For an audio version of this post, please click on the speaker icon (top left).
Those of you who have been following physician financial blogs for a while have likely encountered some incredible insight from a contributor who goes by the moniker Gasem.
Gasem is, and I quote from his email, a “twice retired FL anesthesiologist [get it? Gas ’em], who built 3 practices including a solo FFS [Fee For Service] practice back in the stone age. He started his last practice because with the nature of Medicine, there’s always a great opportunity you can’t pass up. Now he’s free of employment, managing his wad, and having a Gas!”
I was fortunate enough to receive a notification from Gasem asking if I was interested in publishing a submission of his on my blog.
To paraphrase a quote from Don Corleone, “It was an offer I couldn’t refuse.”
So without further adieu, I leave you in the very capable hands of Gasem who shares with us his take on constructing a portfolio so that we, “don’t die poor:”
Gasem:
I’ve been noodling ways to think about retirement portfolio risk.
There are ways to look at a portfolio, as a vehicle of wealth, but also a portfolio can be a purchase of security.
As a purchase of security inherent in that is the notion of risk.
My perspective on a portfolio is that it’s job is to make sure I don’t die poor, while providing an adequate lifestyle.
William J Bernstein talks about splitting the portfolio in two, one is the liability matching portfolio (LMP) which consists of enough “risk free” assets to cover your basic expenses from retirement to death.
It can be, for example, Social Security, pension, annuity money, a long-term TIPS [Treasure Inflation Protected Securities] ladder, something that is more or less guaranteed to pay you a monthly amount.
The other portion of the portfolio is called the risk portfolio (RP).
The RP allows for some risk to be taken with the hope of growth.
So LMP assures your survival and RP is excess money you can spend on luxury or bequeathment.
The advantage is this portfolio has an extremely high likelihood of success, defined as not dying poor.
Consider “to not die poor” as your target somewhere out in your future.
Your goal is to hit the target the way a sniper hits his/her target.
To hit the target, you have to eliminate, as much as possible, any variability (variance) in the flight of your bullet.
The more you control variance the more likely you are “not going to die poor”.
The reason I phrase the target like this, is you can use a tool to analyze your path to the target.
That tool is the Monte Carlo calculator.
A Monte Carlo calculator takes random inputs from -2 to +2 for two standard deviations of risk or -3 to +3 for 3 standard deviations, and calculates a large number of “probable futures” according to a normal distribution.
It calculates the most likely future and a normal distribution of less likely futures.
If you have a good idea of the lowest likelihood, you can calculate how likely your portfolio is to fail, how long it takes to fail, and if it is going to fail.
By knowing the rate of failure, you get some idea of how safe your portfolio is.
Portfolios have a wide number of variables that effect their variance. In the LMP case the variance is very, very low so you have a very high likelihood of “not dying poor.”
But what if you don’t want to hold a kaboodle of TIPS or carry the overhead of some annuity?
Other things that affect a portfolio is the risk and reward of the assets, the amount of time needed to reach the target, withdrawal rate, the economy, alternative sources of income, asset allocation, inflation, SORR [Sequence of Return Risk] etc.
We all know this litany.
By using the Monte Carlo calculation you can calculate your probability of success decades into the future given this myriad of variables. Note this technique is not backward looking but instead forward looking.
It does, however, rely on a period of history to generate an averaged return and standard deviation on the portfolio as a whole.
For example, if you own the Boglehead 3 fund portfolio, those 3 assets, when combined in the usual 50/30/20 way, gives a whole portfolio with a single predicted return and a single predicted risk.
I ran several simulations for various common asset mixes, various withdrawal rates and various periods to get some idea of how risky a given combination of variables is, measured as % success.
I also varied SORR from 0 to 3. 0 as the normal sequence of return of the model:
1 places the worst SORR as the first return.
2 places the first 2 worst returns in the front
3 places the first 3 worst returns in front.
This is a way therefore to stress the portfolio beyond the normal model and gives some insight into how the other variables like WR [Withdrawal Rate], asset mix, longevity etc. affect % success.
