So the Canadian Financial Reddit group had an interesting discussion about what a retirement planning philosophy for your Retirement Savings might be. The question was should you try to create a nest egg that you can live off without having to touch it (live off the interest or dividends) or do you create a block of money that you then “draw down” until it is gone? The answer of course is, it depends, however, then the discussion of the 4% draw down rule came into the discussion.
My simple understanding of the model would be that you attempt to create a pile of savings for your retirement $X and if you retire at age 65, you would then withdraw $X * (1/25) every year to live on. This seems to imply that you will only live for 25 years before you exhaust your funds, however, then you enter the much more interesting discussions about the investment mix of your Pile of Savings $X. I will leave that part of the discussion out for now (i.e. how much do you leave in the stock market, how much in Bonds, how much in GICs, etc.,), however, that is the much more spicy part of the discussion.
So if we assume you have $1.5 M saved by age 65 (good on you!), and you plan on drawing down 4% a year, however, your blob of money also gets 1% growth every year, you will have a lot of money left at age 90.
| Savings Amount at 65 | $1,500,000.00 |
| SavingsGrowth assumption after 65 | 1.00% |
| Amount to draw every year | $60,000.00 |
| Age | Amount Left |
| 65 | $1,440,000.00 |
| 66 | $1,394,400.00 |
| 67 | $1,348,344.00 |
| 68 | $1,301,827.44 |
| 69 | $1,254,845.71 |
| 70 | $1,207,394.17 |
| 71 | $1,159,468.11 |
| 72 | $1,111,062.79 |
| 73 | $1,062,173.42 |
| 74 | $1,012,795.16 |
| 75 | $962,923.11 |
| 76 | $912,552.34 |
| 77 | $861,677.86 |
| 78 | $810,294.64 |
| 79 | $758,397.59 |
| 80 | $705,981.56 |
| 81 | $653,041.38 |
| 82 | $599,571.79 |
| 83 | $545,567.51 |
| 84 | $491,023.19 |
| 85 | $435,933.42 |
| 86 | $380,292.75 |
| 87 | $324,095.68 |
| 88 | $267,336.64 |
| 89 | $210,010.00 |
| 90 | $152,110.10 |
That Isn’t Right!
Before everyone jumps on me, this table does not take into consideration:
- Taxes, if this money lives in your RRSP.
- Cost of Living, however, if you make the assumption that your 4% draw down adjusts with the cost of living, then the scenario could reflect all your investment needs to do is outstrip Cost of Living by 1%
- Catastrophic expenditures (sudden illness and such)
- No significant debt load still in play (or other horrific debts)
I now invite my much smarter readers to point out the flaws in this model, to help me improve on this simplistic set of calculations.
Redux
You could read this revised version: The 4% Draw Down Theory (again), which is a little more on the point in terms of arithmetic.
A couple of suggestions:
The sequence of returns matters whenever you have cashflows. Someone who experiences strong negative early returns might see their money last a full decade less than someone who had strong early returns followed by strong negative returns.
Expenses can be very hard to predict in retirement as you noted.
There is no simple model that will work, but this is better than not having *any* idea and just winging it. I think what is most practical is to approach it much the same way as the approach to saving for retirement: have a basic plan, monitor and revise if necessary. If things work well initially, you might increase your spending (or peace of mind). If things start poorly, one might want to take that into consideration and possibly reduce expenses.
The idea of a perfect plan being one in which your very last cheque bounces is cute, but it’s not prudent planning given the sequence of return risk and expense variability that is unknowable in advance.
There are planning software programs out there that model different sequences of returns and spit out a “probability of success” for given portfolio and spending assumptions (normally based on 1000s of simulations), where success can be defined as “not running out of money based on certain spending requirements”. Up until a few years ago, even these were flawed because they used normal distributions of market returns (which predict market shocks happening like once every 30 years or whatever). New models allow for fat tails and are better still.
In the end, they are all just models, but some models strike me as more prudent than others.