30/70 is as good as 60/40

[watchnerd]

Before I retired, I was at 55 /45. After I retired, I went for 40/60. Fast forward a few years, I'm tired of playing that game and will rebalance this year to 50/50.

Remember that we often give advice to set your asset allocation to the nearest 5%. Why? Going for more precision, like to the nearest 1%, is a lot more work for little added benefit.

There's also an emotional component. If the portfolio drifts a few percent, you won't care. With a 1% step, you'll be wondering what went wrong and want to change things again. So, only select in 5% steps.

Or just don't rebalance at all.

My risk portfolio is global market cap, including the stock bond split.

It goes wherever it goes.

[GaryA505]

Before I retired, I was at 55 /45. After I retired, I went for 40/60. Fast forward a few years, I'm tired of playing that game and will rebalance this year to 50/50.

Remember that we often give advice to set your asset allocation to the nearest 5%. Why? Going for more precision, like to the nearest 1%, is a lot more work for little added benefit.

There's also an emotional component. If the portfolio drifts a few percent, you won't care. With a 1% step, you'll be wondering what went wrong and want to change things again. So, only select in 5% steps.

I like the way you think ...

[HomerJ]

50/50, I plan on just taking from the side that is higher.

Super easy, no math. If I have 800,000 in stocks, 700,000 in bonds, take from stocks.

Slowly automatically rebalances me back to 50/50.

[Barkingsparrow]

50/50, I plan on just taking from the side that is higher.

Super easy, no math. If I have 800,000 in stocks, 700,000 in bonds, take from stocks.

Slowly automatically rebalances me back to 50/50.

That is exactly my plan. Over the years, I have read various recommended books, kept up with a number of "expert" blogs, followed a few Youtube channels, and of course subscribed to innumerable threads on Bogleheads. You can get paralysis by analysis due to the head-spinning variety of opinions on retiree asset allocations and decumulation strategies. As Gandalf once famously said: "Ask not the Bogleheads for advice, because they will tell you both 'yes' and 'no'.”

I gave up and decided that the KISS principle will work just fine, and to just split things down the middle.

[bikechuck]

Retired 12 years and am at 45/55...thinking of maybe upping equities a drop...what do you think?

I am 70 and currently at 43/57, I plan to continue withdrawals/spending from the fixed income side only hopefully resulting in a progressively increasing % of equities. Then if I ever get to 55/45 I will rebalance to 50/50.

[GaryA505]

Retired 12 years and am at 45/55...thinking of maybe upping equities a drop...what do you think?

I am 70 and currently at 43/57, I plan to continue withdrawals/spending from the fixed income side only hopefully resulting in a progressively increasing % of equities. Then if I ever get to 55/45 I will rebalance to 50/50.

I have considered just using 50/50 but I think I'm going to do it somewhat like you are. I'm at 40/60 now at retirement so can spend down the fixed income first, to move towards 50/50 or maybe even 60/40 for the long run.

[xxd091]

U.K. Boglehead
Retired 20+ years now aged 77
Rtd with a 30/70 portfolio -3 global index funds only for equities and bonds
Currently 33/62/5. 5= 2 years cash living expenses
3.2-3.8 % pa withdrawal rate
In practice single yearly withdrawals were mostly from selling equities -only very occasionally used bonds
So as you would expect equity provided the growth-bonds provided the “anchor “
“Anchor “ luckily not needed very often in practice
Don’t seem to need to change Asset Allocation that you originally chose and are happy with
xxd091

[MathWizard]

https://earlyretirementnow.com/2016/12/ ... t-1-intro/

This chart shows probability of the portfolio surviving 30 years using a 4% withdrawal rate with the close 75/25 and 50/50 AAs to be 99% and 95% resp. , while for 25/75 which is close to 30/70 , the probability is only 80%.

How can one believe the probabilities observed by the OP using the nest egg calculator? Is the calculator wrong, or is the interpretation wrong?