The results are illuminating
60/40: 30 Year Horizon| Portfolio | Withdrawal/yr | SORR | % Success | SORR | % Success | SORR | % Success | SORR | % Success |
|---|---|---|---|---|---|---|---|---|---|
| $1,000,000 | $45,000 | 0 | 95.57 | 1 | 92.3 | 2 | 82.07 | 3 | 71.36 |
| $40,000 | 0 | 98.03 | 1 | 96.55 | 2 | 91.19 | 3 | 83.69 | |
| $35,000 | 0 | 99.25 | 1 | 98.88 | 2 | 96.87 | 3 | 93.36 | |
| $30,000 | 0 | 99.79 | 1 | 99.74 | 2 | 99.28 | 3 | 97.77 | |
| $25,000 | 0 | 99.99 | 1 | 99.95 | 2 | 99.94 | 3 | 99.70 | |
| $20,000 | 0 | 100 | 1 | 99.99 | 2 | 99.99 | 3 | 99.99 |
| Portfolio | Withdrawal/yr | SORR | % Success | SORR | % Success | SORR | % Success | SORR | % Success |
|---|---|---|---|---|---|---|---|---|---|
| $1,000,000 | $45,000 | 0 | 88.92 | 1 | 79.67 | 2 | 59.05 | 3 | 39.96 |
| $40,000 | 0 | 94.62 | 1 | 90.13 | 2 | 74.73 | 3 | 58.06 | |
| $35,000 | 0 | 97.45 | 1 | 95.75 | 2 | 88.03 | 3 | 75.61 | |
| $30,000 | 0 | 99.08 | 1 | 98.55 | 2 | 95.95 | 3 | 89.89 | |
| $25,000 | 0 | 99.79 | 1 | 99.69 | 2 | 99.14 | 3 | 97.03 | |
| $20,000 | 0 | 99.97 | 1 | 99.95 | 2 | 99.97 | 3 | 99.59 |
| Portfolio | Withdrawal/yr | SORR | % Success | SORR | % Success | SORR | % Success | SORR | % Success |
|---|---|---|---|---|---|---|---|---|---|
| $1,000,000 | $45,000 | 0 | 87.16 | 1 | 77.52 | 2 | 53.49 | 3 | 34.61 |
| $40,000 | 0 | 93.31 | 1 | 87.67 | 2 | 70.97 | 3 | 51.12 | |
| $35,000 | 0 | 97.07 | 1 | 94.35 | 2 | 85.72 | 3 | 69.99 | |
| $30,000 | 0 | 98.88 | 1 | 98.26 | 2 | 94.68 | 3 | 86.15 | |
| $25,000 | 0 | 99.67 | 1 | 99.68 | 2 | 98.70 | 3 | 96.18 | |
| $20,000 | 0 | 99.92 | 1 | 99.91 | 2 | 99.82 | 3 | 99.47 |
These 3 tables represent a $1,000,000 portfolio of a 60/40 mix of VTSMX and VBMFX [60% Vanguard total stock market index fund/40% Vanguard total bond index fund].
The CAGR [Compound Annual Growth Rate] of this mix is 8.93% and the SD is 9.2%.
The WR [Withdrawal Rate] is between 2% and 4.5%.
I use 3 time horizons 30, 45, and 50 years.
The % success is a measure of how often you would hit the target.
2% withdrawal rate is essentially the equivalent of the LMP portfolio.
2.5% withdrawal rate is pretty good even at 50 years.
The results tail off from there: 4.5% is consistently a disaster.
Boglehead 3 Fund 30 Year Horizon| Portfolio | Withdrawal/yr | SORR | % Success | SORR | % Success | SORR | % Success | SORR | % Success |
|---|---|---|---|---|---|---|---|---|---|
| $1,000,000 | $45,000 | 0 | 80.65 | 1 | 69.09 | 2 | 45.85 | 3 | 31.30 |
| $40,000 | 0 | 86.54 | 1 | 79.35 | 2 | 58.80 | 3 | 43.82 | |
| $35,000 | 0 | 91.3 | 1 | 87.07 | 2 | 72.15 | 3 | 58.54 | |
| $30,000 | 0 | 95.38 | 1 | 93.51 | 2 | 84.31 | 3 | 72.19 | |
| $25,000 | 0 | 97.85 | 1 | 97.18 | 2 | 92.77 | 3 | 84.33 | |
| $20,000 | 0 | 99.35 | 1 | 99.07 | 2 | 97.93 | 3 | 94.18 |
| Portfolio | Withdrawal/yr | SORR | % Success | SORR | % Success | SORR | % Success | SORR | % Success |
|---|---|---|---|---|---|---|---|---|---|
| $1,000,000 | $45,000 | 0 | 66.10 | 1 | 46.12 | 2 | 17.87 | 3 | 8.14 |
| $40,000 | 0 | 73.51 | 1 | 58.94 | 2 | 30.23 | 3 | 14.06 | |
| $35,000 | 0 | 81.69 | 1 | 71.18 | 2 | 44.91 | 3 | 24.58 | |
| $30,000 | 0 | 87.80 | 1 | 82.61 | 2 | 62.48 | 3 | 37.66 | |
| $25,000 | 0 | 93.47 | 1 | 90.50 | 2 | 79.21 | 3 | 55.79 | |
| $20,000 | 0 | 96.92 | 1 | 96.32 | 2 | 91.22 | 3 | 78.08 |
These are the Boglehead 3 fund portfolio at 30 and 50 years. The assets are 50% VTSMX, 30% VGSTX, and 20% VBMFX [50% Vanguard total stock market index fund/30% Vanguard Star Fund/ 20% Vanguard total bond index fund].
