FYI: Retiring without sufficient assets to maintain a minimally acceptable lifestyle (which each person defines in their unique way) is an unthinkable outcome. That’s why, when investors are planning for retirement, the most important question is usually something like: How much can I plan on withdrawing from my portfolio without having a significant chance of outliving my savings?
The answer is generally expressed in terms of what is referred to as a safe withdrawal rate—the percentage of the portfolio you can withdraw the first year, with future withdrawals adjusted for inflation.
Regards,
Ted
http://www.etf.com/sections/index-investor-corner/swedroe-retirements-routes-failure?nopaging=1
Comments
Hooray for Larry Swedroe! In this current article Swedroe identifies Monte Carlo simulators as an important tool when making a retirement decision. He joins many financial advisors who also exploit this useful tool when making that life changing decision. I too have recommended application of Monte Carlo simulators since the early 1990s.
Swedroe emphasizes that a projected failure rate from these Monte Carlo estimates is not sufficient as a standalone output. He argues that the time of potential portfolio exhaustion failure during the retirement lifecycle is also critical. I completely agree.
The code that I frequently recommend, from Portfolio Visualizer, does provide that information in a graphic format. Once again, here is a Link to that excellent website tool:
https://www.portfoliovisualizer.com/monte-carlo-simulation
When I first became interested in the retirement riddle, Monte Carlo calculators were not readily available. So I built my own copy. I too recognized that the time of failure was a critical output. So on my version of a Monte Carlo code I included a user option to reduce withdrawal rate percentage by a user input if the portfolio suffered negative returns for 3 consecutive years.
An input of a 10% withdrawal rate reduction after 3 down markets lowered the portfolio failure rate substantially. These Monte Carlo studies encouraged my early retirement. Other approaches to protect against a portfolio failure exist.
I encourage you guys to visit the Portfolio Visualizer website and to consider using their Monte Carlo code. It's a terrific tool; give it a try.
Best Wishes.
A double hooray is warranted. To complementt the Larry Swedroe endorsement of the Monte Carlo simulator, Professor Snowball has added his informed investment weight. Recently he posted a reply that mentioned a Monte Carlo code that he favored. I'm sure it is a worthy tool. Here is the internal Link to his submittal:
http://www.mutualfundobserver.com/discuss/discussion/31674/suggested-reading-for-a-teenage-investor-next-step
All this is useful stuff when making retirement decisions in a random investment returns environment.
Individual investors who resist this fairly modern tool, which only became available in the1990s, are missing out on a methodology that is specifically designed to focus on uncertain investment outcomes that's generate uncertain end wealth and portfolio survival odds. These outputs give some insights when exploring the retirement quagmire.
I am no longer the Monte Carlo Lone Ranger on this site with both the Swedroe and Snowball endorsements of this methodology. Persistence pays off! With time, Monte Carlo will become as welcomed and as common as our morning coffee.
My Monte Carlo recommendations are certainly not exhaustive. Other more sophisticated Monte Carlo codes are accessible on the Internet. Typically they require more specific inputs because they assess taxable circumstances and tax implications on end wealth. I have not recommended them because of my self imposed simplicity ground rule.
Best Wishes
When I first started work on our retirement plan, Monte Carlo calculators were not readily available. So I built my own projection engine, using a spreadsheet which allowed for multiple variable inputs, especially tailored for worst-case simulation. It also included an option to reduce withdrawal rate percentage if the portfolio suffered negative returns.
I am a firm believer in Murphy's Law: "If anything can go wrong, it will go wrong, and at the worst possible time". True to form, 2008 came along soon after our retirement. Thanks to the planning, we survived very nicely, and have completely recovered.
More sophisticated time related portfolio failure rates are not the problem with the codes. All Monte Carlo codes generate that data as a natural output of their calculation procedure. The problem is that many code designers elect not to include that data in their output summaries.
As an example, I like both the MoneyChimp and the Portfolio Visualizer Monte Carlo codes, mostly because of their easy accessibility and easy input formats. The Portfolio Visualizer code includes a summary graph that depicts portfolio survival as a function of study time; the MoneyChimp code does not. You can choose depending on your needs.
Either code can be a useful part of a retirement planning toolkit. Since these are Monte Carlo tools, precise answers are not possible. Given identical inputs, each code will generate a slightly different portfolio survival rate. Hell, running the same code twice will produce slightly differing output survival odds. That"s the nature of uncertain, unknowable future events.
In terms of the sensitivity of portfolio failure times, the problem is not the code being used. The construction details of the portfolio being examined is the culprit. Portfolios that have a high returns standard deviation are likely candidates for a bunch of some early failures. Being flexible in drawdown is also useful to avoid that outcome.
Thanks for the question. I hope this helps.
Best Wishes
"The problem is that many code designers elect not to include that data in their output summaries."
If the data is not included in the output then the output is certainly not equally sophisticated, nor more sophisticated. That leaves less sophisticated, doesn't it?
Well time hasn't changed much for us. You're the same Old Joe of FundAlarm days. That's not only good for you, it's good for me.
It's amazing how your earlier submittal mirrored my initial entry. I said: " When I first became interested in the retirement riddle, Monte Carlo calculators were not readily available. So I built my own copy." You later said:"When I first started work on our retirement plan, Monte Carlo calculators were not readily available. So I built my own projection engine....". That's an unexpected similarity in words, but not in methodology.
The question has always been how the year-to year-projected returns were selected in your spreadsheet. I used a random generator command to select my annual returns. That's the heart of all current Monte Carlo codes. Random return selections is not so easy a task. Avoiding bias in the selection process is a real challenge.
But let me end this exchange with a famous humorous story. I'm sure you're familiar with the scientist who was awarded a Nobel prize when he discovered an error in the Random Number tables. Anyway, that's my attempt at humor!
I'm pleased that you recovered from the 2008 market debacle. I retired in the early 1990s and our portfolio has fortunately never been in any dangerous survival zone. I'm sure much of that success is pure luck.
I wish you and your family well and much pure luck too.
Best Regards
@MJG: Well said, and thank you. I wish the same for you and yours.
With respect to "mirrored my initial entry" I was struck by our parallel experiences and deliberately utilized (stole; reused) your construction to emphasize the connection.
Cheers! (to steal a salutation from David)
OJ
Nick de Peyster
Undervalued Stocks
I couldn't agree more! Garbage in, garbage out is always an accurate cautionary summary.
And since future investment returns and inflation rate changes are unknowable unknowns, the sensitivity of retirement success to these unknowns should be explored. Monte Carlo codes that are fast running with simple input features make that exploration easy. Simulations like these alert the potential retiree to the risks of that decision and the survival robustness of his portfolio.
For example, Monte Carlo simulations can be used to guide the construction of a portfolio by demonstrating the survival sensitivity to the portfolio's standard deviation. For any given projected returns, reducing the portfolio's standard deviation increases its survival prospects. Running a what-if case will put odds to that generic truism.
I also fully agree with your statement that Monte Carlo is only a tool that plays a small part in the retirement decision. It provides some odds estimates. When gambling, the player who knows the odds will likely do better than a player who is ignorant of those odds.
A successful retirement is much more dependent on the emotional stability and the flexibility of the retiree over any calculation, Monte Carlo or otherwise.
Monte Carlo does not guarantee happiness in retirement. But it can enhance a retiree's comfort feelings if a bunch of what-if Monte Carlo simulations all produced portfolio survival estimates in excess of 90%. Even more comfort if those simulations yielded estimates above 95%. We can work a portfolio to increase its survival likelihood over an extended timeframe.
Thank you for your contribution. I enjoyed the opportunity it provided.
Best Wishes