Do Indexers even need to pay attention to market valuation?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

Scooter57 wrote: Thu Jul 22, 2021 4:58 pm Thing is, 1996 marked the beginning of the masses being introduced to the internet, a development as transformative as the railroad, the automobile, and air conditioning. Now we are working out the details, but it is far from certain that you will see the kinds of new industries and mega companies emerging with enormous growth in the next 20-30 years like we did since 1996.

Current valuations are based on the assumption of tremendous real annual earnings growth for decades to come. (As opposed to financially engineered earnings growth based on borrowing and buybacks IBM-style without sales growth.) If that doesn't happen things get interesting.
It's a good point. This absolutely could just be a 30-year aberration, although electricity was even more of a game-changer in the early 1900s.

Software is pretty easy to scale though. Doesn't take much more effort to sell 2 million instead of 1 million units.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
BJJ_GUY
Posts: 882
Joined: Wed Mar 13, 2019 7:45 am

Re: Do Indexers even need to pay attention to market valuation?

Post by BJJ_GUY »

HomerJ wrote: Thu Jul 22, 2021 4:22 pm
ScubaHogg wrote: Thu Jul 22, 2021 2:52 pmTo use a little reductio ad absurdum, if bonds and what-not stayed roughly at their current yields, but the TSM shot up to a PE10 of 200, would you just say “nobody knows enough” and do nothing? Maybe you would. Heck maybe I would. But I don’t know that just the fact I might not do anything would make it a wise decision.
Since I rebalance a 50/50 portfolio, I would have locked in enough gains to retire comfortably even if TSM thereafter crashed 90% after rising to a PE of 200 :)
(And a PE of 25 shouldn’t seem extreme in the mid 90s when you could look across the pacific from a few years earlier and see a developed market with a PE almost 4x higher. Maybe 25 felt high, but “extreme” would have been hyperbolic)
No, it was considered extreme. Every time CAPE crossed 20 in the past, a crash occurred soon after.

It had only gone higher than 25 once, and that was in 1929, right before the Great Depression (where stocks ultimately dropped 88% at one point from their 1929 highs).

Crossing 25 was HUGE. It hadn't been that high in nearly 70 years. Irrational Exuberance speech.

But actually 1996 ended up a good time to buy. Which is the crazy part... It wasn't that 1996 turned to be not as bad as predicted. It was actually a GOOD time to buy. Highest valuations in 70 years, and it was a good time to buy. Even during the crash of 2000-2003, the market never dropped as low as 1996.

The model was invented in 1988 using 1876-1988 data, and less than 10 years later it utterly failed as a prediction tool. And again and again and again in the following years.

In 2011, CAPE predicted 4.5% real 10-year returns... Instead we've gotten 13% real. That's not just a little bit off. That's wildly wrong. 99% percentile wrong.

Maybe we keep getting lucky. Or maybe, just maybe... the model is missing some important variables. Maybe it's not easy to predict the future.
Valuations don't tell you when something is going to happen. This is interpreting the metric incorrectly (that isn't to say people haven't attempted this as a trading strategy).

Valuations simply give you information about how much you are paying today (PV) in relation to the (best estimation of) long-term value you'll receive far into the future (FV). That's it. Pretty basic math can solve the equation for expected returns over a give time period and based on whatever inputs you use for an approximation for fair value (crude proxy: median valuation, and maybe adjust GDP/revenues and margins to also revert to historical norms).

Investor sentiment is a strong force. It can extend cycles, and ultimately it's what triggers sell-offs and the shape of the sell-off. However, sentiment cannot change the fundamentals that underpin valuation, or the worth of an asset being purchased.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

BJJ_GUY wrote: Thu Jul 22, 2021 4:12 pm
Nathan Drake wrote: Thu Jul 22, 2021 8:57 am I agree that using CAPE to forecast exact returns doesn’t have much predictive power. I can’t look at US CAPE and say it will definitively be poor. Valuations could continue to increase.

The idea that Meb points out is that over the long term, the wider the spreads in CAPE are amongst asset classes with similar expected long term returns would allow us to possibly tactically tilt away from those that are much more highly valued than others.

Which is to say, CAPE won’t tell me if US returns 10% next decade, but it does suggest that if US is greatly overpriced relative to EM, that EM is more likely to have stronger returns than US going forward.

I would argue that we are approaching such a time right now. That still may mean that for you personally no changes to your AA are desired, but I think one can look at the data and take reasonable action knowing the risks.
Keep fighting the good fight, Nathan. I've previously attempted to make many of the same points that you have in this thread, and most folks on this group aren't wired to think about valuations and the fundamental relationship to future returns. If something can't be easily back-tested with perfect results, and/or if they can make a counter-point using the last ten years, then it just isn't going to resonate.

If I had my way, I'd change the way the debate is set-up, and the framed with the following:
1.) Stop using P/E based valuations (like Shiller). A better valuation for the broad market is: Non-financial Market cap / Gross Value Added (or a crude approximation of market cap / GDP) are materially better than P/E versions. A better method would result in fewer arguments about the lack of statistical significance between valuations and subsequent returns.

2.) Valuations can simply be used to understand relative over-priced vs under-priced markets. To utilize this, no one needs to make an assumption on the long term return expectations across various markets. Instead, you can simply figure out how many standard deviations from historical median valuation each broad asset class is.

3.) Think about using valuations as a tool to avoid losses rather than just a way of seeking the best means of increasing returns when everything is based on baseline estimations. Valuations helped smart investors avoid, or at least shade away, from tech in the late 1990s and away from real estate, banks, and other trouble spots in 2007/8. The avoidance, or underweight, to hot markets hurt for many years, and these investors were ridiculed for underperformance. In the end, avoiding the losses of a brief, but painful market drawdown, more than made up for the years of foregone paper gains during over-extended markets.
These are all great points.

There have been some very painful lessons throughout history that I’m afraid some haven’t learned:

66-82 US Bear
80s Japanese (exUS bubble)
00-09 US tech correction
07-21 EM correction (EM overvaluation)

In some cases, these “bubbles” seemed obvious in retrospect. Everyone hand waives Japanese stagnation as some isolated event, but the US has already encountered numerous periods that were quite long and painful.

The typical response is “it doesn’t matter for a long term
Investor”. My response is: can we do better? Why would I place most/all my eggs in one basket and risk a 15 year bear market with only the only response being “well, maybe the next 15 will make up for it”. Of course there’s no guarantee it will.

I’d rather diversify my portfolio in a way that puts odds in my favor for a high likelihood of good performance every 5-10 year period. Will it be BEST at any moment in time? No, diversification will always have winners and losers. But that’s the point - winners aren’t knowable in advance, but using some form of valuation in your AA stacks the odds in your favor of reducing tail outcomes and possibly increasing the likelihood of more smoothed returns (and possibly much higher returns if Value outperforms Growth as is historically the case)
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
atlgenxennial
Posts: 110
Joined: Tue Jul 11, 2017 1:45 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by atlgenxennial »

Buy_N_Hold wrote: Sat Jul 17, 2021 7:17 pm Does it make sense to try to risk off a bit when things seem toppy, and adjust your asset allocation of stocks accordingly, or is this kind of thinking essentially just trying to time the market?
The emphasized word is the key here. You don't know, nor do you pretend to know, by what parameter to sell or "risk off".

To answer your question, yes that would be trying to time the market, and it would be really unwise.
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

BJJ_GUY wrote: Thu Jul 22, 2021 5:10 pmValuations simply give you information about how much you are paying today (PV) in relation to the (best estimation of) long-term value you'll receive far into the future (FV). That's it. Pretty basic math can solve the equation for expected returns over a give time period and based on whatever inputs you use for an approximation for fair value (crude proxy: median valuation, and maybe adjust GDP/revenues and margins to also revert to historical norms).
I see a lot of "best estimates" and "approximation" of multiple variables above.

The pretty basic math is not so basic after all.

For the past 30 years, the "expected" returns have been way off.

We're not talking the expected return is 4.5% plus or minus 8% (which is a HUGE error band), and we actually got 6% real or even 8% real.

The "expected" 10-year return in 2011 was 4.5% with 2 standard deviations up 12.5%, and instead we got 13.5% real over those 10 years. Possible, yes. We could have just hit the 1% chance. But that shouldn't assure anyone that the model has no flaws.