The portfolio CAGR is 7.27% and SD is 12.24%
Notice how the more risky Boglehead 3 fund does compared to the 60/40 portfolios of equivalent longevity.
90/10 30 Year Horizon| Portfolio | Withdrawal/yr | SORR | % Success | SORR | % Success | SORR | % Success | SORR | % Success |
|---|---|---|---|---|---|---|---|---|---|
| $1,000,000 | $45,000 | 0 | 89.98 | 1 | 81.32 | 2 | 60.82 | 3 | 43.59 |
| $40,000 | 0 | 93.53 | 1 | 88.48 | 2 | 72.45 | 3 | 55.64 | |
| $35,000 | 0 | 95.84 | 1 | 93.60 | 2 | 83.09 | 3 | 69.17 | |
| $30,000 | 0 | 98.03 | 1 | 96.84 | 2 | 91.14 | 3 | 79.80 | |
| $25,000 | 0 | 98.68 | 1 | 98.67 | 2 | 96.60 | 3 | 90.86 | |
| $20,000 | 0 | 99.58 | 1 | 98.63 | 2 | 98.93 | 3 | 96.98 |
This is a 90/10 VTSMX VBMFX 30 year portfolio.
It’s CAGR is 9.35% and SD is a whopping 13.1% I included this to show what happens when you have too much risk relative to return.
It’s ok with normal SORR but falls apart dramatically with high SORR and high WR.
I didn’t include a 50 yr portfolio. What’s the point?
VTSMX/VBMFX Portfolio 3% Withdrawal Rate 30 Year Horizon| Asset Mix | SORR 0 | SORR 1 | SORR 2 | SORR 3 | %Return/%Risk |
|---|---|---|---|---|---|
| 90/10 | 97.94 | 96.75 | 91.50 | 80.30 | 9.35/13.01 |
| 80/20 | 98.81 | 98.12 | 94.99 | 86.90 | 9.05/11.53 |
| 70/30 | 99.47 | 99.08 | 97.17 | 93.14 | 8.07/10.09 |
| 60/40 | 99.74 | 99.54 | 99.11 | 97.04 | 8.32/8.69 |
| 50/50 | 99.94 | 99.82 | 99.80 | 99.37 | 7.90/7.35 |
| 40/60 | 99.98 | 99.98 | 99.98 | 99.95 | 7.44/6.09 |
| 30/70 | 100.00 | 100.00 | 100.00 | 100.00 | 6.95/4.95 |
Finally, I included a table that looks at % success for various asset mixes of VTSMX and VBMFX.
For various SORR’s at a constant 3% WR. You can see decreasing risk corresponds to increasing %success but there is some significant effect of SORR on the higher risk mixes.
The calculator was the Monte Carlo calculator from the portfolio visualizer suite.
The simulation model was historical returns which creates an average of many years of returns and SD’s
There are other models you can choose from but I generally like this one.
Bottom line: Lower WR, lower portfolio risk, lower longevity = higher % success = hit the target.
So now you have a quantitative guide on how to plan your portfolio to “not die poor”. Of course planning anything 50 years into the future with any precision is fraught with presumption.
Superpower Take-home points:
- Sequence of Return Risk can have a MAJOR impact on the ability of your portfolio to survive and not allow you to “die poor.”
- Gasem has given great examples of how various asset allocations behave under various early sequence of risk scenarios (including the “perfect storm”)
- Depending on your individual risk profile you can find a % success rate that YOU are comfortable with and design an asset allocation similar to that model.
Again I would like to thank Gasem for his outstanding analysis. I can only imagine the amount of time it took to formulate this and put it into a concise form.
Note:
If you are in search of financial help, please consider enlisting the service of any of the sponsors of this blog who I feel are part of the “good guys and gals of finance.”
