Your Safe Withdrawal Rate Is Not Safe

[CM]

You must be very rich or very frugal (or very daring) to retire today if you depend on your portfolio (rather than a pension). This is especially true if you retire early.

Bond yields are historically low. The Shiller PE (a measure of stock market valuation) is at the third highest peak in history. This means that future returns will be low and the sequence-of-returns-risk is high.

This paper from authors including Wade Pfau provides sobering results: http://corporate.morningstar.com/us/documents/targetmaturity/LowBondYieldsWithdrawalRates.pdf. For example, the safe withdrawal rate from a 60:40 stock:bond portfolio for a 30-year horizon is 2.4% (5% failure rate), and for a 40-year horizon it's 1.8% (5% failure rate).

Your outcomes might be better because the authors impose a 1% fee on returns, but that is perhaps 90 basis points more than a DIY investor will pay.

On the other hand, the paper was written in January 2013 and valuations are worse today. The Shiller PE was about 21.9 then, but it is about 29.1 now. The authors used Ibbotson’s 2012 long-term capital market forecasts for stock returns and projected an arithmetic mean return 2% lower than the historical average. I project a geometric mean return of 3.27% now versus a long-term average of 6.8%, 3.53% lower than the historical average. (This projection is based on the current 2.02% dividend and the 1.25% long-term real growth rate of earnings with no fall in valuation, ever. For slightly more detail, see my comment at 4:26 pm on 11/6/16 to "25 Pearls About the Financial Life of Physicians," https://www.whitecoatinvestor.com/25-pearls-about-the-financial-life-of-physicians/.)

Further, the authors assume that bond yields will drift upward toward the long-term average over time, providing a 2% real return rather than the negative real return we have today. They use the Ibbotson Intermediate-Term Government Bond Index as a proxy for bonds and use a starting yield of 2.5%, but Vanguard's intermediate-term government bond index (VSIGX, Admiral Shares) has an SEC yield of 1.93% today.

I'll consider retirement in 3-6 years if we have a significant bear market in stocks by then, but if valuations remain high I'll look hard for some type of part-time arrangement.

 

[Zaphod]

Itd be nuts to be in a 60/40 portfolio right now, until the regime changes further. We will not get back to "normal" rates. There is no such thing, thats an anchor and recency bias thing. Any younger person with that portfolio set up is asking a lot and not taking an appropriate amount of risk.

Returns do have a high likelihood of being lower going forward, however, real returns may be similar enough that it doesnt matter. Inflation isnt as powerful as it once was, and commissions and fees to invest are nearly zero. This is a huge difference than any other time in history where these drags were very very significant. Thats mostly what the difference in these numbers represent, frictions in the system. You can always shift assets to other areas with lower CAPEs, and as always CAPE ignores market structure change, blows off accounting effects, and cannot predict the future etc....

If you have a couple years worth of liquidity/bonds saved up to last through a bear market without drawing down your portfolio at the worst time, you're probably going to be more than okay.

I'd worry more about this stuff if anyone had any kind of track record of correctly predicting long term market returns outside of it will trend up nominally over time interrupted by some crashes lower.

[TheGipper]

I've read much of Wade Pfau's work, and honestly have learned to be skeptical. I encourage you to be too. Planning as if you can only afford a 2% withdrawal rate is in some ways just as irresponsible as planning for a 6% withdrawal rate.

But if you want to work another 5 years to feel more comfortable, it's your choice.

I do agree it's always nice to retire deep into a bear market, even though people often do the opposite.

If I were retiring today I might temporarily bump my bond allocation up to 40+% for 3-5 years to protect against early retirement sequence of return risk and then let my equity allocation increase back up to 80%+, similar to recent Kitces "bond tent" article approach.

[CM]

Zaphod wrote:

Itd be nuts to be in a 60/40 portfolio right now, until the regime changes further. We will not get back to “normal” rates. There is no such thing, thats an anchor and recency bias thing. Any younger person with that portfolio set up is asking a lot and not taking an appropriate amount of risk.

Returns do have a high likelihood of being lower going forward, however, real returns may be similar enough that it doesnt matter. Inflation isnt as powerful as it once was, and commissions and fees to invest are nearly zero. This is a huge difference than any other time in history where these drags were very very significant. Thats mostly what the difference in these numbers represent, frictions in the system. You can always shift assets to other areas with lower CAPEs, and as always CAPE ignores market structure change, blows off accounting effects, and cannot predict the future etc….

If you have a couple years worth of liquidity/bonds saved up to last through a bear market without drawing down your portfolio at the worst time, you’re probably going to be more than okay.

I’d worry more about this stuff if anyone had any kind of track record of correctly predicting long term market returns outside of it will trend up nominally over time interrupted by some crashes lower.

