SWR

[ToFI]

If I understand correctly, the 4% SWR is based on the historical return of S&P of 9.77% over the past 100 years or so. What if I can achieve double of that return in the long term? Let's say it's 20% annualized, does it mean my SWR can be 8%?

[jacob]

Disclaimer: SWR seems to be used in two completely different ways depending on context. 1) A withdrawal rate in general that's not necessarily safe, specifically Expenses/Invested Net---that should correctly be named WR but it's still called SWR; 2) The withdrawal rate that's predicted to not run out during one's lifetime---the safe "withdrawal rate"---the actual SWR.

Note that SWR is a statistical expectation. It's not a guarantee!!!!!!!!!!!!!!!!!!!

The 4% SWR is based on some historical US equity returns for a portfolio that would be deemed to be safe for _30 years_. This is called the Trinity study. It is based on a statistical sample of the 20th century --- a period in which the US equity market grew more than the markets of most other countries. Counting on a 4% SWR, you're submitting that the worst 30 year period of the 21st century is no worse than the worst 30 year period of the US equity market in the 20th century.

That's somewhat a leap of faith. Supreme trust in statistics, indeed!

The 3% SWR commonly used here (ERE) is based on a 3% real growth rate of the economy of the past few thousand years. It's presumed that a diligent investor will be able to find that return by investing in "something" but that finding that something will require some continuous effort over the several decades the portfolio must last. This might not be equity. Popular investing in stocks is a VERY new thing seen over historic times. It's only in the past some 50 years that the US public got widely involved in it. This also explains why it could rise to much.

This also means that there are few left to buy in. If everybody is invested, it's hard to grow the market faster than GDP. Ponder this until it sinks in!

Statistically, if it was 20% annualized, that does not trivially translate into an 8% SWR. It would depend on the volatility. The point is not to run out.

What you really meant to ask is, if you can achieve higher returns than 9.77% with an equal or lower volatility than the historical scenario, can your SWR be higher. Yes! But can you? It also means that if you can achieve a lower volatility, then you don't need an average 9.77% to get the same SWR.

Keep in mind that SWR is not more "safe" than it's "safe" to cross the freeway without looking x% of the time. You're still vastly better of by looking before you cross.

I find this blind faith in the statistics of the SWR disturbing.

[steveo73]

I find this blind faith in the statistics of the SWR disturbing.

My assumption is that you have relied to a degree on the SWR. Is my assumption correct ?

My take is that a 4% SWR or 25 times expenses is a good target to aim at but you have to tailor this to your own situation. For instance I am not considering my house as an asset for retirement although it might end up being used as an asset to draw down on. I am also more inclined to look at dividend income being greater than my expenses as being my retirement end goal. I suppose this is because I believe that if I can survive basically off dividend income my money won't run out. In stating that I have my goal as 50 times projected minimum realistic expenses. I figure at this point I can splurge a bit if required or cut back as necessary.

[jacob]

I use the SWR studies to get an idea of the ballpark numbers. This way I won't get carried away with hubris thinking I can get away with a WR of 10% because I'm such a super-investor. On the other hand, I figure that's there's a 50/50 chance that the worst of the 21st century will be worse than the worst of the 20th century, so I don't trust my portfolio to autopilot either. Note that S&P recently came out with a revised SWR study that pointed to a number much lower than 4%.

The number I do trust on non-autopilot is 3% but you'd have to pay attention, maybe change asset classes once or twice in your life. I make a big point of this in the book. Some reviewers seem to have missed this point and slam me for not telling them which stocks to buy. Blargh!

On a more general note, the design goal of ERE is to diversify sources of value and not just rely on a single point, the portfolio, to provide. As such the worst thing that could happen would be a kind of semi-retirement and those who are willing to accept that could use higher WRs already.

My current WR is 1.8%.

[workathome]

When you use WR, with something like a Permanent Portfolio or owning a home, would you only count the invested positions against your WR or also include your assets not producing cash flow (Gold, Cash, Real Estate)?

For example, say someone living off $30,000/year with a $500,000 net worth. However, only $250,000 is actually invested in Bonds and Equities. Is that a 6% WR or a 12% WR?

[George the original one]

> Is that a 6% WR or a 12% WR?

Withdrawl rate should always be considered across all your assets. So 6%.

Note that personal housing often isn't counted, though I think it should be. I'm planning on divesting my house around age 80 and switching to rental (probably elder care by then <ahem>).

I suspect that the illiquid nature of housing and the difficulty of substantiating an accurate market price is a prime reason people don't count it.

[jacob]

Once you start including housing, you really have to understand what and why you're doing the calculations you're doing.

Lets assume for a moment that the Trinity mix of equity (and bonds?) will hold "safe" for 30 years at a 4% WR.

Now, assume expenses (food, rent, ...) are $10k/year and the portfolio which is all invested in the Trinity portfolio is $250k. Then you're safe (in the Trinity sense).

Now assume you buy a house for $50k leaving you with $200k for investments. Now by the study, you could withdraw 8k per year and still expect it to hold. You can not withdraw 10k because your housevalue does not grow like stocks and bonds did historically.

By owning your house, it provides some value, namely the rent you used to pay. Therefore you'll be fine withdrawing only 8k because you get 2k in "housing benefits".

This means your 4% rule is predicated on how much you have invested in securities. You can not include your house, your car, or the gold burried in the backyard.

Of course, realizing this, there is of course a separate study that would calculate some other % number under the assumption that 20% of your net is invested in your residence. However, that number is almost surely not 4% (it's probably lower since RE grow slower than equity in the 20th century).

Conclusion: Real life tends to be more complicated than academic studies.

