retirees and nearly assures they will not outlive their wealth. Allowing for a 1%
ruin probability, for example, implies a withdrawal rate of just 0.80%, which provides only $667
per month of income for each $1 million in savings.

and

Even if a couple is willing to bear a 5% ruin probability, the [safe] withdrawal rate is just 2.26%.

This is quite a shock, indeed it was to me. 5% isn't too bad but it's not nothing either. It seems to call for some kind of "safety net" in one's financial planning.

This seems pessimistic to me and doesn't even match a simple smell test.  They're saying that a 40k withdrawal rate requires 5 million dollars invested for 99% certainty that you'll be OK in retirement.  But if you 'invest' that in a mattress and assume 2% inflation and 60 years retirement before death with regular monthly withdrawal from your mattress you would die with about 300,000 left in your 'savings'.  Their argument seems to be that investing money returns nothing to the investor?

I don't think they're insisting that we need to have millions of dollars invested.

Actually, in my opinion, what you write above about cash "in a mattress" is pointing toward the real solution.

The real solution is to have some directly-held fixed income: cash in a high interest savings account and inflation-linked bonds. Not bond funds or anything subject to re-pricing, but directly-held fixed income which delivers a coupon and returns the full principle on maturity.

(Actually Cederburg himself refers to TIPS in both of his YouTube interviews with Ben Felix.)

This fixed income is what provides the "safety net" in case the 5% worst case transpires.

[Maximus28]

Yes it's a "weird" method – the so-called "block bootstrap" sampling. Not what you typically see in the literature.

But I think their method provides useful insight. As they write, this approach is intended to "combat the issues relating to small samples, survivor bias, and easy data bias that impact the conclusions of prior studies". Basically they recognise how unusually good the returns were in the recent past and want to account for that in their models. This seems sensible to me.

I also want to note that in their findings, international stocks offer better long-term safety than US-only or domestic stocks and than bonds. They're not advocating use of bonds - international or domestic. (My opinion is somewhat more nuanced ... see below.)

Safe withdrawal rates are determined by the worst-case scenarios, the unusually good returns of the recent past would not have any effect on the SWRs calculated based upon US market returns. The Great Depression and the extreme inflation of the 1970's are the 2 worst-case scenarios for US stock market returns and those 2 events typically define the SWR.

Asset allocation, and its importance to both portfolio survivability and longevity, has been studied very thoroughly at Early Retirement Now: https://earlyretirementnow.com/start-here/

TIPs are not a permanent solution to sequence of returns risk (SORR); Karsten explains this in detail in this post: https://earlyretirementnow.com/2024/05/16/safety-first-swr-series-part-61/

One of the most effective solutions to SORR is a reverse equity glidepath. The retiree would start with a heavy allocation to bonds (~35%stocks/65% bonds) and slowly "glide" up the equity allocation to (75%stocks/25%bonds) during the first 10 to 15 years of retirement.

Reverse equity glidepaths do not solve all problems, in some scenarios a static allocation like 60/40 or 75/25 will do better. Karsten concludes that reverse equity glidepaths are most effective when CAPE is above 20.

[vand]

I doubt 100% TIPS is an optimal asset allocation for anyone, and it's kindof a backward question; if you have enough in stocks to retire then you have more than enough in stocks/gold or some other more suitable asset allocation to retire. 

[jaysee]

Safe withdrawal rates are determined by the worst-case scenarios, the unusually good returns of the recent past would not have any effect on the SWRs calculated based upon US market returns. The Great Depression and the extreme inflation of the 1970's are the 2 worst-case scenarios for US stock market returns and those 2 events typically define the SWR.

Those are the historical worst cases. But how do you know worse won't happen in future?

The risk might not even show up as nominal downturns.

It might just be stock returns trailing inflation by 3% over a couple of decades. It doesn't seem like much when put in those numbers. But imagine the compounding effect over 20 years. You would lose nearly 50% of your purchasing power. Even if there's a recovery, via higher returns or disinflation/deflation, you'd be starting from a much lower base. Would you recover in time to not exhaust your funds?