Click to expand...

1. The paper provides SWRs for 20%, 40%, 60%, and 80% equity allocations. They're all low because stock returns will be lower than in the past, but real bond returns are negative now.

2. a: Inflation doesn't affect real returns.

b: Commissions and fees are not included in historical returns. (If some estimate of commissions was included in historical returns, they would be much lower than the reported results.) Future returns are likely to be lower than these historical returns due to high valuations rather than changes in commissions and fees.

c: You can look for better investments than the S&P 500, and many international markets do appear to offer better prospective returns now. However, I believe the Bengen and Trinity studies used US stocks and bonds to develop safe withdrawal rates. Those US rates are much higher than most other developed markets because US returns have been higher. That was never a safe assumption going forward, but especially unsafe given current valuations.

d: CAPE does predict future long-term returns (not short-term), albeit with a fairly wide dispersion of results. There are many other valuation measures that predict future returns (if you think CAPE will lose its predictive ability going forward), and they all point in the same direction: lower prospective returns. Valuations predict returns by simple arithmetic (https://www.whitecoatinvestor.com/25-pearls-about-the-financial-life-of-physicians/).

3. Bonds are unlikely to plunge in value like equities, but the problem is that they provide a negative or near-negative real return now, compared to something closer to 2% real historically. Thus the bond portion of your portfolio is locked into a sequence-of-returns problem if you retire now. (Caveat: If you own only treasury bonds, their value will likely appreciate in a stock market crash. Thus I own LT treasury bonds because I think a stock bear market is likely over next 3-6 years.)

[CM]

TheGipper wrote:

I’ve read much of Wade Pfau’s work, and honestly have learned to be skeptical. I encourage you to be too. Planning as if you can only afford a 2% withdrawal rate is in some ways just as irresponsible as planning for a 6% withdrawal rate.

But if you want to work another 5 years to feel more comfortable, it’s your choice.

I do agree it’s always nice to retire deep into a bear market, even though people often do the opposite.

If I were retiring today I might temporarily bump my bond allocation up to 40+% for 3-5 years to protect against early retirement sequence of return risk and then let my equity allocation increase back up to 80%+, similar to recent Kitces “bond tent” article approach.

Click to expand...

This paper uses reliable third-party data, explains assumptions and methods, and reports results. You can draw your own conclusions without regard to what Pfau might think is prudent.

Perhaps I weight the risk of poverty in old age too heavily, but I think the general confidence in the "4% Rule" is badly misplaced. It's an anomaly of the time period and the country. From 1900 through 2012, no developed country posted a SWR of 4% (if we can trust Pfau to utilize Dimson's data correctly): Does International Diversification Improve Safe Withdrawal Rates? (https://www.advisorperspectives.com/articles/2014/03/04/does-international-diversification-improve-safe-withdrawal-rates)

Regarding irresponsible 2% withdrawal rates: Seven of the 20 developed markets had safe withdrawal rates below 2% (all between 0.25% and 1.39%)!

[VagabondMD]

I think that the key is not to accept any rigid rules and aim to be flexible and have more than you need. In theory, you could certainly argue that no amount of money is enough or that 10% returns are expected indefinitely. I consider the 4% rule to be a guideline, one of many, and to protect myself as best I can for any reasonable variations.

[Zaphod]

CM wrote:

TheGipper wrote:

I’ve read much of Wade Pfau’s work, and honestly have learned to be skeptical. I encourage you to be too. Planning as if you can only afford a 2% withdrawal rate is in some ways just as irresponsible as planning for a 6% withdrawal rate.

But if you want to work another 5 years to feel more comfortable, it’s your choice.

I do agree it’s always nice to retire deep into a bear market, even though people often do the opposite.

If I were retiring today I might temporarily bump my bond allocation up to 40+% for 3-5 years to protect against early retirement sequence of return risk and then let my equity allocation increase back up to 80%+, similar to recent Kitces “bond tent” article approach.

Click to expand…

This paper uses reliable third-party data, explains assumptions and methods, and reports results. You can draw your own conclusions without regard to what Pfau might think is prudent.

Perhaps I weight the risk of poverty in old age too heavily, but I think the general confidence in the “4% Rule” is badly misplaced. It’s an anomaly of the time period and the country. From 1900 through 2012, no developed country posted a SWR of 4% (if we can trust Pfau to utilize Dimson’s data correctly): Does International Diversification Improve Safe Withdrawal Rates? (file:///C:/Users/Curt/Downloads/does-international-diversification-improve-safe-withdrawal-rates.pdf)

Regarding irresponsible 2% withdrawal rates: Seven of the 20 developed markets had safe withdrawal rates below 2% (all between 0.25% and 1.39%)!

Click to expand...