The pretty basic math has not worked. If you have a model that predicts a 1 in a 100 year flood at a certain level, and you get a flood that high three times in 15 years, it's reasonable to wonder, if maybe, just maybe, the model has a flaw or if a variable has changed.

I don't understand why this is so hard. Look at past "expected" return calculations that were made at the time, and see how wrong valuations predictions have been since 1992.

DON'T read some article from 2020 that has all the past data, has made a new line, with new expected returns that fit the data from the past 30 years. That proves nothing.

Look at predictions made in 1996 with 1900-1995 data, Look at predictions made in 2007 with 1900-2006 data. Look at predictions made in 2011 with 1900-2010 data... They were wildly off.

Not a little bit off... Wildly off. Valuations proponents can say "Well, it's a been a weird time with the Internet and investor sentiment and Fed interference in the markets". I could accept those reasonings as why valuations as a prediction tool has failed for the past 30 years, and maybe it will work again in the future.

What they CANNOT say is "Oh valuations have totally worked as a prediction tool for the past 30 years."

That's what makes me mad. Valuations as a theory makes total sense to me. But I've lived through the past 30 years, seen the predictions in real-time, watched them fail, and then, infuriatingly, the proponents just make new models with the new data, wave their hands, and proclaim, "Oh valuations have totally worked all this time".

NONE of this is to say that a crash won't happen soon. It absolutely could. I predict nothing. I have no idea what will happen next. I'm not saying valuations predictions are wrong. I'm saying valuations have failed over and over as a prediction tool for the past 30 years so don't put too much stock in them. Don't make large changes to your Asset Allocation based on them.

Risk management is important. Always be prepared for long bear market starting tomorrow. Because it might happen. This is always true. Regardless of valuations.
Last edited by HomerJ on Thu Jul 22, 2021 6:41 pm, edited 1 time in total.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: Do Indexers even need to pay attention to market valuation?

Post by nedsaid »

ScubaHogg wrote: Thu Jul 22, 2021 2:55 pm
nedsaid wrote: Thu Jul 22, 2021 12:01 pm
alex_686 wrote: Sun Jul 18, 2021 6:19 pm
ScubaHogg wrote: Sun Jul 18, 2021 5:57 pm My point is there is some logical limit to the claim that we should ignore valuations and stay the course. Maybe we will never see that limit and shouldn’t worry about it (though a Japanese investor would have been well served by worrying about it) but we should probably acknowledge it exists
Well, maybe atypically as a Boglehead, I don’t advocate holding the course.

One should not panic. Trying to time the market is really hard to do. etc.

However, we always say to be calm and write up a ISP. You base your asset allocation based on your market expectations. But the market changes over time. If the market changes shouldn’t you?

MarkRoulo mentioned the 90s where expected real returns of government bonds was 3%. Compare that to now, where it is around 0%. You should come to 2 different conclusions on what your AA should be, right?
Yep, the economy and the markets are dynamic. There are limits to staying the course no matter what. Problem is, I don't know what those limits are.
Yes, I agree. But even getting people to admit there is some limit is difficult. Personally I’d rather at least ponder the problem in theory so in the tiny chance it occurs I’ve at least put some thought into it.
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
A fool and his money are good for business.
BJJ_GUY
Posts: 882
Joined: Wed Mar 13, 2019 7:45 am

Re: Do Indexers even need to pay attention to market valuation?

Post by BJJ_GUY »

HomerJ wrote: Thu Jul 22, 2021 6:24 pm
BJJ_GUY wrote: Thu Jul 22, 2021 5:10 pmValuations simply give you information about how much you are paying today (PV) in relation to the (best estimation of) long-term value you'll receive far into the future (FV). That's it. Pretty basic math can solve the equation for expected returns over a give time period and based on whatever inputs you use for an approximation for fair value (crude proxy: median valuation, and maybe adjust GDP/revenues and margins to also revert to historical norms).
I see a lot of "best estimates" and "approximation" of multiple variables above.

The pretty basic math is not so basic after all.

For the past 30 years, the "expected" returns have been way off.

We're not talking the expected return is 4.5% plus or minus 8% (which is a HUGE error band), and we actually got 6% real or even 8% real.

The "expected" return in 2011 was 4.5% with a max of 12.5%, and instead we got 13.5% real over those 10 years.

The pretty basic math has not worked.

I don't understand why this is so hard. Look at past "expected" return calculations that were made at the time, and see how wrong valuations predictions have been since 1992.

DON'T read some article from 2020 that has all the past data, has made a new line, with new expected returns that fit the data from the past 30 years. That proves nothing.

Look at predictions made in 1996 with 1900-1995 data, Look at predictions made in 2007 with 1900-2006 data. Look at predictions made in 2011 with 1900-2010 data... They were wildly off.

Not a little bit off... Wildly off. Valuations proponents can say "Well, it's a been a weird time with the Internet and investor sentiment and Fed interference in the markets". I could accept those reasonings as why valuations as a prediction tool has failed for the past 30 years, and maybe it will work again in the future.

What they CANNOT say is "Oh valuations have totally worked as a prediction tool for the past 30 years."

That's what makes me mad. Valuations as a theory makes total sense to me. But I've lived through the past 30 years, seen the predictions in real-time, watched them fail, and then, infuriatingly, the proponents just make new models with the new data, wave their hands, and proclaim, "Oh valuations have totally worked all this time".

NONE of this is to say that a crash won't happen soon. It absolutely could. I predict nothing. I have no idea what will happen next. I'm not saying valuations predictions are wrong. I'm saying valuations have failed over and over as a prediction tool for the past 30 years so don't put too much stock in them.

Risk management is important. Always be prepared for long bear market starting tomorrow. Because it might happen. This is always true. Regardless of valuations.
First, stop using Shiller as a baseline. Just use the buffet indicator (Equity Mkt cap / GDP), or better yet the 'non-financial market cap / gross value added' is a better relationship with subsequent 12 year returns. Either way, both are much better than Shiller CAPE.

Secondly, the implied returns are a function of the valuations. Anyone out in the press make predictions isn't what I'm talking about. To that end, in the late 90s, and mid-2000s the implied returns ended up being pretty accurate. While we are getting to a point where we're approaching the end of the current 12 year period, of course it's possible that realized 12 years may look quite different, but there is value in seeing and understanding this (as we know returns are being borrowed from the future etc.).

I can't really speak directly to all the predictions you referenced, because I'm not sure the source. I'd be happy to read more about it if you can share with me.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

HomerJ wrote: Thu Jul 22, 2021 6:24 pm
BJJ_GUY wrote: Thu Jul 22, 2021 5:10 pmValuations simply give you information about how much you are paying today (PV) in relation to the (best estimation of) long-term value you'll receive far into the future (FV). That's it. Pretty basic math can solve the equation for expected returns over a give time period and based on whatever inputs you use for an approximation for fair value (crude proxy: median valuation, and maybe adjust GDP/revenues and margins to also revert to historical norms).
I see a lot of "best estimates" and "approximation" of multiple variables above.

The pretty basic math is not so basic after all.

For the past 30 years, the "expected" returns have been way off.

We're not talking the expected return is 4.5% plus or minus 8% (which is a HUGE error band), and we actually got 6% real or even 8% real.

The "expected" return in 2011 was 4.5% with a max of 12.5%, and instead we got 13.5% real over those 10 years.

The pretty basic math has not worked.

I don't understand why this is so hard. Look at past "expected" return calculations that were made at the time, and see how wrong valuations predictions have been since 1992.

DON'T read some article from 2020 that has all the past data, has made a new line, with new expected returns that fit the data from the past 30 years. That proves nothing.

Look at predictions made in 1996 with 1900-1995 data, Look at predictions made in 2007 with 1900-2006 data. Look at predictions made in 2011 with 1900-2010 data... They were wildly off.

Not a little bit off... Wildly off. Valuations proponents can say "Well, it's a been a weird time with the Internet and investor sentiment and Fed interference in the markets". I could accept those reasonings as why valuations as a prediction tool has failed for the past 30 years, and maybe it will work again in the future.

What they CANNOT say is "Oh valuations have totally worked as a prediction tool for the past 30 years."

That's what makes me mad. Valuations as a theory makes total sense to me. But I've lived through the past 30 years, seen the predictions in real-time, watched them fail, and then, infuriatingly, the proponents just make new models with the new data, wave their hands, and proclaim, "Oh valuations have totally worked all this time".