[rule of law guy]

I rebalance, but not based upon relative asset size but on my perception of value.

when rates were going up, I rebalanced into more debt because debt was earning more, hence more perceived future value.

with stocks, I am mostly in SPY and I will buy into a correction, and hold on the upswing. with 30/70 I have no need to sell stocks when they get a bit expensive....though I do remember 2000 and if that comes around again, I hope to have my selling hat on

[HicksSt]

https://earlyretirementnow.com/2016/12/ ... t-1-intro/

This chart shows probability of the portfolio surviving 30 years using a 4% withdrawal rate with the close 75/25 and 50/50 AAs to be 99% and 95% resp. , while for 25/75 which is close to 30/70 , the probability is only 80%.

How can one believe the probabilities observed by the OP using the nest egg calculator? Is the calculator wrong, or is the interpretation wrong?

I want to like this chart but the underlying data is a set of overlapping periods. Overlapping periods are not statistically independent. The true sample size is about 5.

[BJJ_GUY]

Ignore valuations. They have been poor at predicting returns ever since they were "discovered".

What is the point of investing in stocks? Do they just increase over time on an entirely arbitrary basis?

It's not a closed system. Billions of humans go to work every day, and energy is pulled from the ground and from the wind and from the sun, and therefore every day more value is input into the system.

So the system becomes more valuable over time.

The only thing I assume, based on the above, is that investing has a positive real return over the long run. That's it.

My point about valuations is that they are not useful for making market-timing moves. They have not predicted returns well since Shiller formulated them in 1988. US Valuations almost immediately went "high" in 1992 and have remained "high" ever since.

I'm honestly not entirely sure what you mean. 'More value is put into the system'? What is value in your definition, and what is the system?

Anyway, the point I was making wasn't about trying to time markets. There is a different between making allocation adjustments based on valuations and making them based on a result in a short time period (or any specific pre-defined time period).

Buying stocks is literally buying a contractual right to a percentage of profits. You are trading cash today in exchange for a perpetual cash stream plus some terminal value.

There is actually really strong relationship between valuation and subsequent stock market returns. Granted, P/E and even adjusted P/E are not the best, and the 10-12 year time-frame is not often exactly correct. Still, there is a high level of statistical significance over time.

So yeah, valuations aren't a great tool for predicting equity market returns in any given 10 year period. But to say valuations don't have a relationship with expected returns makes no sense to me, and would suggest there is zero reason to own them in the first place.

[watchnerd]

Ignore valuations. They have been poor at predicting returns ever since they were "discovered".

What is the point of investing in stocks? Do they just increase over time on an entirely arbitrary basis?

It's not a closed system. Billions of humans go to work every day, and energy is pulled from the ground and from the wind and from the sun, and therefore every day more value is input into the system.

So the system becomes more valuable over time.

The only thing I assume, based on the above, is that investing has a positive real return over the long run. That's it.

My point about valuations is that they are not useful for making market-timing moves. They have not predicted returns well since Shiller formulated them in 1988. US Valuations almost immediately went "high" in 1992 and have remained "high" ever since.

I'm honestly not entirely sure what you mean. 'More value is put into the system'? What is value in your definition, and what is the system?

Anyway, the point I was making wasn't about trying to time markets. There is a different between making allocation adjustments based on valuations and making them based on a result in a short time period (or any specific pre-defined time period).

Buying stocks is literally buying a contractual right to a percentage of profits. You are trading cash today in exchange for a perpetual cash stream plus some terminal value.

There is actually really strong relationship between valuation and subsequent stock market returns. Granted, P/E and even adjusted P/E are not the best, and the 10-12 year time-frame is not often exactly correct. Still, there is a high level of statistical significance over time.

So yeah, valuations aren't a great tool for predicting equity market returns in any given 10 year period. But to say valuations don't have a relationship with expected returns makes no sense to me, and would suggest there is zero reason to own them in the first place.

If the "system" is human beings toiling and the sun and wind providing energy, that's been true since at least the invention of agriculture.

But I don't think I'd describe ancient Egypt or Sumeria economics as representative of modern shared-stock owning capitalism.

[CloseEnough]

50/50, I plan on just taking from the side that is higher.

Super easy, no math. If I have 800,000 in stocks, 700,000 in bonds, take from stocks.

Slowly automatically rebalances me back to 50/50.