Asset allocation, and its importance to both portfolio survivability and longevity, has been studied very thoroughly at Early Retirement Now: https://earlyretirementnow.com/start-here/

TIPs are not a permanent solution to sequence of returns risk (SORR); Karsten explains this in detail in this post: https://earlyretirementnow.com/2024/05/16/safety-first-swr-series-part-61/

True, longevity risk is a problem with TIPS ladders. However I don't see how this fact contradicts my argument about stock returns being unpredictable. It just means that there's a risk on the TIPS side.

In summary:

• On the equities side, we have a risk of unexpected future "black swans", such as big downturns or not keeping up with inflation.
• On the TIPS side, we have a risk of outliving our funds.

Well, I guess each asset type has its risks. This doesn't mean "just give up on TIPS". It means that there are risks to TIPS just as there are to stocks.

How do we deal with the fact that each side has its risks?

I guess one way is to diversify - hold a bit of each and hope that the risks won't show up on both sides simultaneously (that would be very bad luck indeed).

Another potential solution is to purchase real-return annuities in older age, if/when they ever return to the market.

Another solution (one I'm banking on, maybe a little too much) is the government pension as a last resort. In the US, this would be social security.

Finally, and this is where I think Mustache really shines, there is good old fashioned part-time work, combined with maybe a little DIY and belt-tightening. Maybe this is the safest. It's hard to imagine an economy where the stock market has imploded, pensions / social security are gone, TIPS don't work, and inflation is high, where there won't be some form of employment available!

One of the most effective solutions to SORR is a reverse equity glidepath. The retiree would start with a heavy allocation to bonds (~35%stocks/65% bonds) and slowly "glide" up the equity allocation to (75%stocks/25%bonds) during the first 10 to 15 years of retirement.

Reverse equity glidepaths do not solve all problems, in some scenarios a static allocation like 60/40 or 75/25 will do better. Karsten concludes that reverse equity glidepaths are most effective when CAPE is above 20.

Isn't this also based on historical data, which might not necessarily repeat?

[MustacheAndaHalf]

TIPs are not a permanent solution to sequence of returns risk (SORR); Karsten explains this in detail in this post: https://earlyretirementnow.com/2024/05/16/safety-first-swr-series-part-61/

Since I'm not interested in TIPS, I only browsed the first part of the article.  The graph showing negative TIPS yields should include a note that the Fed was dumping TIPS as part of its quantitative tightening.  That may have artificially impacted yields.  That said, I have no interest in retiring on a 100% TIPS strategy, so I didn't read the whole article.

One of the most effective solutions to SORR is a reverse equity glidepath. The retiree would start with a heavy allocation to bonds (~35%stocks/65% bonds) and slowly "glide" up the equity allocation to (75%stocks/25%bonds) during the first 10 to 15 years of retirement.

Reverse equity glidepaths do not solve all problems, in some scenarios a static allocation like 60/40 or 75/25 will do better. Karsten concludes that reverse equity glidepaths are most effective when CAPE is above 20.

The links you provided already were interesting - do you have links showing equity glidepaths vs CAPE?

That's especially relevant now, when CAPE is the third highest in history, soon to be second.  Currently, 2021 ended with a higher CAPE, followed by a crash to undo overvaluation.  The highest was the dot-com crash, another time when valuations crashed back to reality.  Although it is the third highest right now, it will pass up the 2021 levels in a matter of 1-2 months.

dot-com crash: 43.8
2021 peak: 38.58
now: 37.93
https://www.multpl.com/shiller-pe

[ChpBstrd]

IMO, the most attractive rationale for a treasuries-heavy AA is if valuations were getting frothy and you felt very bad news was on the horizon.

The play would be to hide out in treasuries for 1-5 years waiting for the bear market to happen. Then you have a criteria to pivot into riskier assets after markets have fallen to your pre-selected nominal price, yield, or valuation metrics. For example, your IPS might say "Pivot into a stock-heavy AA as soon as the S&P500 is 20% lower than its value on 11/15/2024". Most importantly, you follow that plan no matter how bad things seem to be getting.

This may seem like market timing, but it's more akin to playing the odds. When market valuations are high and people are flinging their money at stupid ideas, there is a higher chance that you'll get an opportunity to buy at much cheaper prices in the future.

