Visualizing the 4% Rule, Trinity Study and Safe Withdrawal Rates

Posted In: Financial Independence | Money

Instructions for using the calculator:

This calculator is designed to let you learn as you play with it. Tweaking inputs and assumptions and hovering and clicking on results will help you to really gain a feel for how withdrawal rates and market returns affect your chance of retirement success (i.e. making it through without running out of money).

Inputs You Can Adjust:

  • Spending and initial balance – This will affect your withdrawal rate.  The withdrawal rate is really the only thing that is important (doubling spending and retirement savings will still yield the same success rate).
  • Asset allocation – Raise or lower your risk tolerance by holding more or less stock vs bonds
  • Adjust retirement length – This affects the number of historical cycles that are used in the simulation, but also increases risk of failure.
  • Add tax rates and investment fees – these will put a drag (i.e. lower) market returns and lower success rates

Options for Visualization:

  • Display all cycles – this is the mess of spaghetti like curves that show all historical cycle simulations
  • Display percentiles – this aggregates the simulations into percentiles to show most likely outcomes
  • Hover/Click on legend years – this will allow you to highlight a single historical cycle (you can also use the arrow keys to step through historical cycles)
  • Bottom graph can show either the sequence of returns (with average returns in 5 year periods) for a single historical cycle or distributions of returns in our historical data (1871 to 2016) and a single historical cycle.  You can choose to look at returns for stocks, bonds or your specific asset allocation.
  • The graph on the right shows a histogram of the ending balance of each historical cycle and color codes them to show percentiles.

What is the 4% Rule?

The 4% rule is a “rule of thumb” relating to safe retirement withdrawals.  It states that if 4% of your retirement savings can cover one years worth of retirement spending (an alternative way to phrase it is if you have saved up 25 times your annual retirement spending), you have a high likelihood of having enough money to last a 30+ year retirement. A key point is that the probabilities shown here are just historical frequencies and not a guarantee of the future. However, if your plan has a high success rate (95+%) in these simulations, this implies that retirement plan should be okay unless future returns are on par with some of the worst in history.

The overall goal of this rule and analysis is identifying a “safe withdrawal rate” or SWR for retirement.  A withdrawal rate is the percentage of your money that you withdraw from your retirement savings each year.  If you’ve saved up $1 million and withdraw $100,000 each year, that is a 10% withdrawal rate.

The “safe” part of the withdrawal rate relates to the fact that if your investments generally grow by more than your annual spending, then your retirement savings should last over the length of your retirement.  But average returns do not tell the whole story as the sequence of returns also plays a very important role, as will be discussed later.

One way to test this is through a backtesting simulation which forms the basis for the “Trinity Study”.

What is the Trinity Study?

The “Trinity Study” is a paper and analysis of this topic entitled “Retirement Spending: Choosing a Sustainable Withdrawal Rate,” by Philip L. Cooley, Carl M. Hubbard, and Daniel T. Walz, three professors at Trinity University. This study is a backtesting simulation that uses historical data to see if a retirement plan (i.e. a withdrawal rate) would have survived under past economic conditions.  The approach is to take a “historical cycle”, i.e. a series of years from the past and test your retirement plan and see if it runs out of money (“fails”) or not (“survives”).

How do you test withdrawal rate?

Given modern equity and bond market data only stretches back about 150 years, there is some, but not a huge amount of data to use in this simulation.  One example of a 30 year historical cycle would be 1900 to 1930, and another is 1970 to 2000.  The Trinity study and this calculator tests withdrawal rates against all historical periods from 1871 until the present (e.g. 1871 to 1901, 1872 to 1902, 1873 to 1903, . . . . 1986 to 2016).  Then across this 115 different historical cycles, it determines how many of these survived and how many failed.

The thinking is that if your retirement plan can survive periods that include recessions, depressions, world wars, and periods of high inflation, then perhaps it can survive the next 30-50 years.

The 4% rule that comes out of these studies basically states that a 4% withdrawal rate (e.g. $40,000 annual spending on a $1,000,000 retirement portfolio) will survive the vast majority of historical cycles (~96%).  If you raise your withdrawal rate, the rate of failure increases, while if you lower your withdrawal rate, your rate of failure decreases.

The goal of this tool is to help you understand the mechanics of the a historical cycle simulation like was used in the Trinity Study and how the 4% rule came to be. This understanding can help you better plan for retirement with the uncertainty that goes along with planning 30+ years into the future. If you want to also see how longevity and life expectancy play a role in retirement planning, you can take a look at the Rich, Broke and Dead calculator.

This post and tool is a work in progress. I have a number of ideas that I will implement and add to it to help improve the visualization and clarity of these concepts.

Data source and Tools Historical Stock/Bond and Inflation data comes from Prof. Robert Shiller. Javascript is used to create the interactive calculator tool and the create the code in the simulations to test each historical cycle and aggregate the results, and graphed using open-source, javascript graphing library.

4% rule trinity study

No Comments »

Leave a Reply

Your email address will not be published. Required fields are marked *