That is exactly my plan. Over the years, I have read various recommended books, kept up with a number of "expert" blogs, followed a few Youtube channels, and of course subscribed to innumerable threads on Bogleheads. You can get paralysis by analysis due to the head-spinning variety of opinions on retiree asset allocations and decumulation strategies. As Gandalf once famously said: "Ask not the Bogleheads for advice, because they will tell you both 'yes' and 'no'.”

I gave up and decided that the KISS principle will work just fine, and to just split things down the middle.

Yes that is exactly my plan too. Except at 60/40. And except I will plan for a different allocation to bridge to SS, which could be viewed as a different allocation maybe more like 50/50, or an extra bucket above the 60/40 of fixed income "safe" money to cover expense when W-2 income is gone.

[watchnerd]

you have the misfortune of living longer than planned.

time to take up skydiving at 80

[exodusing]

https://earlyretirementnow.com/2016/12/ ... t-1-intro/

This chart shows probability of the portfolio surviving 30 years using a 4% withdrawal rate with the close 75/25 and 50/50 AAs to be 99% and 95% resp. , while for 25/75 which is close to 30/70 , the probability is only 80%.

How can one believe the probabilities observed by the OP using the nest egg calculator? Is the calculator wrong, or is the interpretation wrong?

No it doesn't. It shows the probability of a portfolio surviving 30 years during a specific time period in the past which may or may not bear any relation to the future. There were only a handful of independent 30 year periods studied, those periods may not be comparable to today (the world has changed, investment structures have changed, etc.) and the data may not be reliable.

We just don't know the future odds. Chance dominates realized returns.

[petulant]

50/50, I plan on just taking from the side that is higher.

Super easy, no math. If I have 800,000 in stocks, 700,000 in bonds, take from stocks.

Slowly automatically rebalances me back to 50/50.

I run the retirement finances for an older family member, and this is how we set her plan up. It works really well to keep her on board even though she is not really a finance person.

[xxd091]

Withdrawals should always be made so that the portfolio is rebalanced back to the desired Asset Allocation
In other words a yearly rebalance if only one withdrawal per year is the plan
Works as well in Accumulation phase ie additional monies should be added to rebalance portfolio
Keep the systems simple understandable and cheap!
xxd091

[Claudia Whitten]

I love 50/50.

I'm never wrong. If stocks are up, I'm smart for having 50% in stocks. If stocks are down, I'm smart for having 50% in fixed income.

Plus the math is easy

It's not so easy if the stock part of that 50 percent goes down by 50 percent and stays down--and you need the money. It helps to think of the stock percentage in dollar terms, not percentage terms. If that 50 percent is $1M and turns into $500K while you're taking distributions, how does that feel? Doable?

Remember: more risk is more risk. In retirement, for many, the game is capital preservation more than capital appreciation.

"It won't happen to me" or "If that happens, we'll have bigger problems" are not good rationales for taking on more risk than necessary.

[CloseEnough]

Remember: more risk is more risk. In retirement, for many, the game is capital preservation more than capital appreciation.

"It won't happen to me" or "If that happens, we'll have bigger problems" are not good rationales for taking on more risk than necessary.

Depends on what kind of risk you are talking about. Reduction of one type of risk can often mean increasing another type of risk.

[HomerJ]

Ignore valuations. They have been poor at predicting returns ever since they were "discovered".

What is the point of investing in stocks? Do they just increase over time on an entirely arbitrary basis?

It's not a closed system. Billions of humans go to work every day, and energy is pulled from the ground and from the wind and from the sun, and therefore every day more value is input into the system.

So the system becomes more valuable over time.

The only thing I assume, based on the above, is that investing has a positive real return over the long run. That's it.

My point about valuations is that they are not useful for making market-timing moves. They have not predicted returns well since Shiller formulated them in 1988. US Valuations almost immediately went "high" in 1992 and have remained "high" ever since.

I'm honestly not entirely sure what you mean. 'More value is put into the system'? What is value in your definition, and what is the system?

I don't know how to explain it.

It seems self-explanatory to me. Even if valuations remain constant, a business can be worth more and more over time.

A business is worth more over time because of the work put into it.

An IPhone is worth more than the ore taken from the ground that is used to make the components. The software written by multiple humans over time makes the IPhone more valuable as well.

Apple doesn't "increase over time on an entirely arbitrary basis".