Winning this gamble means locking in higher long-term asset returns and a higher SWR than if you bought at a higher price.

Losing this gamble means missing out on the difference between treasury returns and the return of an riskier AA.

If we quantify these outcomes, the comparison gets easier. For example, we could say "winning means buying a million dollars worth of assets at 20% cheaper" and "losing means having a million dollars earn 3% less than it could have earned per year for up to 5 years". We can calculate the Expected Value of these outcomes. Winning would be worth $200k. Losing would cost $-159,274. Your specific numbers may vary. The breakeven odds on the numbers I'm using as an example would be a 44.34% odds of bear market and 55.66% odds of no bear market (i.e. 200k * 0.4434 is within a few bucks of 159,274 * 0.5566). So it would make sense to take the treasuries route if you believed the true odds of the bear market scenario were greater than 44.34%.

Of course the world is more complicated than this and the possibilities are not binary. Stocks could rise 20% before they go down 20%, for example, or only fall 19% before zooming higher, and not trigger the pivot in one's IPS. Plus, as we saw in 2022, sitting in long-duration treasuries can result in losses if interest rates rise. So if that was your plan, add a contingency with estimated odds in which you lose money due to higher rates. Finally, the gap between stock returns and treasury yields could prove to be a bad estimate. But in general this sort of math is persuasive if one assigned a high probability to a cheaper entry point in the future.

[green2bend]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

Likely but not guaranteed. Are you willing to take the ~5% risk of financial ruin?

I thought the question was whether to sell your existing portfolio and go 100% inflation protected treasuries. It seems like a strawman argument to compare a 100% inflation protected treasury portfolio to a 100% equities portfolio. As others have said, even if you really have a 100% equities portfolio then you still only have a 1.7% withdrawal rate which is crazy low. I think that at that withdrawal rate, the chances that a portfolio with broad based index funds fails but a 100% inflation protected treasury portfolio succeeds is 0%. At that point, the world has completely collapsed in which case no kind of money would help. And if it's not 0%, then something happened that's never happened before and I would think TIPS would be just as likely to be wiped out as equities.

[SeattleCPA]

No. Stocks should significantly outperform over the long term, and I wouldn't be willing to pay that opportunity cost. Also, I'm way too lazy to have to maintain a TIPS ladder.

Likely but not guaranteed. Are you willing to take the ~5% risk of financial ruin?

+1

[Maximus28]

I guess one way is to diversify - hold a bit of each and hope that the risks won't show up on both sides simultaneously (that would be very bad luck indeed).

Another potential solution is to purchase real-return annuities in older age, if/when they ever return to the market.

Another solution (one I'm banking on, maybe a little too much) is the government pension as a last resort. In the US, this would be social security.

Finally, and this is where I think Mustache really shines, there is good old fashioned part-time work, combined with maybe a little DIY and belt-tightening. Maybe this is the safest. It's hard to imagine an economy where the stock market has imploded, pensions / social security are gone, TIPS don't work, and inflation is high, where there won't be some form of employment available!

A 60/40, 70/25 or reverse equity glidepath are already diversified. One can add international exposure to the stocks and bonds components if there is a desire to further diversify based on US specific risk. As I understand, the scenario you are proposing is that a 1.7% withdrawal rate cannot be sustained with any of those portfolios. This has never happened, so you are discussing an event worse than the Great Depression. If that happened, I don't think you could rely on US annuities or US government pensions either. An international weighting may help in this scenario, but that isn't guaranteed either.

Reverse equity glidepaths do not solve all problems, in some scenarios a static allocation like 60/40 or 75/25 will do better. Karsten concludes that reverse equity glidepaths are most effective when CAPE is above 20.

Isn't this also based on historical data, which might not necessarily repeat?

Sure, what other method would you use for modeling purposes if 100 years of stock market data cannot be relied upon to create a basic foundation for the portfolio?

