A Tail of Two Risks

I’m going to create 2 portfolios, one high risk and one low risk, and run them through the meat grinder called Monte Carlo. I’m using a simple 2 fund portfolio of US Stocks and Total US Bonds, such that both portfolios reside on the efficient frontier. One is an asset allocation of 40/60, a bond heavy allocation the other 70/30 a stock heavy allocation. There are plenty of people that run 70/30 or even worse.

I’m starting with 4M in each portfolio, and I’m taking the first 1M off the top and stuffing that in a Roth IRA in each instance. The Roth serves the purpose of a backup portfolio, which I don’t count in the day to day WR of the other 3M. In other words the Roth stays closed to withdrawal unless needed. A portfolio closed to withdrawal does not suffer SORR. It’s end value will simply be it’s start value plus interest, inflation adjusted. A portfolio open to WR is liable to SORR and inflation. This study will look at a 25 year ride (actually 10,000 25 year rides) and create a distribution of what the future might look like for each portfolio.

These are the efficient frontier data for each portfolio. You can read the risk and return and Sharpe ratio for each.

When I plug the 70/30 portfolio /into Monte Carlo I choose 3M and 110,000 as portfolio size and fixed withdrawal rate inflation adjusted. I choose historic inflation. 3M and 110,000 is a WR of 3.7% in either portfolio. Monte Carlo has a cool feature in that it dissects the internals of various statistics and each portfilio has a very different ride through its 10,000 simulations.

You can read off things like Max Drawdown, the percentage of drawdown the portfolio suffered according to centile, quite informative! in 10% of the cases the portfolio suffered a 95.71% drawdown and an ending balance of only $128,636 left in the bank after 25 years. It tells you safe withdrawal rate and perpetual withdrawal rate for 25 years, an important statistic The WR is tied to the term of withdrawal, so you don’t get to pull fuzzy numbers out of the air and try to apply them to other scenarios like assuming if 4% never fails over 30 years it won’t fail over 50 years. If you want to treat the portfolio as a perpetual source of money you have to reduce the WR from 3.7 to 2.94 or $88,500/yr.

Look at the difference between 10% and 25%! Inflation adjusted end value for the 25% cohort is 1.7M while its 65K for the 10% cohort. Here is the graph

In the 70/30 case at the 10% centile if you start with 3M you’ll end with just under 2M and your chances of success are 9887/10000 for 25 years.

Plugging in 40/60 all else the same

In the 40/60 case max drawdown is -40% (not 95%) but the end balance is 1.4M not 65K much safer. Both portfolios give you 25 years of 110,000 buying power inflation adjusted, but the 40/60 is a kinder gentler ride. The success is 9998/10000 only 2 failures. The graph

the inflation adjusted end balance of the 40/60 is 2.27M or 320K more than the 70/30. The reason of course is the drawdown. It takes a LONG time to recover a 95% drawdown. If it’s 95% down it’s 190% back up. Perpetual withdrawal in the 40/60 is a little bit better at 3.02% or $90,900/yr

So there ya go, a quantitative way to determine your post retirement risk profile through number crunching. Remember you also have that Roth over to the side which has been growing unencumbered by SORR. About the only thing that can derail you is if we actually switch to a bullet based economy where money is no good. No need for side gigs

8 Replies to “A Tail of Two Risks”

  1. That is a really cool feature of this Monte Carlo analysis (the perpetual withdrawal rate especially). Is this something that is available to the general public? (via purchase, etc)? The only Monte Carlo simulation I have is through Personal Capital and it doesn’t have this information available.

    It is interesting that in regards to the perpetual withdrawal rate (something I am more interested in), that the rate is higher only at the 10th percentile mark for the 40/60 split. Every other scenario favors the “riskier” portfolio. I wonder what percentile mark occurs where both portfolios have the same perpetual withdrawal number (obviously it is lower than the 25 percentile mark).

    Which means if you play the odds you come out ahead more than 3 out of 4 times with the riskier portfolio, correct? I guess it all depends on risk tolerance. Would be interesting to see what a 50/50 and 60/40 split would do.