NONE of this is to say that a crash won't happen soon. It absolutely could. I predict nothing. I have no idea what will happen next. I'm not saying valuations predictions are wrong. I'm saying valuations have failed over and over as a prediction tool for the past 30 years so don't put too much stock in them. Don't make large changes to your Asset Allocation based on them.

Risk management is important. Always be prepared for long bear market starting tomorrow. Because it might happen. This is always true. Regardless of valuations.
Again confusing valuations at different points of time for the same asset class instead of the relative valuations of different asset classes during the same sample period.

I don’t think anyone disagrees with the former, it is the latter that’s the main point being discussed.

There are usually alternatives. I don’t need to be 80-100% US large cap in my equity portion
Last edited by Nathan Drake on Thu Jul 22, 2021 6:38 pm, edited 1 time in total.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
Normchad
Posts: 5648
Joined: Thu Mar 03, 2011 6:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Normchad »

HomerJ wrote: Thu Jul 22, 2021 6:24 pm
BJJ_GUY wrote: Thu Jul 22, 2021 5:10 pmValuations simply give you information about how much you are paying today (PV) in relation to the (best estimation of) long-term value you'll receive far into the future (FV). That's it. Pretty basic math can solve the equation for expected returns over a give time period and based on whatever inputs you use for an approximation for fair value (crude proxy: median valuation, and maybe adjust GDP/revenues and margins to also revert to historical norms).
I see a lot of "best estimates" and "approximation" of multiple variables above.

The pretty basic math is not so basic after all.

For the past 30 years, the "expected" returns have been way off.

We're not talking the expected return is 4.5% plus or minus 8% (which is a HUGE error band), and we actually got 6% real or even 8% real.

The "expected" return in 2011 was 4.5% with a max of 12.5%, and instead we got 13.5% real over those 10 years.

The pretty basic math has not worked.

I don't understand why this is so hard. Look at past "expected" return calculations that were made at the time, and see how wrong valuations predictions have been since 1992.

DON'T read some article from 2020 that has all the past data, has made a new line, with new expected returns that fit the data from the past 30 years. That proves nothing.

Look at predictions made in 1996 with 1900-1995 data, Look at predictions made in 2007 with 1900-2006 data. Look at predictions made in 2011 with 1900-2010 data... They were wildly off.

Not a little bit off... Wildly off. Valuations proponents can say "Well, it's a been a weird time with the Internet and investor sentiment and Fed interference in the markets". I could accept those reasonings as why valuations as a prediction tool has failed for the past 30 years, and maybe it will work again in the future.

What they CANNOT say is "Oh valuations have totally worked as a prediction tool for the past 30 years."

That's what makes me mad. Valuations as a theory makes total sense to me. But I've lived through the past 30 years, seen the predictions in real-time, watched them fail, and then, infuriatingly, the proponents just make new models with the new data, wave their hands, and proclaim, "Oh valuations have totally worked all this time".

NONE of this is to say that a crash won't happen soon. It absolutely could. I predict nothing. I have no idea what will happen next. I'm not saying valuations predictions are wrong. I'm saying valuations have failed over and over as a prediction tool for the past 30 years so don't put too much stock in them. Don't make large changes to your Asset Allocation based on them.

Risk management is important. Always be prepared for long bear market starting tomorrow. Because it might happen. This is always true. Regardless of valuations.
I understand the intuitive appeal of valuations. “Hey, I’m spending $40 to buy $1 of future earnings. That’s historically expensive. This can’t go on….”

But as you say, it just hasn’t panned out.

Part of why I ignore it, it seems to be just another form of “technical analysis”; guys looking at graphs and reading way too much into them.

CAPE is a data mined observation, right? Is there a theory that goes with it? And if so how does the theory explain the ways it’s changed in the past? I guess that’s my biggest issue, it seems like it’s just an observation without an intellectual underpinning that makes sense to me. I’m sure there is one, but nobody is agreeing on what to do with this information.

But for the regular indexer, ignore this. This board will tell you bonds stink. And cash stinks. And based on valuations, equities are going to stink. But you have to put your money somewhere. So which of the predicted-to-stinky places should it be?

Set your AA. Follow your AA.
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: Do Indexers even need to pay attention to market valuation?

Post by nedsaid »

HomerJ wrote: Thu Jul 22, 2021 6:49 pm
nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return.

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
I don't advocate market timing but certainly you can take some off the top when markets are in euphoria. The decision was easier back in 1999 because you could effortlessly collect 6% from your bonds with very little risk. Today, your bonds don't even cover inflation.

No need to time things perfectly but certainly you can take risk off the table when markets get euphoric. As an investor, I have experienced this just once. 2007 wasn't euphoria and 2021 isn't either. These market bubbles are a once in a lifetime experience for most investors unless you live a very long time. I was nine years old back in 1968, pretty doubtful I would have had an investment portfolio back then. 1992 wasn't euphoria, 1996 was getting there but I don't think things got truly euphoric until maybe 1997 or 1998. To me, a co-worker trying to day trade her 403(b) in 1999 was a sign that something was amiss.

If things get insanely euphoric, you can take some off the top. But it is more about reducing risk rather than increasing returns. Market melt-ups, while enjoyable, do add risk to the portfolio. When your crazy brother-in-law mortgages the house and hocks the kids to buy stocks, you know the top is nigh.

Somewhere, way back in my very foggy memory banks, I do remember you making this same point to someone about Tesla. At some point, so much optimism gets priced in that a great company can be turned into a poor investment. This can happen to entire markets as well.

Like I said, Japan was an extreme example, and Japan had a rather extreme bear market that STILL hasn't ended. Think someone said that Japan had P/E's approaching 80, their Real Estate prices were insane as well. They bought Hawaii from us and we bought it back after Japan crashed. :wink:

Pretty much, this might be actionable once in a lifetime. This is a pretty academic discussion. The one thing we CAN do is have a Global portfolio. Vanguard hasn't indexed the stock market on Mars yet so we can't get interplanetary diversification.
Last edited by nedsaid on Thu Jul 22, 2021 7:02 pm, edited 1 time in total.
A fool and his money are good for business.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

HomerJ wrote: Thu Jul 22, 2021 6:49 pm
nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
HootingSloth
Posts: 1050
Joined: Mon Jan 28, 2019 2:38 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HootingSloth »

Nathan Drake wrote: Thu Jul 22, 2021 7:01 pm
HomerJ wrote: Thu Jul 22, 2021 6:49 pm
nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
Pretty sure HomerJ is not a 100% VTSAX kind of person.

Other people skeptical of the valuations-are-actionable thesis, like myself, are also not exactly rosy-eyed optimists. An early 20th century German investor with a globally-diversified stock portfolio still experienced a real drawdown lasting 57 years. That had nothing to do with valuations. That is one kind of outcome, among many, that I try to prepare for. Valuations don't really enter into that planning process because they have not been sufficiently reliable in practice and there are strong theoretical reasons to question their reliability.

Others have plans that rely on valuations and that is OK for them. The risk that those valuations are misleading is one risk they are willing to bear, and it may help them mitigate certain other kinds of risks to a degree. We all have to evaluate these tradeoffs ourselves.
Last edited by HootingSloth on Thu Jul 22, 2021 7:13 pm, edited 1 time in total.
Global Market Portfolio + modest tilt towards volatility (80/20->60/40 as approach FI) + modest tilt away from exchange rate risk (80% global+20% U.S. stocks; currency-hedge bonds) + tax optimization
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

HootingSloth wrote: Thu Jul 22, 2021 7:11 pm
Nathan Drake wrote: Thu Jul 22, 2021 7:01 pm
HomerJ wrote: Thu Jul 22, 2021 6:49 pm
nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
Pretty sure HomerJ is not a 100% VTSAX kind of person.

Other people skeptical of the valuations-are-actionable thesis, like myself, are also not exactly rosy-eyed optimists. An early 20th century German investor with a globally-diversified stock portfolio experienced a real drawdown lasting 57 years. That had nothing to do with valuations. That is one kind of outcome, among many, that I try to prepare for. Valuations don't really enter into that planning process because they have not been sufficiently reliable in practice and there are strong theoretical reasons to question their reliability.
He's an 80% VTSAX kind of person. Just enough to get a little dangerous and place him on Bogle's naughty list.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: Do Indexers even need to pay attention to market valuation?