The paper you referenced earlier states that a 0.80% withdrawal rate is needed to achieve 1% probability of failure. To me, that is a red flag. Since 1972, a money market fund would have returned 0.6% real return and no risk. I'm guessing the failures in this paper driving the withdrawals rates so low were due to heavy weighting in a specific country that lost a war while drawing down the portfolio. Without the details of the specific failure cases i.e. time frame, portfolio composition, this is just guess work to try to determine why their SWRs are so low.

[jaysee]

A 60/40, 70/25 or reverse equity glidepath are already diversified.
One can add international exposure to the stocks and bonds components if there is a desire to further diversify based on US specific risk.

The Cederburg study finds that international with home country bias gives a better return, mainly because bonds, long-term, cannot keep up with inflation, and thus drag down the portfolio.

This is why I emphasise directly-held TIPS, which might keep up better with inflation and would give you coupons and principle protection in the event of a crash during interest rate increases (such as what we had recently).

As I understand, the scenario you are proposing is that a 1.7% withdrawal rate cannot be sustained with any of those portfolios. This has never happened, so you are discussing an event worse than the Great Depression. If that happened, I don't think you could rely on US annuities or US government pensions either. An international weighting may help in this scenario, but that isn't guaranteed either.

I believe it has happened more often than just the Great Depression. According to the Cederburg study, there's a 5% risk of total loss. Yikes!

Reverse equity glidepaths do not solve all problems, in some scenarios a static allocation like 60/40 or 75/25 will do better. Karsten concludes that reverse equity glidepaths are most effective when CAPE is above 20.

I guess ERN / Karsten's approach is different from Cederburg's and that's why they get different results. I'm not really sure who to believe. Probably I should investigate Karsten more.

Isn't this also based on historical data, which might not necessarily repeat?

Sure, what other method would you use for modeling purposes if 100 years of stock market data cannot be relied upon to create a basic foundation for the portfolio?

My understanding is that Cederburg & co uses a "block bootstrap simulation". So they take historical data as a starting point, but then use it to build some kind of random simulations, and then measure which % of those simulations succeed vs. fail.

This is supposed to remove certain biases, e.g. survivorship bias.

One problem though - maybe this suffers from what Taleb calls the "ludic fallacy". Even though it's randomised, you're still using the historical data as input. This means you're limiting the range of possible outcomes to what is capture in the historical data. But there's no reason the future should be anything like the past. So the future could have wildly different outcomes.

I guess making predictions is really hard, especially about the future. 😄

The paper you referenced earlier states that a 0.80% withdrawal rate is needed to achieve 1% probability of failure. To me, that is a red flag. Since 1972, a money market fund would have returned 0.6% real return and no risk. I'm guessing the failures in this paper driving the withdrawals rates so low were due to heavy weighting in a specific country that lost a war while drawing down the portfolio. Without the details of the specific failure cases i.e. time frame, portfolio composition, this is just guess work to try to determine why their SWRs are so low.

Yes that's correct and he said as much in the RR interview.

My fear is that, even if it's not war, some other kind of "unknown-unknown" could cause stocks to fail.

However I guess you could say the same about TIPS or anything really. At that point you're just hand waving and not actually doing "science".

So this all leaves me feeling, maybe TIPS aren't really a safe-haven, but more of a temporary diversification/hedge in case stocks do poorly during a specific period. Beyond that, we can't expect anything more from them than we can expect from stocks. There's no way to hedge against a complete unknown.

[GuitarStv]

I don't think they're insisting that we need to have millions of dollars invested.

Actually, in my opinion, what you write above about cash "in a mattress" is pointing toward the real solution.

From my perspective cash in a mattress is quite extreme.  Again, this 'investment' approach works best only if you assume that all investments will fail the majority of the time - so I agree, they're not insisting that you need millions invested because the upshot of the article is entirely anti-investment.

I'd be interested to re-read the study if instead of using dozens of countries teetering on the precipice of disaster and collapse they instead used a global market index like VTI (extrapolating back using the same stock identifiers to simulate what the results would have been in the past were the index around prior to inception) to do their calculations.  My suspicion is that the conclusions would be radically different and heavily tilted towards investing rather than stuffing money in a mattress.

This is a classic case of bad data and initial assumptions leading to bad results.