All the products on Wal-marts shelves are worth more than their component parts. The guys in Walmart logistics who move the merchandise around more effectively from warehouse to store at lower costs make Walmart more profitable too. All that work matters.

That's the main reason stocks rise in value over time. Sure, people can value companies at different prices at different times, based on future expectations (valuations), and the price can change just purely because of increased or decreased valuations.

But the long-term rise of stock and value of all the companies is not based on valuations.

Anyway, the point I was making wasn't about trying to time markets. There is a different between making allocation adjustments based on valuations and making them based on a result in a short time period (or any specific pre-defined time period).

Buying stocks is literally buying a contractual right to a percentage of profits. You are trading cash today in exchange for a perpetual cash stream plus some terminal value.

There is actually really strong relationship between valuation and subsequent stock market returns. Granted, P/E and even adjusted P/E are not the best, and the 10-12 year time-frame is not often exactly correct. Still, there is a high level of statistical significance over time.

So yeah, valuations aren't a great tool for predicting equity market returns in any given 10 year period. But to say valuations don't have a relationship with expected returns makes no sense to me, and would suggest there is zero reason to own them in the first place.

That relationship has changed over time, and therefore has not proven be predictive. That's it.

People here like to say "high" valuations predict lower returns. But the definition of "high" has changed over time. So CAPE of 25 now predicts close to historical average returns. Less than 30 years ago, a CAPE of 25 predicted 0% 10-year real returns. That's not a small change.

Valuations are not actionable, because the relationship has not been stable.

And here's the real thing. If you believe in valuations, then you have to believe in them BOTH ways.

If indeed, "high" valuations predicts lower returns, and we actually get the lower returns, then valuations will become "low", and start predicting higher returns.

So, in the long run, it all averages out. And that's what we see in history. Cycles of poor returns, followed by cycles of good returns, repeat, repeat, repeat.

The long-term historical averages of the US stock market AND the ex-US stock market INCLUDES all the crashes and bad years.

Just buying and holding and ignoring valuations, so far, in US and ex-US, has made investors wealthy over the long run.

[exodusing]

All else being equal, valuations predict returns. This should be obvious - look at bonds, where the YTM at purchase predicts the return.

However, all else is decidedly not equal. For stocks, future earnings, dividends and valuations change, essentially eliminating any useful predictive power from valuations. The future is unknowable, over any given timeframe.

Valuations have been drifting up since about 1980, although hardly in anything resembling a straight line. See https://www.multpl.com/s-p-500-pe-ratio and https://www.multpl.com/shiller-pe

[BJJ_GUY]

Ignore valuations. They have been poor at predicting returns ever since they were "discovered".

What is the point of investing in stocks? Do they just increase over time on an entirely arbitrary basis?

It's not a closed system. Billions of humans go to work every day, and energy is pulled from the ground and from the wind and from the sun, and therefore every day more value is input into the system.

So the system becomes more valuable over time.

The only thing I assume, based on the above, is that investing has a positive real return over the long run. That's it.

My point about valuations is that they are not useful for making market-timing moves. They have not predicted returns well since Shiller formulated them in 1988. US Valuations almost immediately went "high" in 1992 and have remained "high" ever since.

I'm honestly not entirely sure what you mean. 'More value is put into the system'? What is value in your definition, and what is the system?

I don't know how to explain it.

It seems self-explanatory to me. Even if valuations remain constant, a business can be worth more and more over time.

A business is worth more over time because of the work put into it.

An IPhone is worth more than the ore taken from the ground that is used to make the components. The software written by multiple humans over time makes the IPhone more valuable as well.

Apple doesn't "increase over time on an entirely arbitrary basis".

All the products on Wal-marts shelves are worth more than their component parts. The guys in Walmart logistics who move the merchandise around more effectively from warehouse to store at lower costs make Walmart more profitable too. All that work matters.

That's the main reason stocks rise in value over time. Sure, people can value companies at different prices at different times, based on future expectations (valuations), and the price can change just purely because of increased or decreased valuations.

But the long-term rise of stock and value of all the companies is not based on valuations.

Anyway, the point I was making wasn't about trying to time markets. There is a different between making allocation adjustments based on valuations and making them based on a result in a short time period (or any specific pre-defined time period).