Post by nedsaid »

Nathan Drake wrote: Thu Jul 22, 2021 7:01 pm
HomerJ wrote: Thu Jul 22, 2021 6:49 pm
nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
Yep. The old Swedroe shuffle. You are right, it is never all or nothing. Sometimes you can just shift into cheaper stocks.
A fool and his money are good for business.
HootingSloth
Posts: 1050
Joined: Mon Jan 28, 2019 2:38 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HootingSloth »

Nathan Drake wrote: Thu Jul 22, 2021 7:12 pm
HootingSloth wrote: Thu Jul 22, 2021 7:11 pm
Nathan Drake wrote: Thu Jul 22, 2021 7:01 pm
HomerJ wrote: Thu Jul 22, 2021 6:49 pm
nedsaid wrote: Thu Jul 22, 2021 6:34 pm
HomerJ keeps going on and on about valuations not mattering. I wonder how Japanese investors feel about that? They did matter in 1929, 1968, and 1999 here in the U.S. We had essentially flat markets for years after each of those three bull market peaks: after 1929 it probably took until 1948 or so to get back to new all-time highs, after 1968 it took until 1984, and after 1999 new all-time highs weren't reached again until 2013. The Japanese Stock Market peaked in 1989, not sure they are even back to even after all of these years, 32 years now and counting.
All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
Pretty sure HomerJ is not a 100% VTSAX kind of person.

Other people skeptical of the valuations-are-actionable thesis, like myself, are also not exactly rosy-eyed optimists. An early 20th century German investor with a globally-diversified stock portfolio experienced a real drawdown lasting 57 years. That had nothing to do with valuations. That is one kind of outcome, among many, that I try to prepare for. Valuations don't really enter into that planning process because they have not been sufficiently reliable in practice and there are strong theoretical reasons to question their reliability.
He's an 80% VTSAX kind of person. Just enough to get a little dangerous and place him on Bogle's naughty list.
He can correct me if I'm wrong, but I thought he had about 40% of his portfolio in US stocks (because 50% is in bonds).

In any event, I have more international than that in my stock portion (80% global + 20% US tilt), but it has precisely nothing to do with valuations. There are other ways to determine an appropriate AA than looking at relative valuations.
Global Market Portfolio + modest tilt towards volatility (80/20->60/40 as approach FI) + modest tilt away from exchange rate risk (80% global+20% U.S. stocks; currency-hedge bonds) + tax optimization
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

HootingSloth wrote: Thu Jul 22, 2021 7:16 pm
Nathan Drake wrote: Thu Jul 22, 2021 7:12 pm
HootingSloth wrote: Thu Jul 22, 2021 7:11 pm
Nathan Drake wrote: Thu Jul 22, 2021 7:01 pm
HomerJ wrote: Thu Jul 22, 2021 6:49 pm

All I can say to that is that long-term historical average of 10% nominal INCLUDES those periods.

Valuations don't matter in the long-run because, so far, bad years are followed by good years, and it all averages out to a healthy good return

(This is actually part of valuation theory, right? If we have a crash or a long sideways period, valuations will come down, and start to predict higher returns going forward - If you believe in valuations, then you should believe in cycles).

Sure, if you can find a timing mechanism that lets you skip the bad years, and just get the good years, you'll make more.

But if your timing mechanism fails, you could end up with less.

And since, doing nothing, so far, makes you 10% a year in the long-run... Seems like doing nothing is a pretty smart move.

Valuations is not an accurate timing mechanism, which is why I say valuations don't matter.

You said valuations mattered in 1999. But anyone following valuations got partially out in 1992, and mostly (or even fully) out in 1996.

And that was way too early, and they likely would have ended up with less money trying to avoid the dot-com crash.
In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
Pretty sure HomerJ is not a 100% VTSAX kind of person.

Other people skeptical of the valuations-are-actionable thesis, like myself, are also not exactly rosy-eyed optimists. An early 20th century German investor with a globally-diversified stock portfolio experienced a real drawdown lasting 57 years. That had nothing to do with valuations. That is one kind of outcome, among many, that I try to prepare for. Valuations don't really enter into that planning process because they have not been sufficiently reliable in practice and there are strong theoretical reasons to question their reliability.
He's an 80% VTSAX kind of person. Just enough to get a little dangerous and place him on Bogle's naughty list.
He can correct me if I'm wrong, but I thought he had about 40% of his portfolio in US stocks (because 50% is in bonds).

In any event, I have more international than that in my stock portion (80% global + 20% US tilt), but it has precisely nothing to do with valuations. There are other ways to determine an appropriate AA than looking at relative valuations.
Yes, that's true. 80% was for Homer's equity portion. So 50% Bonds, 40% VTSAX, 10% VTIAX

This type of portfolio is at very high risk of SWR failure in retirement from where we are right now if you think the past will still hold. I would certainly not bet on it. He'll need to hope US has a Japan style melt-up otherwise he could run into SORR issues due to over reliance on bonds and US mega cap equities. Or ensure a smaller SWR or more flexible SWR strategy.
Last edited by Nathan Drake on Thu Jul 22, 2021 7:24 pm, edited 1 time in total.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
BJJ_GUY
Posts: 882
Joined: Wed Mar 13, 2019 7:45 am

Re: Do Indexers even need to pay attention to market valuation?

Post by BJJ_GUY »

Normchad wrote: Thu Jul 22, 2021 6:38 pm I understand the intuitive appeal of valuations. “Hey, I’m spending $40 to buy $1 of future earnings. That’s historically expensive. This can’t go on….”

But as you say, it just hasn’t panned out.

Part of why I ignore it, it seems to be just another form of “technical analysis”; guys looking at graphs and reading way too much into them.

CAPE is a data mined observation, right? Is there a theory that goes with it? And if so how does the theory explain the ways it’s changed in the past? I guess that’s my biggest issue, it seems like it’s just an observation without an intellectual underpinning that makes sense to me. I’m sure there is one, but nobody is agreeing on what to do with this information.

But for the regular indexer, ignore this. This board will tell you bonds stink. And cash stinks. And based on valuations, equities are going to stink. But you have to put your money somewhere. So which of the predicted-to-stinky places should it be?

Set your AA. Follow your AA.
The appeal to using fundamentals to value a company (or a broad market), is precisely because of the intuition. Obviously there is some variable inputs for equity valuations, but nothing crazy is needed. It's very similar to assessing an investment in a bond. If it cost $120 to buy a 10 year zero coupon bond that will pay $100 at maturity, you can back into the implied YTM. Of course that bond could always trade up to $150 in the interim period, but that doesn't change the fact that the security is only worth $100 (future value). Equities are perpetual and not as simple as that example, obviously, but the idea is still rooted in the same place: How much must you pay today to have claim on perpetual cash flows/earnings plus terminal value.

CAPE is not data mined. They've changed the way they smooth earnings (because earnings are cyclical, volatile, and notoriously less useful than revenues, for example). The underpinnings to CAPE are essentially all the same as the basic outline in the above paragraph. current price is divided by earnings.
HootingSloth
Posts: 1050
Joined: Mon Jan 28, 2019 2:38 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HootingSloth »

Nathan Drake wrote: Thu Jul 22, 2021 7:21 pm
HootingSloth wrote: Thu Jul 22, 2021 7:16 pm
Nathan Drake wrote: Thu Jul 22, 2021 7:12 pm
HootingSloth wrote: Thu Jul 22, 2021 7:11 pm
Nathan Drake wrote: Thu Jul 22, 2021 7:01 pm

In 1999 you didn’t have to get out. You just had to increase exposure to SCV, VXUS, and/or EM as they looked more attractively valued on a relative basis

This is the point I keep hammering home

But if you like dead decades, I guess have fun YOLOing all in
VTSAX for life
Pretty sure HomerJ is not a 100% VTSAX kind of person.

Other people skeptical of the valuations-are-actionable thesis, like myself, are also not exactly rosy-eyed optimists. An early 20th century German investor with a globally-diversified stock portfolio experienced a real drawdown lasting 57 years. That had nothing to do with valuations. That is one kind of outcome, among many, that I try to prepare for. Valuations don't really enter into that planning process because they have not been sufficiently reliable in practice and there are strong theoretical reasons to question their reliability.
He's an 80% VTSAX kind of person. Just enough to get a little dangerous and place him on Bogle's naughty list.
He can correct me if I'm wrong, but I thought he had about 40% of his portfolio in US stocks (because 50% is in bonds).

