2% Withdrawal Rate - Stocks Only?

Investing / Portfolios
35

The ability to be frugal when it’s required is a real superpower imo. There’s only so much you can do by changing asset allocation (little to no marginal benefit at some point), but on the spending side, you have enormous leverage

2 Likes
36

Not sure if I was writing about it already.

I wonder why there seems to be no discussion of portfolio withdrawal strategies based on the selling of fixed amount of fund units. Combined with averaging over few years and flexible spending, you are guaranteed to do well.

1 Like
37

Perhaps because this means a fixed number of years until the portfolio has been depleted? Sure you could plan to use it up in 50 or 60 years, but that means you’d withdraw much less than you could without realistically depleting the portfolio.

Also you’d be increasing the amount by, say, 5-7% per year on average, which should be way more than inflation, so you’d be withdrawing more and more each year.

1 Like
38

Yes. Even with most optimistic estimates, I don’t expect to live past 110 years old. Are people that afraid to think about their own death?

If I knew future market returns, I would have enough money already 20 years ago.

Do you think one should spend more at the beginning of a retirement? This is all very theoretical for me, but I will upload this question into my subconscious thought queue and see what will surface out.

1 Like
40

I’m intrigued by the concept since I haven’t seen it mentioned anywhere until now. How would you handle that?

On a superficial level, my guess as to why it is not more widely discussed is that it seems to me not to allow for consequential planning. You basically roll a dice as to what your allowed expenses for a specific year would be (I don’t see smoothing it over a few years as really getting rid of that problem).

It would work out, to me, if we’re talking mostly discretionary spending, with consequential OASI/Pension base income but I’m not sure how I would handle it if my income came mostly from limited OASI (because early retirement) and stocks/bonds.

1 Like
41

Probably because it makes no sense. People typically need a certain amount of money per month so they might consider withdrawing certain fixed $ amount (not unit amount which varies in price) to meet their needs (needs focussed) or alternatively fixed % of funds which focusses on the amount they are able to withdraw depending on their financial capacity (available resources focussed).

2 Likes
42

That’s not my point. My point is that you either define beforehand you’ll die at, say, 85, then you won’t have any money left after that, or you plan to have enough until turning unrealistically old (e.g. 110), in which case you’d need to amass a much larger fortune than with a traditional 4% p.a. model to cover the same costs.

That’s also not my point. Your idea would lead to much lower withdrawals compared to the 4% p.a. model, which has been tested in a billion different ways. Of course nobody knows future market returns, but we’d at least expect to achieve, on average, a certain profit, why else investing?

No, but I see no reason why my spending should increase or decrease exactly in tandem with the markets. I’d want the ups and downs to be at least greatly flattened, because my costs of living don’t change along with markets, and I also want my withdrawals to increase in tandem with inflation on average.

3 Likes
43

Need to look into these RE funds, which ones do you favor?

1 Like
44

These are hardly anything more than some very theoretical thoughts. It is also similar to bond or cash tent, duration buckets and similar approaches. It is based on the fact that long term stocks have the highest return despite the volatility.

But the concept should look something like this:

Starting configuration: Based on 4% rule, 10% of the portfolio is reserved as cash for the expenses in next 2.5 years = 30 months. Because a part of this cash won’t be used for 1 or 2 years, it can go into fixed term deposit, the rest can go into savings accounts, money markets, etc. The goal of this part of the portfolio is not to lose value in nominal CHF terms.

The rest goes into stocks. Let’s say, the simple way is to have a portfolio of one accumulating all-world ETF. We count them all to be sold over certain amount of years, say, 50.

First month we spend 1/30th part of the cash holdings. In the end of one month we sell 1/600 of the fund units and add the proceeds to the cash holdings. Or we sell 1/100 every 6 months, or something else. This should provide smoothing of available cash, to avoid short term fluctuations.

Nevertheless, the process is to regularly sell the same amount of units and spend the same percentage of cash.