Buying stocks is literally buying a contractual right to a percentage of profits. You are trading cash today in exchange for a perpetual cash stream plus some terminal value.

There is actually really strong relationship between valuation and subsequent stock market returns. Granted, P/E and even adjusted P/E are not the best, and the 10-12 year time-frame is not often exactly correct. Still, there is a high level of statistical significance over time.

So yeah, valuations aren't a great tool for predicting equity market returns in any given 10 year period. But to say valuations don't have a relationship with expected returns makes no sense to me, and would suggest there is zero reason to own them in the first place.

That relationship has changed over time, and therefore has not proven be predictive. That's it.

People here like to say "high" valuations predict lower returns. But the definition of "high" has changed over time. So CAPE of 25 now predicts close to historical average returns. Less than 30 years ago, a CAPE of 25 predicted 0% 10-year real returns. That's not a small change.

Valuations are not actionable, because the relationship has not been stable.

And here's the real thing. If you believe in valuations, then you have to believe in them BOTH ways.

If indeed, "high" valuations predicts lower returns, and we actually get the lower returns, then valuations will become "low", and start predicting higher returns.

So, in the long run, it all averages out. And that's what we see in history. Cycles of poor returns, followed by cycles of good returns, repeat, repeat, repeat.

The long-term historical averages of the US stock market AND the ex-US stock market INCLUDES all the crashes and bad years.

Just buying and holding and ignoring valuations, so far, in US and ex-US, has made investors wealthy over the long run.

Companies that generate profits, and can grow their free cash flow over time, is what creates economic value in this context. A valuation has nothing to do with this concept. A valuation is just a method of showing a standardized relationship between price paid per unit of value being purchased.

To say valuations aren't information, and mathematically, and economically meaningful because valuations are not constant is the same as saying arithmetic is not reliable or useful because 8/2=4 but 8/4=2. The fact that the answer to those to calculations change is not an indictment of using division.

I don't think we're probably going to get much use out of discussing the strength of the relationship between valuations and subsequent returns. This relationship is statistically significant over time. But we are using different ways of defining the strength of the relationship. I'm referencing explanatory power based on statistics. I am not saying a given point-to-point time period will allow a valuation to predict with accuracy what the return will be exactly 10 years later. So, basically we will continue talking past each other based on the definition of success

[exodusing]

I don't think we're probably going to get much use out of discussing the strength of the relationship between valuations and subsequent returns. This relationship is statistically significant over time. But we are using different ways of defining the strength of the relationship. I'm referencing explanatory power based on statistics. I am not saying a given point-to-point time period will allow a valuation to predict with accuracy what the return will be exactly 10 years later. So, basically we will continue talking past each other based on the definition of success

Please post your statistical evidence.

[watchnerd]

All else being equal, valuations predict returns. This should be obvious - look at bonds, where the YTM at purchase predicts the return.

However, all else is decidedly not equal. For stocks, future earnings, dividends and valuations change, essentially eliminating any useful predictive power from valuations. The future is unknowable, over any given timeframe.

Valuations have been drifting up since about 1980, although hardly in anything resembling a straight line. See https://www.multpl.com/s-p-500-pe-ratio and https://www.multpl.com/shiller-pe

Sometimes I feel like Jack Bogle when he said something to the effect of:

"Sometimes I don't even know why I own stocks"

(when bonds were yielding 7%, letting you double your money in 10 years)

[HomerJ]

To say valuations aren't information, and mathematically, and economically meaningful because valuations are not constant is the same as saying arithmetic is not reliable or useful because 8/2=4 but 8/4=2. The fact that the answer to those to calculations change is not an indictment of using division.

The analogy is flawed. It's not math itself (like division) that has changed, it's the equation that has changed.

It's a pure slope, y-intercept equation. y=mx+b

And the slope and y-intercept are different now than they were 25 years ago. So all calculations using the old equation gave the wrong answer, even with large error bars.

And we can't be certain it won't change again, and so we can't be sure calculations using today's equation will give more accurate answers.