In any event, I have more international than that in my stock portion (80% global + 20% US tilt), but it has precisely nothing to do with valuations. There are other ways to determine an appropriate AA than looking at relative valuations.
Yes, that's true. 80% was for Homer's equity portion. So 50% Bonds, 40% VTSAX, 10% VTIAX

This type of portfolio is at very high risk of SWR failure in retirement from where we are right now if you think the past will still hold. I would certainly not bet on it. He'll need to hope US has a Japan style melt-up otherwise he could run into SORR issues due to over reliance on bonds and US mega cap equities. Or ensure a smaller SWR or more flexible SWR strategy.
Yes, some may prefer to mitigate risk by paying for it explicitly through increased savings and flexibility in spending. Others think that they will be able to avoid risk by being clever with tactical allocations. Different approaches.
Global Market Portfolio + modest tilt towards volatility (80/20->60/40 as approach FI) + modest tilt away from exchange rate risk (80% global+20% U.S. stocks; currency-hedge bonds) + tax optimization
ScubaHogg
Posts: 3573
Joined: Sun Nov 06, 2011 2:02 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by ScubaHogg »

Nathan Drake wrote: Thu Jul 22, 2021 3:45 pm
Everyone has a different threshold. To me a PE above 30 is giving me an estimated return of 3%, maybe less if valuations decrease, maybe more if they continue increasing. As PEs get higher and higher from long term trends, the less likely I am to factor in PE expansion so I will discount for reversion. That leaves an estimate of 1-3%. That’s quite low for having money at risk in equities.

So US stocks are even higher than this at 35-40. At these levels I’m unlikely to add new contributions and may start rebalancing/tilting towards other assets with lower valuations, but I won’t get out completely.

I imagine there would come a point where I reduce exposure completely in favor of other assets, PE50 or so.

Thank you. Its good to hear someone’s thresholds
“Conventional Treasury rates are risk free only in the sense that they guarantee nominal principal. But their real rate of return is uncertain until after the fact.” -Risk Less and Prosper
ScubaHogg
Posts: 3573
Joined: Sun Nov 06, 2011 2:02 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by ScubaHogg »

HomerJ wrote: Thu Jul 22, 2021 4:22 pm
ScubaHogg wrote: Thu Jul 22, 2021 2:52 pmTo use a little reductio ad absurdum, if bonds and what-not stayed roughly at their current yields, but the TSM shot up to a PE10 of 200, would you just say “nobody knows enough” and do nothing? Maybe you would. Heck maybe I would. But I don’t know that just the fact I might not do anything would make it a wise decision.
Since I rebalance a 50/50 portfolio, I would have locked in enough gains to retire comfortably even if TSM thereafter crashed 90% after rising to a PE of 200 :)

:moneybag :sharebeer
ScubaHogg wrote: Thu Jul 22, 2021 2:52 pm
And a PE of 25 shouldn’t seem extreme in the mid 90s when you could look across the pacific from a few years earlier and see a developed market with a PE almost 4x higher. Maybe 25 felt high, but “extreme” would have been hyperbolic)
No, it was considered extreme. Every time CAPE crossed 20 in the past, a crash occurred soon after.

It had only gone higher than 25 once, and that was in 1929, right before the Great Depression (where stocks ultimately dropped 88% at one point from their 1929 highs).

Crossing 25 was HUGE. It hadn't been that high in nearly 70 years. Irrational Exuberance speech by the Fed Chairman himself.

(Imagine what would happen today if the Fed Chairman, at the next televised meeting, said "Stocks prices are crazy high, ya'll! I mean stupid high!")

But actually 1996 ended up a good time to buy. Which is the crazy part... It wasn't that 1996 turned to be not as bad as predicted. It was actually a GOOD time to buy. Highest valuations in 70 years, and it was a good time to buy. Even during the crash of 2000-2003, the market never dropped as low as 1996.

The model was invented in 1988 using 1876-1988 data, and less than 10 years later it utterly failed as a prediction tool. And again and again and again in the following years.

In 2011, CAPE predicted 4.5% real 10-year returns... Instead we've gotten 13% real. That's not just a little bit off. That's wildly wrong. 99% percentile wrong.

Maybe we keep getting lucky. Or maybe, just maybe... the model is missing some important variables. Maybe it's not easy to predict the future.
See, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.

At a PE of 200 you are talking about a 90% drop to bring things back to historical norms. Or conversely it doesn’t drop, and you just dribble out a tiny return from earnings growth while accepting the full risk of equities.

These are qualitatively different things
“Conventional Treasury rates are risk free only in the sense that they guarantee nominal principal. But their real rate of return is uncertain until after the fact.” -Risk Less and Prosper
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
BJJ_GUY
Posts: 882
Joined: Wed Mar 13, 2019 7:45 am

Re: Do Indexers even need to pay attention to market valuation?

Post by BJJ_GUY »

HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
Try using a different method. CAPE ratio is more flawed than others. The inverted CAPE to estimate returns is not great, and the newer version ECY (excess cape yield) is even worse because the assumptions for the inflation adjustment are highly flawed.
am
Posts: 4233
Joined: Sun Sep 30, 2007 9:55 am

Re: Do Indexers even need to pay attention to market valuation?

Post by am »

HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
+1. Another way of saying that no one knows and stay the course.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

am wrote: Fri Jul 23, 2021 8:18 am
HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
+1. Another way of saying that no one knows and stay the course.
Nobody knows nothin!

*final last words of Japanese meme stock buyer in 1980s*
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

BJJ_GUY wrote: Fri Jul 23, 2021 2:15 am
HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
Try using a different method. CAPE ratio is more flawed than others. The inverted CAPE to estimate returns is not great, and the newer version ECY (excess cape yield) is even worse because the assumptions for the inflation adjustment are highly flawed.
Umm.. that's exactly my point. As new annual return data fails to fit an existing model, the PhDs just data-mine all the data again (including the new recent stuff that didn't work) until they find a new method/model that fits the new data.

And then write papers showing if you had used the new model 30 years ago you would have done really well timing the market (Well, duh... it's a data-mined model DERIVED from the past 30 years of data)

The new model didn't exist 30 years ago, so one couldn't have used it. They were using a completely different model then that failed terribly going forward. That's why they had to come up with a new method/model.

That was my point.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
Triple digit golfer
Posts: 10433
Joined: Mon May 18, 2009 5:57 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by Triple digit golfer »

No, and in fact, nobody should pay attention to market valuation unless they're trying to sell a product. For investing purposes, it's utterly meaningless.
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

Nathan Drake wrote: Fri Jul 23, 2021 8:46 am
am wrote: Fri Jul 23, 2021 8:18 am +1. Another way of saying that no one knows and stay the course.
Nobody knows nothin!

*final last words of Japanese meme stock buyer in 1980s*
Sure. But if tactical asset allocation is a reasonably good/proven strategy prospectively, why aren't there many mutual funds that try it? Vanguard had a few and they folded. Seems hard to put into practice usefully.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Fri Jul 23, 2021 11:37 am
Nathan Drake wrote: Fri Jul 23, 2021 8:46 am
am wrote: Fri Jul 23, 2021 8:18 am +1. Another way of saying that no one knows and stay the course.
Nobody knows nothin!

*final last words of Japanese meme stock buyer in 1980s*
Sure. But if tactical asset allocation is a reasonably good/proven strategy prospectively, why aren't there many mutual funds that try it? Vanguard had a few and they folded. Seems hard to put into practice usefully.
Because as with any strategy of the sort, it may take 10 years or more to start realizing that outsized performance

But to me it’s not necessarily about outsized performance per se as it is about protecting downside risk

I’m generally an optimist, and if a huge collection of countries has been beaten down with bad returns and low valuations I don’t expect that to persist over the long term. And as such I don’t feel there’s much downside if pessimism is already high. Maybe returns won’t be robust, but they’re unlikely to be poor

On the contrary, a highly valued country is more prone to sudden and large corrections should the expectations fail to materialize
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

Nathan Drake wrote: Fri Jul 23, 2021 11:50 am
Da5id wrote: Fri Jul 23, 2021 11:37 am
Nathan Drake wrote: Fri Jul 23, 2021 8:46 am
am wrote: Fri Jul 23, 2021 8:18 am +1. Another way of saying that no one knows and stay the course.
Nobody knows nothin!