I know what you are thinking: recessions, drawdowns, bear markets. However I suspect that this is the loss aversion speaking. 20% drawdown after a 50% increase is still an excellent investment return.

Completely agree!

Or you should have so much that a 50% decrease in spendings won’t affect your life much.

2 Likes
45

I think for all intents and purposes starting at like 40 year horizon, the calculation is the same for a finite death date or unlimited. We could enter a 60% drawdown any time.

1 Like
47

Well, I already see myself what are the problems with this approach.

According to, for example, calculations by @thepoorswiss , even inflation adjusted 3.5% annual withdrawal rate is super safe for 50 years with a 70-80% of stocks in a portfolio. Add some flexibility, like, consume less in downturns, and it should be totally fine. A global high dividend stocks ETF, like the one from Vanguard IE, distributes 3-3.5% self-adjusting for inflation, permanently, without selling anything. In my scenario, one starts with the withdrawal rate of 2%. That means that at the beginning, when you actually have more use for money, you withdraw much less than you can, and later much more, and leave a lot in the end.

I am convinced that one have to sell more fund units at the beginning than later :grinning:.

48

In my accumuIation phase I used to practice value averaging* (with ETFs) before the stock picking sirens lured me in.**

Your question of a viable withdrawal stategy thus made my neurons (almost certainly prematurely) jump to this conclusion: reverse value averaging,*** i.e. selling varying dollar amounts based on an expected valuation path alongside a drawdown path with some flexibility on your consumption end.

The “classic” x% withdrawal path would be to withdraw a fixed % amount of dollars based on your projected needs every year, e.g. 2%, 3%, 4%, regardless of market valuation.
The reverse value withdrawal/cost averaging path would be to withdraw as much as your projected value withdrawal path allows, i.e. in years of market upswings, you withdraw more, in years of market downswings, you withdraw less.
You deal with the variation in withdrawal amounts by adjusting your lifestyle or by saving up a little more when the market upswings allowed you to withdraw more and vice versa.

The value averaging withdrawal path looks as follows:

It’s setting a (close to) zero target t years out, setting an expected value path of your portfolio for the years approaching t, estimating a yearly withdrawal rate, and — in its simplest form — withdrawing each year your expected withdrawal plus or minus the delta based on your expected valuation path (instead of a fixed dollar amount).****

* Value averaging is discussed in this forum e.g. here:

** Live shot here of the situation by one of my mates with a very early iPhone prototype available at the time:

image
image1101×738 88.2 KB

Look at these beauties … how could I have not deliberately cut myself loose from the mast?

*** Quick intro / refresher for those not familiar with value averaging:

  • if you were to DCA in its simlpest version, and you could set aside (e.g. nominal) $1’000 per year. In 30 years, you’d invest $36’000 (36 x $1000). If you expect a return of (e.g.) 5% p.a. then you expect to end up with about $70’000 in year 30.
    Your schedule would be as follows:

    1. Start of year 1: invest $1000. End of year 1: expected value is $1’050.
    2. Start of year 2: invest $1000. End of year 2: expected value is $2’153.
    3. …
    4. Start of year 15: invest $1000. End of year 15: expected value is $22’657.
    5. …
    6. Start of year 30: invest $1000. End of year 30: expected value is about $70’000.

  • translate this to value averaging.

    • Some math, for those inclined (skip over this if you’re not the math equation person and see the value plan below):
      1. you set a goal of $70k in 30 years
      2. you set up your value path with formulas
      V_t = C * t * (1 + R)^t
      where
      R = (r+g)/2
      V is your value at time t, C is your contribution initial (or average net) investment contribution,
      r is your expected return over period t, g is your expected growth of your contribution after every time period t (essentially a proxy for you making more and wanting to adjust for inflation).
      3. For example, following the DCA example above, you want to end up at $70k over 30 years, you’re willing to increase your contribution by 1% per year and you expect a 5% return every year, you get R = 3%. You put this into the first formula above and solve for C.
    $70'000 = C * 30 * 1.03^{30}
    C = $961
    1. Your value path is
    V_t = 961 * t * 1.03^t

    Sample points on your value path are as folllows:

    1. Start of year 1: invest $961. End of year 1: expected value is $990.
    2. End of year 2: expected value is: $2040.
    3. …
    4. End of year 15: expected value is: $22’465.
    5. …
    6. End of year 30: expected value is: $70’000.