In 2012, Cliff Asness, CORRECTLY, using the data he had at the time, wrote this:

https://www.aqr.com/Research-Archive/Re ... Shiller-PE

The S&P 500 Shiller P/E, a particularly useful measure of the valuation of the entire U.S. stock market, was 22.2 on September 30, 2012. While that is not close to historic excesses — it is almost exactly half of its peak value during the 1999–2000 stock market bubble and about two-thirds its height in late 1929 — it is high versus history generally. In fact, it’s higher than it has been 80% of the time since 1926.

Based on the past, the 2012 level of Shiller P/E — the ratio of stock prices to an inflation-adjusted 10-year rolling average of corporate earnings — suggested that the average annual real stock market return over the next decade would not exceed 1%. At similar levels in the past, the worst case horrendous: –4.4%. The best case is very good — about 8.3% annually — though it is less wonderful than the much better best cases from lower starting Shiller P/E’s.

Cliff Asness, smarter than most, PhD, professional, thought 22.2 CAPE was "high" (correctly at the time), and expected returns were 1% real. Now, he at least gave us a range of historical returns from -4.4% real to 8.3% real.

And then we got 10% a year real instead. An equation that spits out 1% real a year, and we actually get 10% real a year is not just a little bit wrong.

And NOW CAPE of 22 is no longer considered "high". Because 2003-2013, 2004-2014, 2005-2015, 2006-2016, etc. data points exist now. Data above 20 CAPE was somewhat limited in the past anyway, but recently we have a ton of 10-year data points with CAPEs in the 20s, and they all show decent returns.

A mere eleven years ago, 22 CAPE was still considered "high", forecasting very poor returns. Now it's considered "normal" (some might even say it's a buying opportunity)

A metric that changes that drastically in a mere 11 years is not a useful tool.

I'm not saying it has zero information. I'm saying it doesn't have enough information to be useful.

[GaryA505]

To say valuations aren't information, and mathematically, and economically meaningful because valuations are not constant is the same as saying arithmetic is not reliable or useful because 8/2=4 but 8/4=2. The fact that the answer to those to calculations change is not an indictment of using division.

The analogy is flawed. It's not math itself (like division) that has changed, it's the equation that has changed.

It's a pure slope, y-intercept equation. y=mx+b

And the slope and y-intercept are different now than they were 25 years ago. So all calculations using the old equation gave the wrong answer, even with large error bars.

And we can't be certain it won't change again, and so we can't be sure calculations using today's equation will give more accurate answers.

In 2012, Cliff Asness, CORRECTLY, using the data he had at the time, wrote this:

https://www.aqr.com/Research-Archive/Re ... Shiller-PE

The S&P 500 Shiller P/E, a particularly useful measure of the valuation of the entire U.S. stock market, was 22.2 on September 30, 2012. While that is not close to historic excesses — it is almost exactly half of its peak value during the 1999–2000 stock market bubble and about two-thirds its height in late 1929 — it is high versus history generally. In fact, it’s higher than it has been 80% of the time since 1926.

Based on the past, the 2012 level of Shiller P/E — the ratio of stock prices to an inflation-adjusted 10-year rolling average of corporate earnings — suggested that the average annual real stock market return over the next decade would not exceed 1%. At similar levels in the past, the worst case horrendous: –4.4%. The best case is very good — about 8.3% annually — though it is less wonderful than the much better best cases from lower starting Shiller P/E’s.

Cliff Asness, smarter than most, PhD, professional, thought 22.2 CAPE was "high" (correctly at the time), and expected returns were 1% real. Now, he at least gave us a range of historical returns from -4.4% real to 8.3% real.

And then we got 10% a year real instead. An equation that spits out 1% real a year, and we actually get 10% real a year is not just a little bit wrong.

And NOW CAPE of 22 is no longer considered "high". Because 2003-2013, 2004-2014, 2005-2015, 2006-2016, etc. data points exist now. Data above 20 CAPE was somewhat limited in the past anyway, but recently we have a ton of 10-year data points with CAPEs in the 20s, and they all show decent returns.

A mere eleven years ago, 22 CAPE was still considered "high", forecasting very poor returns. Now it's considered "normal" (some might even say it's a buying opportunity)

A metric that changes that drastically in a mere 11 years is not a useful tool.

I'm not saying it has zero information. I'm saying it doesn't have enough information to be useful.

Nobody knows nuthin. Some days I think I should be 30/70 (according to some experts) and some days I think I should be 70/30 (according to some other experts). I think I'll just go with 50/50.

