Hi Guys,
The retirement decision is one of the most crucial in our lifetime. It is never easy because it involves the uncertainties of forecasting the future. That's a tough challenging task. Ben Carlson examines the retirement issues in a recent article. Here is a Link to that article:
http://awealthofcommonsense.com/2017/01/the-many-uncertainties-involved-with-retirement-planning/The article tabulates a ton of uncertainties. That gives me an opportunity to once again recommend a tool that was developed specifically to explore uncertainties. Of course, that tool is Monte Carlo simulations. Again I'll recommend that you guys give the Portfolio Visualizer website some serious consideration because it offers an excellent Monte Carlo simulator. Here is the Link to the website and to the Monte Carlo code:
https://www.portfoliovisualizer.com/monte-carlo-simulationI encourage you all to give it a chance to make your retirement decision just a little easier and better informed. You can run a large number of simulations in a short time that will give you a feel of the fragility of your retirement funds to various spend and market assumptions. Good luck when exploring your various what-if future scenarios.
Best Regards.
Comments
As for assumptions, I would encourage folks to use very conservative expectations for investment returns, know that medical expenses will become a bigger part of your cash flow needs the older you get, and be bluntly realistic in your spending projections meaning that you are willing to reduce spending following years when investment returns are poor.
MJG makes good points about looking at what the variables might do to projections, but if potential cash flow won't sustain the plan even with conservative assumptions, rest assured it will only be worse with a major market downturn. I still contend that most folks should have a goal to pay off the mortgage before they retire. That alone can mean the difference between squeaking by and having a full retirement experience.
Wise words. Many thanks for saying them. I have been retired for over twenty years and practice much of what you recommended. It works!
Best Wishes.
Based on your comment I assume you tried a portfolio that included an Emerging Markets component. As you discovered, the database for that asset class is limited to a starting year of 1995. Other asset class databases go back in time much further, like to the early 1970s..
Workaround options exist, like imputing your own return estimates. The tool, like all tools, has limits. You can extract maximum benefit from it by experimentally testing its various options, and there are plenty of options provided.
I hope you have as much fun playing with this Monte Carlo code as I do. Regardless, good luck.
Best Wishes.
I have played with it many times and for a very long time. Prefer some of the others, I think.
Sorry that you're experiencing difficulties using the Portfolio Visualizer code. I don't have any problems. Perhaps that's because I input my own data sets for each asset class in my portfolio.
So I input the expected average return and standard deviation for each class as well as its percentage in the portfolio. I secure that needed data from sources like Morningstar (many other sources are easily available)'as a baseline and modify them to test my portfolio's sensitivity to uncertainties in the future market environment. In the referenced code, the returns option that I often use is called "Forecasted Returns". Good luck if you persist in your effort.
What Monte Carlo,code do you favor?
Best Wishes.
I ran different Return (and other) variables, sure.
Always get well above 95% no matter what I do.
The Vanguard site is less optimistic, but (and probably because) way less granular, at least to an outsider.
Well congratulations on getting a persistent 95% plus success likelihood running different scenarios. That should be a confidence builder.
Over 2 decades ago, I used the 95% value as my threshold retirement target. I believe the statistical clan call that a statistically significant finding given the huge number of cases explored in these Monte Carlo simulations. It certainly increased my confidence level when I finally pulled the plug. It has worked out well for me and my wife.
You might try running a numerical experiment that checks the success probability drop-off as a function of increasing annual spending. That number is one way to quantify your safety margin.
Most financial institutions did not offer Monte Carlo codes when I made my decision. Much progress has been made in this field. Good for all of us.
Best Wishes.