PS: You can see how the accounting gets complicated real fast once you own assets.
For instance, if I own my house clear and fair, do I count the opportunity cost of not having this money invested as part of my expenses. Say I spend 10k/year and live in a 200k/house. Should I calculate my expenses as 10k+200k*4% = 18k or just stick with the 10k. Clearly not accounting for the house makes a big difference in the accounting, because if I was paying rent instead, my expenses probably would be in the ballpark of 18k.

[Tyler9000]

With the Permanent Portfolio, you absolutely count gold and cash in your assets. They're important parts of the package. Just because the Trinity Study ignored whole asset classes doesn't mean you need to. In fact, the 25% cash is especially helpful to your WR in retirement as it allows you to ride out market dips for a few years without selling assets.

However, for WR purposes your primary house is generally considered a consumption item, not an investment. You can't sell 3-4% of your house every year to cover expenses.

[jacob]

@Tyler9000 - If you're using the PP, the "magic number" is no longer 4% though. It's in the ballpark of 4%, surely. Maybe higher, maybe lower.

However, logically you can't take the 4% which is based on portfolio A and then apply the same number to portfolio B.

Which leads me back to my take-away. The "4% rule" is not a rule. It's a statistical expectation.

[Tyler9000]

@Jacob - absolutely agree. SWR discussions get you in the ballpark of where you need to be, but following a "safe" withdrawal rate blindly implies a huge list of assumptions that one would do well to reevaluate periodically. Life has a way of throwing you curveballs that the no spreadsheet can account for.

(Written from my couch while recovering from surgery)

[workathome]

Off-topic:

What factors would predicate subnormal historical returns for PP similar to Demographic limit in the stock market?

Something like markets underperforming while interest rates go up, and Gold only matching inflation while inflation slowly eats at cash?

I don't understand why the PP performs at 9% historically and not only 4%ish. Cash value erodes, bonds and gold only seem to match inflation, while equities are the only true growth opportunity.

[chicago81]

I've determined that my "safe" withdrawal rate is going to be when dividends from my equities holdings reach somewhere around 110% to 125% of my regular living expenses (with the assumptions that my primary residence is owned outright and paid off, and I have zero debt.) This is probably much more conservative than it needs to be. If I put a "number" to it, it would probably be much less than 4%SWR... probably close to 3%. That's just what I feel comfortable with. I don't want to ever have to go back to work (even part-time) when I pull the trigger.

[George the original one]

> That's just what I feel comfortable with.

In the beginning, this is what is important. In the end, actual performance is what matters. And the SWR studies are there to keep us from having unrealistically positive expectations.

[Tyler9000]

@Workathome - You should ask your PP questions at gyroscopicinvesting.com. The crowd there is very helpful and knowledgeable.

[steveo73]

SWR discussions get you in the ballpark of where you need to be

I think that this is the key. The SWR gives me an initial goal.

I am working on this basis:-

1. Pay off the mortgage.
2. Get to 25 times minimum expected expenses once we are not supporting the 3 kids. For instance now we spend about 40 grand per year but I am budgeting for 25 grand per year when the kids take care of themselves. That to me equates to 500 grand in super (I'm Australian so this is like your 401k).
3. Then get another 500k to last until that super is accessible.

To me this will give me more like a 2% WR however we might spend less or more some years during retirement. I can't see this failing because we can downsize the house and inheritance might help out as well even though I am not considering this at all in my figures.

I think that is plenty for me however I'm also 40 and if I retire at 50 my youngest child will be 12 whereas the oldest will be 21. Basically my kids could cost me more for a number of years and I will never be an early retirer.

[steveo73]

I've determined that my "safe" withdrawal rate is going to be when dividends from my equities holdings reach somewhere around 110% to 125% of my regular living expenses (with the assumptions that my primary residence is owned outright and paid off, and I have zero debt.) This is probably much more conservative than it needs to be. If I put a "number" to it, it would probably be much less than 4%SWR... probably close to 3%. That's just what I feel comfortable with. I don't want to ever have to go back to work (even part-time) when I pull the trigger.

This is what I am looking at as well. I also don't want to go back to work although within reason I may. I might for instance want to go on a big cruise even though that isn't really my style.

[ToFI]

Don't you see I am here to make fun of SWR? SWR came from academia right? Lots of idea from academia are not practical in real life. I said 8% because I pretend like them saying how it can be put into a number.
Lots of variable can affect how long the withdrawal can last.
Such as:
A person can spend more after retirement.
Individual investment performance can be different.
What if a person panics and sell all investment during recession like 2009?

Bottom line, I don't just want the withdrawal to last. I want it to grow while it last so I can give a big chunk to charity before I die.
I use different methods to boost return: Buy and hold high quality businesses, value investing, selling options.

To me, it's not hard to have a 4% withdrawal rate. Maybe 8% is pretty safe.
My target return is 15 to 20% per year. YTD return is 30%.

[jacob]

@ToFI - Hah, be careful about being funny. It can be hard to distinguish between informed&funny and uninformed&serious. Especially on the internet.

[workathome]

My target return is 15 to 20% per year. YTD return is 30%.

That's pretty awesome, 10% alpha.

[ToFI]

My target return is 15 to 20% per year. YTD return is 30%.

That's pretty awesome, 10% alpha.

15 to 20% per year is generated from Charlie Munger style buy and sit on ass.

I got lucky with a few value play this year. It added about 10%.. They show up more often than we think. Sometimes, it's as simple as buying at 52 weeks low. Find a couple value play a year, and I am set for 25%. People think I am nuts to declare this kind of target return. Oh well. Maybe they're mutual fund sales man.