Yes Im aware of the outlying nature of the US returns, its very interesting stuff, especially given length of draw downs, etc...

Remember the trinity study does not say that 4% is a sustainable withdrawal rate for 60/40, it says to be safe, one should never have less than a 75/25 allocation. That is wildly different than how it is misinterpreted and applied. Thats not a great starting point, a straw man argument. It also says for longer periods that one should slightly lower the rate and have more stock that 50/50. Lets recall most of those scenarios ended up with many multiples of their starting amount. A 100% probability is very likely too conservative of an overall plan. This also does not account for spending patterns and other assets (as physicians we dont need nearly the income once we retire to even live the same lifestyle given taxes and other pay to play issues we have). Pfau also uses different bond investments which of course means the methodology and interpretation are going to be different no matter what.

Pfau says returns will be lower. Maybe, or maybe he has no idea of the future like everyone else. Agree its a high probability, today. However, the money that you invested from 2008-2013 will enjoy out sized returns over your lifetime, even if the 2013-17 is lame they will average out to hopefully good enough. Then in the next down turn that money invested again has out sized returns. Thats the cycle. Unless you have all of your money in a bank and deployed it this year, the extrapolation from this single year is a whole lot less important. Our savings rate is whats most important and we control that, and even with lower returns packing it in while we can is what gives us an advantage.

On a real basis bonds have been a loser most of the time, inflation is their enemy. Thats not really their role, but it does bother me a lot.

[Live Free MD]

I agree that there is no completely safe (100% effective) withdrawal rate.  However, I think it is reasonable to pick a "reasonably safe" withdrawal rate, such as 3%, and then try to have multiple sources of income during an early "retirement", which could include not only your investment income, but also real estate income, part-time consulting, a super low stress part-time job (like working at REI), passive income from a book or website, or (for the lucky few) a pension.

[CM]

Live Free MD wrote:

I agree that there is no completely safe (100% effective) withdrawal rate.  However, I think it is reasonable to pick a “reasonably safe” withdrawal rate, such as 3%, and then try to have multiple sources of income during an early “retirement”, which could include not only your investment income, but also real estate income, part-time consulting, a super low stress part-time job (like working at REI), passive income from a book or website, or (for the lucky few) a pension.

Click to expand...

.

[Zaphod]

CM wrote:

Live Free MD wrote:

I agree that there is no completely safe (100% effective) withdrawal rate.  However, I think it is reasonable to pick a “reasonably safe” withdrawal rate, such as 3%, and then try to have multiple sources of income during an early “retirement”, which could include not only your investment income, but also real estate income, part-time consulting, a super low stress part-time job (like working at REI), passive income from a book or website, or (for the lucky few) a pension.

Click to expand…

Nice job on your blog. My annual expenses have been similar to yours, but I’m not comfortable walking away on $2 million even at an older age: http://www.livefreemd.com/how-to-earn-your-freedom/.

I already “retired” once on $1.28 million at about 42 (about $1.789 million in today’s dollars) because I was so burned out I couldn’t stand another day. I wasn’t familiar with the 4% rule then, but I estimated that I could spend about $40K/year (about $55,900 in today’s dollars, a 3.125% withdrawal rate). The lifestyle was ok, but I always felt the Sword of Damocles was dangling overhead. There wasn’t much margin for error, and untoward events could have turned my modest existence into something significantly worse. It’s comfortable to live on $60K when you have a steady income, but it feels precarious (at least to me) when you surrender all human capital.

That is why I favor your idea of part-time consulting/low stress jobs/book/blog etc. These sorts of things still allow you to live a stress-free life with the time and energy to pursue your interests, but they preserve some human capital which is a great insurance policy against poverty.

In my case, I created some new human capital by studying accounting and finance with most of my free time. That led to a job as a healthcare equity analyst, a position that almost fell into my lap. Years later I returned to cardiology in a slightly less taxing position than my original solo private practice.

Despite that good fortune, I might not be successful if I try to return to work in my late 60s after lying on the beach for years on end.

Click to expand...

The margin of safety is the real issue, and mostly with regard to the cost of healthcare since it seems still very likely to cause undue hardship for things you dont really have control over. Even if you're a frugal person, you dont want to have to rely on that just to make it.

[PhysicianOnFIRE]

I've linked to it several times on my site, but if you want to read an incredibly thorough investigation of safe withdrawal rates, a well-studied Ph.D. economist has published 10 posts in a series on Safe Withdrawal Rates. Start here.

The take-home points for me:

1. To be truly safe (as safe as safe can be), plan on a starting withdrawal rate of 3.5% or less and have some flexibility.