Even a steadfast DIY’er can sometimes gain benefit from the occasional professional input.
-Xrayvsn
NOTE: The website XRAYVSN contains affiliate links and thus receives compensation whenever a purchase through these links is made (at no further cost to you). As an Amazon Associate I earn from qualifying purchases. Although these proceeds help keep this site going they do not have any bearing on the reviews of any products I endorse which are from my own honest experiences. Thank you- XRAYVSN
Gasem, great discussion. (I’ve enjoyed your comments in many different blogs…I too was a ham radio operator “back in the day”).
I consider myself very fortunate that Gasem gave me the opportunity to publish one of his works. I too was impressed with his commentary on other blogs. I actually have two more of his posts scheduled to publish in the upcoming months
Thanks PD your comments are kind. You should get back into the ham biz the technology today is amazing
It’s funny…I still sometimes “think in CW”. Flying a lot knowing code is extremely useful. Does anyone do CW anymore?
Enjoyed this very much. People focus a lot on asset allocation, but there isn’t enough talk about sequence of return risk. This was a great illustration of how important it can be.
The hard truth is there is always going to be some uncertainty when planning for retirement. The key is to be flexible so you can respond to changes in the market, or in your personal situation as the retirement years tick past.
-Ray
You hit the nail on the head. Flexibility is key. There is no law that says you have to take out the same percentage each year. In a down market, especially if early in retirement, you can tighten your belt and withdraw less. That greatly increases your portfolio survivability
Hi Ray Flexibility is important but the decisions we make during accumulation is magnified to great consequence in the portfolio in the distant future. A 50 year horizon is far more dangerous and therefore requires more robust targeting Thanks for your insight
That is why I always build back ups for my back up plans. You might get lucky but also might not. I agree with Gasem’s tenet of not dying poor. Anyone who understands money knows that any amount can be depleted if one tries hard enough. It only takes a few unanticipated set backs to blow things up.
Being careful with money and planning to the end is a great idea. I see many of us want to continue working part time even when we seem to have enough.
Dying poor is definitely not my end game either. I agree that there are so many unexpected events life can throw at you and obviously you cannot anticipate them all. Most of the readers who follow sites like this are much more savvy (as well as typically more conservative) than the general population and likely will die with a net worth even higher than they started out with. But that’s okay because peace of mind having a big buffer is also in it’s own right something that brings happiness.
Always provocative MB. I think some kind of side gig is almost necessary if you RE. It provides an incredible source of non correlated asset diversity. I look at retirement in epochs and early retirement is a different epoch than late retirement. In late retirement if you did it right your ability to out live your mistakes is reduced to the point where you can afford to let the gig go.
Provocative and on point as always, Gasem. I imagine few folks who retire early rely solely on their winter chestnut supply to carry them the next 40-50 years. Instead they allow that stockpile to grow and compound while working just enough to cover expenses and keep from dipping into their retirement. Since they front-load savings anyway, they start with a tailwind. By adopting a glide path they work less, cover expenses, and weather the storms flexibly as they come. While it could be characterized as one more year syndrome , I see it as the ability to take on work… Read more »
Crispy, that’s been my goal….to be at the point where I don’t *have* to make money, but that I have the ability to say “no” and turn down things I don’t enjoy or don’t meet my personal “pain:reward” ratio.
SORR scares the $&*t out of me. I remember stagflation all too well.
My man CD! I wrote this as an objective prediction of failure. It represents one tail of the bell curve. There is the other tail where you wind up an eleventy bazillionaire that is not addressed. I have two more articles coming out on XRAYVSN this summer which will deal with some strategy on how to position yourself to preserve yourself in the storm. I’m considering a third article as well compromising a 4 horseman series. These strategies require a little long range planning to allow differing risk profiles and compounding to provide insurance to the portfolio to get you… Read more »
Gasem, I’m looking forward to reading them!
…Continued discussion with Gasem from one of my recent posts on conflicted financial advice… Man, we are not that far off. I am not a total believer in the boglehead 3 fund portfolio, but in index fund investing which your two fund portfolio still espouses. I think the three fund portfolio is a good example of how simple it can be, but yours is even simpler with only two funds. I just don’t like making it any more complicated than it needs to be. I like the way you looked at this and found it fascinating. You still have a… Read more »
[…] math (4% x25) you got it covered. You don’t. I wrote a guest post on this reality using some quantitative methodology looking at this. Pick your […]