[BJJ_GUY]

I don't think we're probably going to get much use out of discussing the strength of the relationship between valuations and subsequent returns. This relationship is statistically significant over time. But we are using different ways of defining the strength of the relationship. I'm referencing explanatory power based on statistics. I am not saying a given point-to-point time period will allow a valuation to predict with accuracy what the return will be exactly 10 years later. So, basically we will continue talking past each other based on the definition of success

Please post your statistical evidence.

Link (below) for one of many examples out there. And I'm not saying this one analysis is the single answer to this topic. My point was rather that there are useful methodologies that are different from just using P/E, and especially more useful than putting together a dot plot which charts a handful of calendar year returns (just as an example for comparison).

https://www.hussmanfunds.com/pdf/Hussma ... ePaper.pdf

[dozer183e]

deleted

[exodusing]

deleted

[BJJ_GUY]

To say valuations aren't information, and mathematically, and economically meaningful because valuations are not constant is the same as saying arithmetic is not reliable or useful because 8/2=4 but 8/4=2. The fact that the answer to those to calculations change is not an indictment of using division.

The analogy is flawed. It's not math itself (like division) that has changed, it's the equation that has changed.

It's a pure slope, y-intercept equation. y=mx+b

And the slope and y-intercept are different now than they were 25 years ago. So all calculations using the old equation gave the wrong answer, even with large error bars.

And we can't be certain it won't change again, and so we can't be sure calculations using today's equation will give more accurate answers.

In 2012, Cliff Asness, CORRECTLY, using the data he had at the time, wrote this:

https://www.aqr.com/Research-Archive/Re ... Shiller-PE

The S&P 500 Shiller P/E, a particularly useful measure of the valuation of the entire U.S. stock market, was 22.2 on September 30, 2012. While that is not close to historic excesses — it is almost exactly half of its peak value during the 1999–2000 stock market bubble and about two-thirds its height in late 1929 — it is high versus history generally. In fact, it’s higher than it has been 80% of the time since 1926.

Based on the past, the 2012 level of Shiller P/E — the ratio of stock prices to an inflation-adjusted 10-year rolling average of corporate earnings — suggested that the average annual real stock market return over the next decade would not exceed 1%. At similar levels in the past, the worst case horrendous: –4.4%. The best case is very good — about 8.3% annually — though it is less wonderful than the much better best cases from lower starting Shiller P/E’s.

Cliff Asness, smarter than most, PhD, professional, thought 22.2 CAPE was "high" (correctly at the time), and expected returns were 1% real. Now, he at least gave us a range of historical returns from -4.4% real to 8.3% real.

And then we got 10% a year real instead. An equation that spits out 1% real a year, and we actually get 10% real a year is not just a little bit wrong.

And NOW CAPE of 22 is no longer considered "high". Because 2003-2013, 2004-2014, 2005-2015, 2006-2016, etc. data points exist now. Data above 20 CAPE was somewhat limited in the past anyway, but recently we have a ton of 10-year data points with CAPEs in the 20s, and they all show decent returns.

A mere eleven years ago, 22 CAPE was still considered "high", forecasting very poor returns. Now it's considered "normal" (some might even say it's a buying opportunity)

A metric that changes that drastically in a mere 11 years is not a useful tool.

I'm not saying it has zero information. I'm saying it doesn't have enough information to be useful.

Where is the slope and intercept in a valuation? It's literally price divided by earnings. It really is simple arithmetic. Understanding the meaning of a valuation seems to be the challenge in this discussion.

There Is no regression required to first understand the logic behind valuations. That's just a way of measuring how one variable explains the variability of another variable. Which is fine as an ex post tool, especially if someone were trying to say valuations are great at predicting precise point-to-point results, and should be used to time markets. But that isn't what I've said, and I don't think anyone else has.

What I've said is that long term returns, over time, are largely explained by valuations. However, there are absolutely times where the market overshoots significantly, and for longer than one might expect. That's because there are multiple factors that can impact the short-term pricing (investor behavior, structural changes, central banks etc.).

The bottom line is price relative to future value is the equation that matters. This is why bubbles may persist, but not into perpetuity. Economic reality has a funny way of getting in the way.