*final last words of Japanese meme stock buyer in 1980s*
Sure. But if tactical asset allocation is a reasonably good/proven strategy prospectively, why aren't there many mutual funds that try it? Vanguard had a few and they folded. Seems hard to put into practice usefully.
Because as with any strategy of the sort, it may take 10 years or more to start realizing that outsized performance

But to me it’s not necessarily about outsized performance per se as it is about protecting downside risk

I’m generally an optimist, and if a huge collection of countries has been beaten down with bad returns and low valuations I don’t expect that to persist over the long term. And as such I don’t feel there’s much downside if pessimism is already high. Maybe returns won’t be robust, but they’re unlikely to be poor

On the contrary, a highly valued country is more prone to sudden and large corrections should the expectations fail to materialize
You are arguing for diversification across countries. That is a separate argument from being valuation aware or tactically allocating IMO. Seems like none of the big balanced or target date funds choose to be valuation aware and dynamically allocate their assets. Perhaps that is because it is not historically at all reliable.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Fri Jul 23, 2021 1:12 pm
Nathan Drake wrote: Fri Jul 23, 2021 11:50 am
Da5id wrote: Fri Jul 23, 2021 11:37 am
Nathan Drake wrote: Fri Jul 23, 2021 8:46 am
am wrote: Fri Jul 23, 2021 8:18 am +1. Another way of saying that no one knows and stay the course.
Nobody knows nothin!

*final last words of Japanese meme stock buyer in 1980s*
Sure. But if tactical asset allocation is a reasonably good/proven strategy prospectively, why aren't there many mutual funds that try it? Vanguard had a few and they folded. Seems hard to put into practice usefully.
Because as with any strategy of the sort, it may take 10 years or more to start realizing that outsized performance

But to me it’s not necessarily about outsized performance per se as it is about protecting downside risk

I’m generally an optimist, and if a huge collection of countries has been beaten down with bad returns and low valuations I don’t expect that to persist over the long term. And as such I don’t feel there’s much downside if pessimism is already high. Maybe returns won’t be robust, but they’re unlikely to be poor

On the contrary, a highly valued country is more prone to sudden and large corrections should the expectations fail to materialize
You are arguing for diversification across countries. That is a separate argument from being valuation aware or tactically allocating IMO. Seems like none of the big balanced or target date funds choose to be valuation aware and dynamically allocate their assets. Perhaps that is because it is not historically at all reliable.
Im arguing for my own personal allocation of overweighting regions with long term underperformance and lower valuations, particularly when spreads are substantial relative to historical norms

Target date funds are a fine choice but they also can’t stray too much from global cap weight for various institutional regions. It’s not really about long-term performance not being reliable, but the consumers of such funds may get antsy over short term underperformance to a given benchmark. Target date funds also have zero exposure to things like small cap value with higher expected returns
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
ScubaHogg
Posts: 3573
Joined: Sun Nov 06, 2011 2:02 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by ScubaHogg »

HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
Just to be clear, I’m not predicting anything. I’m simple stating how many dollars one must spend to “buy” a dollar if earnings. It’s not a model insomuch as it is just logic
“Conventional Treasury rates are risk free only in the sense that they guarantee nominal principal. But their real rate of return is uncertain until after the fact.” -Risk Less and Prosper
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

ScubaHogg wrote: Fri Jul 23, 2021 5:57 pm
HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
Just to be clear, I’m not predicting anything. I’m simple stating how many dollars one must spend to “buy” a dollar if earnings. It’s not a model insomuch as it is just logic
That's not "yield". You are mixing up your terms.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

Nathan Drake wrote: Fri Jul 23, 2021 4:22 pm
Da5id wrote: Fri Jul 23, 2021 1:12 pm You are arguing for diversification across countries. That is a separate argument from being valuation aware or tactically allocating IMO. Seems like none of the big balanced or target date funds choose to be valuation aware and dynamically allocate their assets. Perhaps that is because it is not historically at all reliable.
Im arguing for my own personal allocation of overweighting regions with long term underperformance and lower valuations, particularly when spreads are substantial relative to historical norms

Target date funds are a fine choice but they also can’t stray too much from global cap weight for various institutional regions. It’s not really about long-term performance not being reliable, but the consumers of such funds may get antsy over short term underperformance to a given benchmark. Target date funds also have zero exposure to things like small cap value with higher expected returns
OK. But as far as I'm aware historically (non back tested) strategies to use valuations tactically/strategically were unsuccessful. Do you disagree with the claim that using CAPE or other valuation measures to set asset allocations wasn't a win? If so why you think strongly that it is different now?
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Fri Jul 23, 2021 11:02 pm
Nathan Drake wrote: Fri Jul 23, 2021 4:22 pm
Da5id wrote: Fri Jul 23, 2021 1:12 pm You are arguing for diversification across countries. That is a separate argument from being valuation aware or tactically allocating IMO. Seems like none of the big balanced or target date funds choose to be valuation aware and dynamically allocate their assets. Perhaps that is because it is not historically at all reliable.
Im arguing for my own personal allocation of overweighting regions with long term underperformance and lower valuations, particularly when spreads are substantial relative to historical norms

Target date funds are a fine choice but they also can’t stray too much from global cap weight for various institutional regions. It’s not really about long-term performance not being reliable, but the consumers of such funds may get antsy over short term underperformance to a given benchmark. Target date funds also have zero exposure to things like small cap value with higher expected returns
OK. But as far as I'm aware historically (non back tested) strategies to use valuations tactically/strategically were unsuccessful. Do you disagree with the claim that using CAPE or other valuation measures to set asset allocations wasn't a win? If so why you think strongly that it is different now?
I haven’t seen any studies that indicate when X market gets valued at Y levels of standard deviation away from market Z it was a losing strategy over the next twenty years.

In fact, the “value” premium shows fairly strong evidence for its long term existence, which is more nuanced strategy on CAPE timing.

It’s a winning strategy but it can require long patience for the outperformance to play out.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

Nathan Drake wrote: Fri Jul 23, 2021 11:12 pm
Da5id wrote: Fri Jul 23, 2021 11:02 pm
Nathan Drake wrote: Fri Jul 23, 2021 4:22 pm
Da5id wrote: Fri Jul 23, 2021 1:12 pm You are arguing for diversification across countries. That is a separate argument from being valuation aware or tactically allocating IMO. Seems like none of the big balanced or target date funds choose to be valuation aware and dynamically allocate their assets. Perhaps that is because it is not historically at all reliable.
Im arguing for my own personal allocation of overweighting regions with long term underperformance and lower valuations, particularly when spreads are substantial relative to historical norms

Target date funds are a fine choice but they also can’t stray too much from global cap weight for various institutional regions. It’s not really about long-term performance not being reliable, but the consumers of such funds may get antsy over short term underperformance to a given benchmark. Target date funds also have zero exposure to things like small cap value with higher expected returns
OK. But as far as I'm aware historically (non back tested) strategies to use valuations tactically/strategically were unsuccessful. Do you disagree with the claim that using CAPE or other valuation measures to set asset allocations wasn't a win? If so why you think strongly that it is different now?
I haven’t seen any studies that indicate when X market gets valued at Y levels of standard deviation away from market Z it was a losing strategy over the next twenty years.

In fact, the “value” premium shows fairly strong evidence for its long term existence, which is more nuanced strategy on CAPE timing.