    1. Here is the Monte Carlo I used Here is the Efficient Frontier I used. Yes they are free to use. Since my personal portfolio closely mimics an EF 2 fund I use the 2 fund as a proxy for my portfolio. You can get anal and model your exact portfolio if you like. The perpetual feature is cool but it means you need a BIG (but not huge) portfolio to assure yourself of a 110K/yr inflation adjusted payout.

      I figured the break is somewhere around 18% but that’s just a guess. If your objective is winning play the odds. If your objective is “not loosing” hug the 10% line. The 10% line is the same line (more or less) Personal Capital uses as is the 50% line shared between programs. You can “what if” your butt off testing different conditions. The MC calc allows you to model SORR into the mix as well and change periods of portfolio deflation like 25 and 45 years

      The MC calc has a graph I didn’t publish of “how soon” and “to what extent” portfolios start to fail. The slope and associated data of that line tells the story of failure. You want failure to be far in the future and a small number.

      This brings up a philosophical difference between accumulation mentality and spend down mentality. Nobody in accumulation is subject to SORR because they are living off the W2 and the benefits, paid vaca health care and the like. Since your butt is covered you think little of the risk involved and focus on the return. When you spend down you take back all the risk your employer now covers for you, you’re on your own subject to inflation and SORR as well as health care etc. From a risk perspective spend down is much more risky than accumulation. Having a portfolio that also contains a lot of risk (70/30) makes your risk profile worse. Having a portfolio that is risk averse (40 /60) unloads risk from your life. People talk about their “number” and then proceed to try and get MORE. What the hell if 3M is good 4M is better right. But in spend down getting MORE may be a deadly mistake. It would really suck to run out of money at 88 and you have 4 more years to live and a newly minted NDD diagnosis

  2. Enjoyed reading some of your articles. Keep it up.
    I retired a year ago at 60. Couldn’t take the 2am cases and stress.

    1. Maurice, do you speak the pompatus of love? (see the Joker lyrics) Sorry I was the Joker in college. Tnx for the accolade I try to keep it interesting but my stuff lives out on the perimeter compared to most FIRE blogs. I hear ya about 2 AM cases

  3. Thanks for doing these simulations. I certainly am planning for a perpetual portfolio.

    Thinking about risk management is more interesting to me than growing a number to the stratosphere. I am already thinking and planning around all these issues you are pointing out.

    Keep doing this Gasem. It is very helpful indeed.

    I am currently on a road trip in USA. I love US Denny’s 2.00 pancake breakfasts. Otherwise I eat my own oatmeal. ?

    1. HI MB hope you fare well. Denny’s is a good deal but kind of carby for me but I can always get a plate of eggs and bacon which is more my style. Love the waffles but again too many carbs when you get syrup and all involved.

      Glad you like the concepts. I have one advantage in that I’m already retired so for me it’s not a simulation but living out a reality. The “numbers” I use are not accurate to my actual accounts but the numbers are in the ball park of many FIRE types and might give a target for people to aim for as well as the tool chest to actually create their own scenario. Once you pass 65 it becomes all about risk management and tax management. If you just let it pass to the governments default position, they have big plans for their part of your money.

  4. You capture the continuum of youth to age in the contrast of stretching for gains (stock heavy) and reduction of risk (bond heavy).

    Stretch your neck too far and it winds up on someone else’s chopping block.

    Like the concept of the Roth bucket as a separate insurance policy, although I would think for many the temptation to tap into a little of the Roth each year to avoid entering the next highest bracket would be significant.

    While you have already provided adequate funds without needing the Roth so it can serve it’s intended purpose of redundancy, I could envision a tax optimization strategy calling for small Roth withdrawals.

    Interesting stuff to turn over in my head…

    1. The Roth strategy depends on the size of the Roth. Certainly if you start retirement with 1M in the Roth in 5 years you can start a small withdrawal every once in a while to enhance your life or rejigger your taxes. I talk about it the way I talk about it to enhance the sharp contrast of the utility of that money as a source of risk management. It’s harder to envision that contrast if the Roth is also liable to SOR. If you have a small Roth say 200 -300K don’t tap it and let it grow so it can CYA. Big Roth buy a new car every 5 years or split for the coast in a van.

      No future in loosing your head! I’m, a physician I know these things.

Leave a Reply to Maurice Cancel reply

Your email address will not be published.