2. For a longer time frame (i.e. greater than 30 years), plan to be heavy on stocks.

3. This guy knows more about the topic than I ever will.

[CM]

Thanks PoF. That is a terrific series. I think this is the most useful installment: https://earlyretirementnow.com/2016/12/21/the-ultimate-guide-to-safe-withdrawal-rates-part-3-equity-valuation/.

"Quite intriguingly, over the 30-year horizon (top panel) and for equity weights greater than 40%, every single failure of the 4% rule occurred when the CAPE was above 20 at the start of the retirement. In contrast, for all CAPE<20 you have a 100% success rate. I wish the original authors of the Trinity Study had dug deeper into when those failures occur."

However, as I recall the author assumes "normal" (6.6% real) equity returns in the future, even though the current CAPE is 29 (27 at the time of publication). I believe he reports there is no volatility in his return assumption for future years. If so, this will lead to an underestimation of failure rates going forward. He also lumps all CAPE values between 20 and 30 in one category, but we're at the very top of the range now.

He is also reporting historical outcomes based on a limited number of overlapping periods. (There aren't many points in history with CAPE > 29.) That's why I prefer a Monte Carlo simulation (as in the first paper above) based on likely returns and volatility given the circumstances today.

I should probably suggest assumptions and ask him to run a Monte Carlo simulation. He asks for reader requests. I'll try to remember to do that. (First I'll need to read the newer posts, and make sure I'm familiar with all of the assumptions and methods he used.)

[Live Free MD]

Zaphod wrote:

CM wrote:

Live Free MD wrote:

I agree that there is no completely safe (100% effective) withdrawal rate.  However, I think it is reasonable to pick a “reasonably safe” withdrawal rate, such as 3%, and then try to have multiple sources of income during an early “retirement”, which could include not only your investment income, but also real estate income, part-time consulting, a super low stress part-time job (like working at REI), passive income from a book or website, or (for the lucky few) a pension.

Click to expand…

Nice job on your blog. My annual expenses have been similar to yours, but I’m not comfortable walking away on $2 million even at an older age: http://www.livefreemd.com/how-to-earn-your-freedom/.

I already “retired” once on $1.28 million at about 42 (about $1.789 million in today’s dollars) because I was so burned out I couldn’t stand another day. I wasn’t familiar with the 4% rule then, but I estimated that I could spend about $40K/year (about $55,900 in today’s dollars, a 3.125% withdrawal rate). The lifestyle was ok, but I always felt the Sword of Damocles was dangling overhead. There wasn’t much margin for error, and untoward events could have turned my modest existence into something significantly worse. It’s comfortable to live on $60K when you have a steady income, but it feels precarious (at least to me) when you surrender all human capital.

That is why I favor your idea of part-time consulting/low stress jobs/book/blog etc. These sorts of things still allow you to live a stress-free life with the time and energy to pursue your interests, but they preserve some human capital which is a great insurance policy against poverty.

In my case, I created some new human capital by studying accounting and finance with most of my free time. That led to a job as a healthcare equity analyst, a position that almost fell into my lap. Years later I returned to cardiology in a slightly less taxing position than my original solo private practice.

Despite that good fortune, I might not be successful if I try to return to work in my late 60s after lying on the beach for years on end.

Click to expand…

The margin of safety is the real issue, and mostly with regard to the cost of healthcare since it seems still very likely to cause undue hardship for things you dont really have control over. Even if you’re a frugal person, you dont want to have to rely on that just to make it.

Click to expand...

It is challenging to budget for the cost of health insurance if you stop working once you reach financial independence.  One approach would be to estimate health insurance costs based upon the current cost of obtaining health insurance on the healthcare.gov exchange.  Depending upon the cost of healthcare in your area, this might be around 15-20K per year for a couple.  Another approach would be to plan on picking up a chill part-time job once you are financially independent that may provide health insurance as well as some extra spending money.

[CM]

Just noticed something (hat tip PoF and ERN): I projected 3.27% annual real prospective return for S&P 500 in post above.

Based on the first scatter plot in this post, https://earlyretirementnow.com/2016/12/21/the-ultimate-guide-to-safe-withdrawal-rates-part-3-equity-valuation/, the Early Retirement Now author reports a historical 10-year annualized real return of about 3% for CAPE (i.e., Shiller PE) values between 20 and 30 (the red line in scatter plot).

[PhysicianOnFIRE]

CM wrote:

Just noticed something (hat tip PoF and ERN): I projected 3.27% annual real prospective return for S&P 500 in post above.

Click to expand...

He is ridiculously thorough.

The Big Question is whether or not higher P/E or CAPE ratios represent a "new normal". Will we see a reversion to the mean or collectively accept a higher P/E for American companies? It's not difficult to look at a graph of the last 50+ years and see a gradual upwards trend.

Part 12, not yet published, will discuss dynamic withdrawal rates based on CAPE. That should be interesting.