It’s a winning strategy but it can require long patience for the outperformance to play out.
Wait I thought this is all about valuations like CAPE or P/E. And using them tactically or dynamically. Now it's just single factor (value) plus beta investing?
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Fri Jul 23, 2021 11:16 pm
Nathan Drake wrote: Fri Jul 23, 2021 11:12 pm
Da5id wrote: Fri Jul 23, 2021 11:02 pm
Nathan Drake wrote: Fri Jul 23, 2021 4:22 pm
Da5id wrote: Fri Jul 23, 2021 1:12 pm You are arguing for diversification across countries. That is a separate argument from being valuation aware or tactically allocating IMO. Seems like none of the big balanced or target date funds choose to be valuation aware and dynamically allocate their assets. Perhaps that is because it is not historically at all reliable.
Im arguing for my own personal allocation of overweighting regions with long term underperformance and lower valuations, particularly when spreads are substantial relative to historical norms

Target date funds are a fine choice but they also can’t stray too much from global cap weight for various institutional regions. It’s not really about long-term performance not being reliable, but the consumers of such funds may get antsy over short term underperformance to a given benchmark. Target date funds also have zero exposure to things like small cap value with higher expected returns
OK. But as far as I'm aware historically (non back tested) strategies to use valuations tactically/strategically were unsuccessful. Do you disagree with the claim that using CAPE or other valuation measures to set asset allocations wasn't a win? If so why you think strongly that it is different now?
I haven’t seen any studies that indicate when X market gets valued at Y levels of standard deviation away from market Z it was a losing strategy over the next twenty years.

In fact, the “value” premium shows fairly strong evidence for its long term existence, which is more nuanced strategy on CAPE timing.

It’s a winning strategy but it can require long patience for the outperformance to play out.
Wait I thought this is all about valuations like CAPE or P/E. And using them tactically or dynamically. Now it's just single factor (value) plus beta investing?
Value investing is a similar concept to observing CAPE with defined measures of value for specific basket of stocks within a broad market.

CAPE is looking at the total of a market relative to another. So similar concepts, not exactly the same.

Regardless, the idea of buying cheaper stocks (higher discount rates) leading to increased risk/return is a fairly universal concept. Markets can stay irrational for long periods but eventually reality will hit.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Do Indexers even need to pay attention to market valuation?

Post by YRT70 »

Nathan Drake wrote: Fri Jul 23, 2021 11:30 pm Value investing is a similar concept to observing CAPE with defined measures of value for specific basket of stocks within a broad market.

CAPE is looking at the total of a market relative to another. So similar concepts, not exactly the same.

Regardless, the idea of buying cheaper stocks (higher discount rates) leading to increased risk/return is a fairly universal concept. Markets can stay irrational for long periods but eventually reality will hit.
That's certainly possible, and it's also possible that the US market could continue to beat CAPE based estimates.

And it's also possible the value premium may have disappeared after publication.
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

YRT70 wrote: Sat Jul 24, 2021 2:58 am
Nathan Drake wrote: Fri Jul 23, 2021 11:30 pm Value investing is a similar concept to observing CAPE with defined measures of value for specific basket of stocks within a broad market.

CAPE is looking at the total of a market relative to another. So similar concepts, not exactly the same.

Regardless, the idea of buying cheaper stocks (higher discount rates) leading to increased risk/return is a fairly universal concept. Markets can stay irrational for long periods but eventually reality will hit.
That's certainly possible, and it's also possible that the US market could continue to beat CAPE based estimates.

And it's also possible the value premium may have disappeared after publication.
Meh. I don't choose to tilt, but I see the appeal I guess.

I don't consider long term tilting to value (or more commonly SCV) in the same vein as tactically shifting ones investments in response to relative CAPEs of different segments of the domestic US market or CAPEs of other nations markets. *That* is what I'm saying is historically an unproven strategy. You maybe can data-mine ways to make it work, but tactical allocation hasn't succeeded by any historical measures using data at the time, and has been a commercial failure.

SCV is also wrong for many (most?) people because, even if the theory holds in the future (plausible, but not guaranteed) it requires significant commitment and ability to persevere through very long periods of underperformance.
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Do Indexers even need to pay attention to market valuation?

Post by YRT70 »

Da5id wrote: Sat Jul 24, 2021 8:00 am
YRT70 wrote: Sat Jul 24, 2021 2:58 am
Nathan Drake wrote: Fri Jul 23, 2021 11:30 pm Value investing is a similar concept to observing CAPE with defined measures of value for specific basket of stocks within a broad market.

CAPE is looking at the total of a market relative to another. So similar concepts, not exactly the same.

Regardless, the idea of buying cheaper stocks (higher discount rates) leading to increased risk/return is a fairly universal concept. Markets can stay irrational for long periods but eventually reality will hit.
That's certainly possible, and it's also possible that the US market could continue to beat CAPE based estimates.

And it's also possible the value premium may have disappeared after publication.
Meh. I don't choose to tilt, but I see the appeal I guess.

I don't consider long term tilting to value (or more commonly SCV) in the same vein as tactically shifting ones investments in response to relative CAPEs of different segments of the domestic US market or CAPEs of other nations markets. *That* is what I'm saying is historically an unproven strategy. You maybe can data-mine ways to make it work, but tactical allocation hasn't succeeded by any historical measures using data at the time, and has been a commercial failure.

SCV is also wrong for many (most?) people because, even if the theory holds in the future (plausible, but not guaranteed) it requires significant commitment and ability to persevere through very long periods of underperformance.
I think you're responding to Nathan?

I wouldn't have mentioned value if he hadn't.
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

YRT70 wrote: Sat Jul 24, 2021 8:54 am
Da5id wrote: Sat Jul 24, 2021 8:00 am
YRT70 wrote: Sat Jul 24, 2021 2:58 am
Nathan Drake wrote: Fri Jul 23, 2021 11:30 pm Value investing is a similar concept to observing CAPE with defined measures of value for specific basket of stocks within a broad market.

CAPE is looking at the total of a market relative to another. So similar concepts, not exactly the same.

Regardless, the idea of buying cheaper stocks (higher discount rates) leading to increased risk/return is a fairly universal concept. Markets can stay irrational for long periods but eventually reality will hit.
That's certainly possible, and it's also possible that the US market could continue to beat CAPE based estimates.

And it's also possible the value premium may have disappeared after publication.
Meh. I don't choose to tilt, but I see the appeal I guess.

I don't consider long term tilting to value (or more commonly SCV) in the same vein as tactically shifting ones investments in response to relative CAPEs of different segments of the domestic US market or CAPEs of other nations markets. *That* is what I'm saying is historically an unproven strategy. You maybe can data-mine ways to make it work, but tactical allocation hasn't succeeded by any historical measures using data at the time, and has been a commercial failure.

SCV is also wrong for many (most?) people because, even if the theory holds in the future (plausible, but not guaranteed) it requires significant commitment and ability to persevere through very long periods of underperformance.
I think you're responding to Nathan?

I wouldn't have mentioned value if he hadn't.
I probably should have directly responded to his post rather than yours, yes.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Sat Jul 24, 2021 8:00 am
YRT70 wrote: Sat Jul 24, 2021 2:58 am
Nathan Drake wrote: Fri Jul 23, 2021 11:30 pm Value investing is a similar concept to observing CAPE with defined measures of value for specific basket of stocks within a broad market.

CAPE is looking at the total of a market relative to another. So similar concepts, not exactly the same.

Regardless, the idea of buying cheaper stocks (higher discount rates) leading to increased risk/return is a fairly universal concept. Markets can stay irrational for long periods but eventually reality will hit.
That's certainly possible, and it's also possible that the US market could continue to beat CAPE based estimates.

And it's also possible the value premium may have disappeared after publication.
Meh. I don't choose to tilt, but I see the appeal I guess.

I don't consider long term tilting to value (or more commonly SCV) in the same vein as tactically shifting ones investments in response to relative CAPEs of different segments of the domestic US market or CAPEs of other nations markets. *That* is what I'm saying is historically an unproven strategy. You maybe can data-mine ways to make it work, but tactical allocation hasn't succeeded by any historical measures using data at the time, and has been a commercial failure.

SCV is also wrong for many (most?) people because, even if the theory holds in the future (plausible, but not guaranteed) it requires significant commitment and ability to persevere through very long periods of underperformance.
Small Cap Value outperforms *most* of the time. Just like the equity premium outperforms *most* of the time. But you can experience long stretches where it doesn't, such as the one we are in now.

But you have to look at the data and see whether or not there was a fundamental reason for the underperformance or did the spreads widen? If the latter, I would expect mean reversion to eventually occur, so the odds are in your favor.

Same reason for CAPE or some general valuation model of regions investment tilting. Funds like these haven't worked out because investors are mostly concerned with 1,3,5 year performance periods and if a fund doesn't take off then it's doomed to failure.