  • Note that you will have invested $33’428 with value averaging (including growing your investment sum every year by 1%) versus having invested $36’000 with DCA and you still end up with the nominal $70’000 after year 30.
    That’s the edge that value averaging claims over DCA.

**** Typically, in textbooks, the working out of an exact procedure is left as an exercise to the reader. :wink:
If there is any interest here, I’m happy to give it a shot at describing the value averaging drawdown plan in terms of formulae and sample numbers. Until then, be well and invest well!

5 Likes
49

Seems more like an acadamic example to me, of which the usefulness can be illustrated by another acadamic example :wink:

You are at year 15 and have saved precisely $22,465, in line with your plan. Next year, you project your portfolio to grow to $24,674. This means that if markets do not move at all throughout the entire year, you need to save $2,209. Unfortunately, markets tank by 50%. Instead of having to save $2,209, you need to save now (in the best case, assuming you only invest after the crash) $13,442 - ouch. Either this amount is huge for you and you need to be prepared to live on the street (bring your family along if you have one…) or it ain’t and you could regularly easily invest much more than you chose to (ie, you regularly through a lot of money “outside the window”).

There was a good article around this on BIG ERNS blog discussing similar “strategies” for the withdrawal phase. If I recall correctly, it had a very similar conclusion: either you are prepared to underspend massively each year or you need to be prepared to live on the street for years or even a decade. Neither sounds very appealing :wink:

3 Likes
50

I would add that you somehow suddenly make the dividend yield of your assets a relevant metric, and not total returns anymore. As you sell fund units, your dividend income decreases, decreasing your total income and increasing your need to sell fund units to finance your expenses, which you won’t because you are selling a fixed amount of them.

With a distributing fund, which would have my preference in the withdrawal phase, you’d effectively reduce your retirement income as time passes even if the markets stay flat and the value of your fund units doesn’t change.

51

Not at all. In most cases I mean the total return. I have just mentioned the dividend yield of a high dividend strategy (no selling) for comparison.

52

Well, I’m confused.

In my understanding of your proposal, you have:

  • a cash buffer, of which you withdraw a set percentage of funds each month (or other amount of time).
  • an investing portfolio, in which you sell “units” each month (or other amount of time) to replenish the cash buffer.

I understood “unit” to mean “amount of shares”. Am I wrong in that?
Where do the dividends go in your scenario, are they reinvested in the investing portfolio or directly added to the cash buffer?

53

If people don’t believe in timing the market in accumulation, why try to time it when spending it down?

If you think money is better invested and you want to invest it as soon as possible instead of waiting for a crash, then similarly, instead of keeping a large cash buffer, you keep everything in stocks (or whatever your asset allocation is) and sell only when you need the cash, say, monthly to keep it simple.

Assuming you are not 100% stocks, you would sell from asset classes to maintain your preferred allocation and/or re-balance periodically.

54

I would really recommend to talk to some financial advisors as well. Maybe there is a solution which includes all of the following

Stocks
Real estate
Fixed income
Annuity
Others
You will also have AHV at some point

I understand DIY rocks but experienced planners do this for a living. They might have some good suggestions.

55

This is to smooth out the amount of money available after the selling of the fund units.

For simplicity, I have assumed from the beginning that these units are accumulating. Otherwise one can, for my original scenario, calculate how many units each distribution worth and sell less when there are dividends.

But well, upon some more thinking, I have decided that it is not s good way.

56

Accumulation is easy, you can outwait anything if you wait long enough. When you live from your portfolio, you need money to be available.