[Maximus28]

My understanding is that Cederburg & co uses a "block bootstrap simulation". So they take historical data as a starting point, but then use it to build some kind of random simulations, and then measure which % of those simulations succeed vs. fail.

This is supposed to remove certain biases, e.g. survivorship bias.

One problem though - maybe this suffers from what Taleb calls the "ludic fallacy". Even though it's randomised, you're still using the historical data as input. This means you're limiting the range of possible outcomes to what is capture in the historical data. But there's no reason the future should be anything like the past. So the future could have wildly different outcomes.

I think this phenomenon makes these "block bootstrap simulations" unreliable. As I understand those simulations, they are iterating thousands of different portfolio compositions with huge variations in weighting to specific countries. The failure cases probably do not make logical sense to hold but they are generated in that methodology. Again, we would have to see the failures and what portfolio was held, but if some of the failure cases were 100% German equities in/after WW2, what do we really learn from that? Don't hold majority equity in a country that just lost a war? That's pretty obvious, but a lot harder to implement unless you have a crystal ball.

To me, ERNs approach is much more scientific. Stress test a few specific, reasonable portfolios with the goal of maximizing SWR in all scenarios in the last 100 years. This bootstrap simulation is completely backwards, they just run an enormous amount portfolios (most of them probably garbage/unrealistic) and look for the worst SWR that emerges from all of those cases.

However I guess you could say the same about TIPS or anything really. At that point you're just hand waving and not actually doing "science".

So this all leaves me feeling, maybe TIPS aren't really a safe-haven, but more of a temporary diversification/hedge in case stocks do poorly during a specific period. Beyond that, we can't expect anything more from them than we can expect from stocks. There's no way to hedge against a complete unknown.

Bonds perform the hedge you are describing as well, not just TIPS. The reverse equity glidepath is probably the most effective way to hedge against the unknown. The downside to this approach is that the investor will miss out on some gains if equity returns are great in the first 10 years of retirement. However, if the goal is to minimize risk of failure, then that tradeoff is acceptable. In the attached graph, ERN modeled a linear 25-year reverse glidepath and compared that to other portfolios with retirement starting at the great depression.

ERN also did a great analysis on international diversification of equities, and it turns out that rarely helps in US recessions. Typically when the US is in recession, global economies go into recession as well so the correlations usually align and both domestic and international equities drop at the same time. That is the opposite of diversification.

See his post here about international diversification: https://earlyretirementnow.com/2017/08/23/how-useful-is-international-diversification/

[MustacheAndaHalf]

My understanding is that Cederburg & co uses a "block bootstrap simulation". So they take historical data as a starting point, but then use it to build some kind of random simulations, and then measure which % of those simulations succeed vs. fail.
...
I guess making predictions is really hard, especially about the future. 😄

The Cederburg simulation couldn't even predict the past, assigning 5x higher failure rates than seen in reality.  You are putting a lot of faith in something that did not match reality.

[jaysee]

My understanding is that Cederburg & co uses a "block bootstrap simulation". So they take historical data as a starting point, but then use it to build some kind of random simulations, and then measure which % of those simulations succeed vs. fail.
...
I guess making predictions is really hard, especially about the future. 😄

The Cederburg simulation couldn't even predict the past, assigning 5x higher failure rates than seen in reality.  You are putting a lot of faith in something that did not match reality.

That's to its credit, it's not meant to predict the past, it's meant to predict the future. It's future returns that we care about.

[GuitarStv]

My understanding is that Cederburg & co uses a "block bootstrap simulation". So they take historical data as a starting point, but then use it to build some kind of random simulations, and then measure which % of those simulations succeed vs. fail.
...
I guess making predictions is really hard, especially about the future. 😄

The Cederburg simulation couldn't even predict the past, assigning 5x higher failure rates than seen in reality.  You are putting a lot of faith in something that did not match reality.

That's to its credit, it's not meant to predict the past, it's meant to predict the future. It's future returns that we care about.

Ah.  So fallacy is a pro in your mind?

If your weatherman was wrong 100% of the time in the past eighty years would you trust him when he says it's going to rain today, even though there aren't any clouds in the sky?