Read Meb's article:
https://mebfaber.com/2019/01/06/you-wou ... ood-thing/
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
ScubaHogg
Posts: 3573
Joined: Sun Nov 06, 2011 2:02 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by ScubaHogg »

HomerJ wrote: Fri Jul 23, 2021 6:24 pm
ScubaHogg wrote: Fri Jul 23, 2021 5:57 pm
HomerJ wrote: Fri Jul 23, 2021 12:38 am
ScubaHogg wrote: Thu Jul 22, 2021 10:31 pmSee, we are talking about different things. In that model of “extreme” (once again, just a fraction of what just happened across the pacific) a 20% drop, approximately, would bring the market back to historical averages. And even at 25 you are talking about a 4% yield.
Just FYI, that's the new model. Back in the day, CAPE 25 meant 0% 10-year real return. Again, remember the Fed Chairman himself in 1996 gave a speech basically stating "This is crazy, ya'll. Stocks are insanely overpriced" (and then they more than doubled again before crashing only 40%)

NOW, it predicts 4%, because they've changed the model with all the new data. But they pretend the model always said 4% at CAPE 25.

It didn't.

I lived through all this. There was no 1/CAPE model until far later, when data-mining all the new data showed 1/CAPE makes a pretty good line.

If the data is different over the next 20 years, they will change the model again, and you might see some young people posting here how the CAPE model has ALWAYS predicted 6% returns for CAPE 25...

And you will scratch your head, and say "I could I have sworn that once CAPE 25 predicted 4%"

And the young people will tell you that you have no idea what you're talking about. "Look at this article from 2040 that plainly shows that CAPE 25 has always predicted 6% real returns..."
Just to be clear, I’m not predicting anything. I’m simple stating how many dollars one must spend to “buy” a dollar if earnings. It’s not a model insomuch as it is just logic
That's not "yield". You are mixing up your terms.
Fair. “Earnings Yield” then

https://www.investopedia.com/terms/e/earningsyield.asp
“Conventional Treasury rates are risk free only in the sense that they guarantee nominal principal. But their real rate of return is uncertain until after the fact.” -Risk Less and Prosper
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Do Indexers even need to pay attention to market valuation?

Post by Call_Me_Op »

alex_686 wrote: Sat Jul 17, 2021 8:18 pm What we can say is that long run expected returns will be lower in the future. So you should adjust your allocation according.
How do you do that when expected returns of other assets classes are also lower (compared to their respective historical averages)?
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

Nathan Drake wrote: Sat Jul 24, 2021 11:52 am Small Cap Value outperforms *most* of the time. Just like the equity premium outperforms *most* of the time. But you can experience long stretches where it doesn't, such as the one we are in now.

But you have to look at the data and see whether or not there was a fundamental reason for the underperformance or did the spreads widen? If the latter, I would expect mean reversion to eventually occur, so the odds are in your favor.

Same reason for CAPE or some general valuation model of regions investment tilting. Funds like these haven't worked out because investors are mostly concerned with 1,3,5 year performance periods and if a fund doesn't take off then it's doomed to failure.

Read Meb's article:
https://mebfaber.com/2019/01/06/you-wou ... ood-thing/
Doesn't address my point.

I looked at his article. I'm not interested in back-tested p-hacked allocation switching strategies. I want to see strategies that worked in out-of-sample data over long periods. So calls made in the past based on data available at the time to significantly tactically switch. And those generally have been poor. And even if they appear to work I'd need to see that they continue to work in the long haul out of sample, i.e. aren't luck. And heck, even then I'd probably not trust them to persist into the future.

Buying and sticking with value or SCV or other factor based funds isn't the above. To buy and hold SCV, in the context of this thread, isn't to "pay attention to market valuation" as you can (and likely should) completely ignore valuation changes to execute that strategy.

Valuations do matter somewhat, I just continue question the ability to use them to make on the fly switches "paying attention to market valuation" in ones allocation in response. I'm OK with my 60/40 total US/total int'l. Not looking to get the best of it, just to match the markets. SCV is a reasonable strategy, just not one I've chosen to do.
Actin
Posts: 259
Joined: Wed May 16, 2018 8:40 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by Actin »

It makes no difference to me. I can't do anything about the scam companies like Tesla or the dozen of zombie companies.

Indexing is set it and forget it. If I have to put more thought into it than that, I'd go to different investment
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Sun Jul 25, 2021 8:53 am
Nathan Drake wrote: Sat Jul 24, 2021 11:52 am Small Cap Value outperforms *most* of the time. Just like the equity premium outperforms *most* of the time. But you can experience long stretches where it doesn't, such as the one we are in now.

But you have to look at the data and see whether or not there was a fundamental reason for the underperformance or did the spreads widen? If the latter, I would expect mean reversion to eventually occur, so the odds are in your favor.

Same reason for CAPE or some general valuation model of regions investment tilting. Funds like these haven't worked out because investors are mostly concerned with 1,3,5 year performance periods and if a fund doesn't take off then it's doomed to failure.

Read Meb's article:
https://mebfaber.com/2019/01/06/you-wou ... ood-thing/
Doesn't address my point.

I looked at his article. I'm not interested in back-tested p-hacked allocation switching strategies. I want to see strategies that worked in out-of-sample data over long periods. So calls made in the past based on data available at the time to significantly tactically switch. And those generally have been poor. And even if they appear to work I'd need to see that they continue to work in the long haul out of sample, i.e. aren't luck. And heck, even then I'd probably not trust them to persist into the future.

Buying and sticking with value or SCV or other factor based funds isn't the above. To buy and hold SCV, in the context of this thread, isn't to "pay attention to market valuation" as you can (and likely should) completely ignore valuation changes to execute that strategy.

Valuations do matter somewhat, I just continue question the ability to use them to make on the fly switches "paying attention to market valuation" in ones allocation in response. I'm OK with my 60/40 total US/total int'l. Not looking to get the best of it, just to match the markets. SCV is a reasonable strategy, just not one I've chosen to do.
Then stick to your strategy. I’m not saying MCW is bad at all. I think you can likely do better if there seem to be points where one asset class gets wildly out of sync with other asset classes.

My approach isn’t all or nothing switching. It’s as Cliff Asness would say “sinning a little”. I believe I will be rewarded given how wide the spreads are, you can disagree if you want. If I’m wrong, it’s doubtful there will be much harm.

What studies of tactical switching are you referring to? Why didn’t they work? It’s probably due to not staying the course long enough.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Da5id »

Nathan Drake wrote: Sun Jul 25, 2021 10:52 am My approach isn’t all or nothing switching. It’s as Cliff Asness would say “sinning a little”. I believe I will be rewarded given how wide the spreads are, you can disagree if you want. If I’m wrong, it’s doubtful there will be much harm.

What studies of tactical switching are you referring to? Why didn’t they work? It’s probably due to not staying the course long enough.
I have no studies in mind. Just the generic historical failures of peoples attempts at market timing. Of which this clearly is an instance.
Nathan Drake
Posts: 6234
Joined: Mon Apr 11, 2011 12:28 am

Re: Do Indexers even need to pay attention to market valuation?

Post by Nathan Drake »

Da5id wrote: Sun Jul 25, 2021 10:58 am
Nathan Drake wrote: Sun Jul 25, 2021 10:52 am My approach isn’t all or nothing switching. It’s as Cliff Asness would say “sinning a little”. I believe I will be rewarded given how wide the spreads are, you can disagree if you want. If I’m wrong, it’s doubtful there will be much harm.

What studies of tactical switching are you referring to? Why didn’t they work? It’s probably due to not staying the course long enough.
I have no studies in mind. Just the generic historical failures of peoples attempts at market timing. Of which this clearly is an instance.
Market timing is generally short term trading. I have no delusions that I can predict what the market is going to do in the short term or the long term. I can try and put the odds in my favor, but I agree there is no guarantee I will be correct. Historically you can see countless examples of one country having a humongous gap in P/E only to be an awful investment in the future relative to assets that haven’t done as well recently.

I am more comfortable limiting my exposure a bit when that occurs.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Do Indexers even need to pay attention to market valuation?

Post by HomerJ »

Nathan Drake wrote: Sun Jul 25, 2021 10:52 am I believe I will be rewarded given how wide the spreads are, you can disagree if you want. If I’m wrong, it’s doubtful there will be much harm.
What percentage of International are you now?

You went from 50% to 60% 7 years ago based on valuations differences... Since valuations spreads have only gotten larger... Have you increased your International since then?
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
Post Reply