A widely accepted rule of thumb when it comes to retirement spending is the ‘4% rule’, which says if you spend 4% of your assets in retirement per year, you have a good chance of making your money last 30 years.

I recently came across a set of articles by Wade Pfau from the Retirement Researcher Blog which I thought were interesting. The author shifts the focus from spending in retirement to savings during the working years. Although the study uses historical data from the US market under certain assumptions, there are some lessons we can take away from the articles.

The author uses historical data and evaluates what would have happened to the retirees in any given rolling period up to 1980. This is a similar idea to what is presented in Unveiling the Retirement Myth which also advocates looking at historical market returns to plan for retirement. No guesses are made as to how the markets will perform in the future.

The study assumes that at retirement, the retiree wants to have a retirement income of 50% of their final income from their savings. It also assumes the retiree has saved for 30 years, needs retirement income for 30 years, and maintains an asset allocation of 60% stocks and 40% bonds.

For example, it someone earned $100,000 every year for 30 working years, and wanted $50,000 per year for 30 years in retirement (assume no inflation), what would it take to achieve this goal?

All the figures and tables shown in this article are from the original study referenced at the end of this post.

**PROBLEMS WITH THE 4% RULE**

The 4% rule may not be optimal as the actual withdrawal rates which could have sustained a 30 year retirement have shown a large variation as shown in the figure below. The range goes from about 4% to around 10% between retirees who retired between 1870 and 1980. From the chart below, a person retiring around 1920 could have safely withdrawn 10% of their assets every year, whereas someone retiring in mid 1960s could have only withdrawn around 4%.

In addition, this rule hasn’t worked in every country. Only 4 out of 17 developed countries have been able to sustain the 4% rule under ideal assumptions, with Canada being one of these 4. Details can be found in the article An International Perspective on Safe Withdrawal Rates from Retirement Savings: The Demise of the 4 Percent Rule?** **

**JUST SAVE ‘ENOUGH’ AND THE REST WILL TAKE CARE OF ITSELF?**

Based on the rolling historical data, if retirees were to follow the 4% rule, the amount of money they would have had to have saved as a percent of their salary also varies a lot. At times, that number would not have been practical in my opinion. As shown in the chart below by the blue line, someone retiring around 1920 would have had to save over 35% of their income every year for 30 years.

However, by studying the historical data, the author concludes that in the **worst** case scenario, a retiree would have to save **16.62%** of their salary for 30 years. Regardless of how much wealth this had accumulated, it was enough to support a 50% salary rate for a 30 year retirement period without running out of money. The black line in the chart below shows the various savings rates that would have been required for someone retiring in any given year.

Of course, it’s not possible to know how what the acceptable savings rate will be for you in advance. The savings rate however was in a range from about 10% (best case) to about 17% (worst case). This is a much tighter and manageable range. Theoretically, if we save 17% of our desired final salary every year, we should be able to support the 50% income need.

**KEEPING COSTS LOW**

The data above assumes there are no taxes or fees. If a fee of 1% is added, the worst-case savings rate jumps from 16.62% to 22.15%, a fairly substantial increase. So, keeping costs low and staying away from expensive investment products becomes very important.

**IMPORTANCE OF STARTING EARLY**

As is often advocated, starting to save early makes a big difference on how much one needs to save to reach a given target. In the table below, the required worse case savings rate goes from 16.62% of income to 35.91% if one delays saving for 10 years. The numbers are even scarier if you want to replace 70% of your salary instead of just 50%.

**MID-CAREER PROGRESS TRACKING**

Since most of us probably did not save a lot of money at a younger age, the Getting on Track for Retirement provides a basic guideline to evaluate when one could retire based on their age and how much they can save.

**SUMMARY**

How much money one ends up with after a lifetime of savings can vary a lot and has a luck factor based on how the markets happened to perform over the period you were saving. Instead of focusing on a fixed ending dollar value before retirement, focus on an adequate savings rate to ensure the income you want during retirement.

Although all the above numbers are based on very specific assumptions and the US markets, we can take away some lessons from it. Namely,

1) Saving a reasonable amount of your salary. The worst case (highest required) savings rate was 16.62% in the study mentioned in this article in order to save enough to replace 50% of pre-retirement income for 30 years. I would suggest trying to save probably 20-25% as soon as one is able to (if not more) to account for taxes, fees, and adding a margin of safety.

2) Start saving as early as possible. A 10-year delay in starting to save will require you to save twice as much every year.

3) Keep investments cost low. Use low cost ETFs over mutual funds if you have a choice.

4) Don’t use the 4% rule as a static rule as it may or may not work. You may get away with a greater than 4% withdrawal rate, but if you are withdrawing more than 10% of your retirement assets every year, your funds are unlikely to last too long.

**WHERE IS THE TECHNOLOGY?**

It would be great to have (free?), easy-to-use software available to allow users to plan their retirement using historical simulations. Ideally, the software would allow users to

- Enter their asset allocation
- Choose multiple global stock and bond indices to represent the portfolio design, not just use U.S data
- Incorporate assets located in tax deferred, tax exempt, and taxable accounts
- Calculate required savings rate required to reach the desired goal based on average and worst case scenarios
- Provide a best-guess estimate on the current sustainable withdrawal rate based on a some market valuation metrics
- Factor in annuities

Perhaps someday soon we will have something like this available.

**REFERENCES**