Page 1 of 2

Re: Golden Butterfly rebalancing bands

Posted: Mon Apr 06, 2020 9:23 pm
by CJ043332
pmward wrote:
Mon Apr 06, 2020 8:17 pm
CJ043332 wrote:
Mon Apr 06, 2020 8:05 pm
pmward wrote:
Mon Apr 06, 2020 7:10 pm
CJ043332 wrote:
Mon Apr 06, 2020 7:05 pm
pmward wrote:
Mon Apr 06, 2020 6:51 pm
CJ043332 wrote:
Mon Apr 06, 2020 6:28 pm
Here is a very rough example:

Overpriced market:

25% S&P 500
25% Tbills
25% Long Term Treasuries
25% Gold

Fairly priced market:

20% S&P 500
12.5% Small Cap Value
22.5% Tbills
22.5% Long Term Treasuries
22.5% Gold

Underpriced market:

25% Small Cap Value
15% S&P 500
20% Tbills
20% Long Term Treasuries
20% Gold

You can see that the allocations adjust based on how "cheap" the overall stock market it. As it heats up, more automatically gets allocated to the other asset classes as insurance since the risk is higher. Nobody knows for sure what will happen... but that doesn't mean that statistically the predictability is at zero either.
The problem, as it always is with valuation based investing, is that the market can stay irrational longer than you can stay solvent. There is also the problem of "overpriced", "underpriced", and "fairly priced" being completely arbitrary. Over the years I have seen the "fairly priced" marker get quoted over a wide range of numbers. Picking one fundamental metric is arbitrary, and all fundamental metrics are relative. They also place no emphasis on timing at all. I mean Warren Buffet has underperformed the S&P 500 for what... 17 years now? Can you really handle that kind of underperformance? Is that even acceptable? Every fund manager alive would be fired for that kind of performance, yet WB is still looked at as a hero/genius/etc. If someone wants to be tactical in their allocations then they need some kind of a technical or quantitative basis behind it. Otherwise, people are just going to wind up juggling their portfolio based purely on feelings, and that tends to get people in to trouble. I would just be careful with trying to time the market based on your perception of what the market valuation is. Valuations tend to only really matter over very long multi-decade time frames. You're likely to get chopped up and capitulate long before you reap any valuation premium.


Any long term investor that invests in a PP or GB would needs to be able to stomach under-performance. Besides, at 40% equities it is unlikely you will go insolvent.

Furthermore, the CAPE was not chosen by random. A study by Vanguard showed that P/E 10 and PE/1 were basically the only metrics that amounted to anything. As stated above,
Shiller CAPE doesn't have a very strong predictability over a 10 year period... but it isn't exactly a very weak predictability either.
Are you attempting to state:

1. CAPE literally has a 0% predictability rate.

or

2. CAPE has some predictability rate, however that rate should be disregarded entirely for investing purposes even in the smallest of ways.
CAPE has 0 predictability when it comes to timing. Like I said, valuation premiums are a multi-decade kind of thing. Most investors do not go multiple decades without tinkering with their portfolio. I see no reason for a long term investor to skew their portfolio up or down based on CAPE. If someone wants a tactical portfolio allocation then it should be based using more technical or quantitative measures. Otherwise, it's probably best to just pick the portfolio allocation and rebalance over time, letting the rebalancing help you buy low and sell high.
A 2012 research study funded by Vanguard authored by Joseph Davis, Ph.D., Roger Aliaga-Diaz, Ph.D., and Charles J. Thomas, CFA appear to have concluded something quite differently. They did an analysis of several popular metrics and back-tested their predictability.

Do you have any research studies to support your claim of zero predictability over a 10 year period for CAPE?

Now, is there a chance that those Ph.D.'s got it wrong? Sure. Is it also possible that a GB is a worse allocation than a PP? Sure. Worst case scenario you end up with basically a PP or a GB. Back where you started.

What if they are right?

If I play roulette and 40% of the numbers are red and 60% are black (no green on this imaginary wheel)... my investment allocation should not be 50/50. Granted, if I was playing only 1 hand (i.e. a 1 year investment time-frame), I would be crazy to make a huge bet. But the longer you play, the more the true odds have a real impact. If you want to have the highest chance of breaking even over a long horizon (i.e. over a 10 year period), you better be sure you have a 40% red and 60% black allocation.

I would also point out that it appears that the vast majority of Permanent Portfolio investors have in fact adjusted their asset allocations to a GB allocation as equities (over the past 10 years) have become more expensive based on CAPE. Surely, you must see the irony in that.
I do not have any "PhD studies", and I personally do not care about these things as academics are always getting these things wrong. A PhD means absolutely nothing in the markets. Case in point, look at the famous Fama-French studies. They pretty much called the top on the "value premium" as soon as they published their research. The value factor has substantially underperformed for 18 years straight now. Small caps have also been much weaker, especially in the last 10 years. I would take these research studies with a grain of salt. Pretty much everyone who ever took Fama-French seriously has lost. CAPE metrics have been better for international than U.S. for over 10 years now, and the "undervalued" international stocks seem just find new ways to become more "undervalued". There are much better ways to be tactical than using valuation metrics.
I see... the CAPE is wrong AND the Ph.D's that backtested the CAPE at Vanguard are ALWAYS wrong... oh wait, but wouldn't your argument than make the CAPE right (since their argument was that the CAPE was not a predictor in the short term)? :o

I kid, but seriously...

TLDR, "You have no real evidence to support your claim of 0% predictability of CAPE, and dislike people with PhDs." :)

No judgement... I have no PhD, so I am not offended.

Re: Golden Butterfly rebalancing bands

Posted: Mon Apr 06, 2020 9:46 pm
by pmward
CJ043332 wrote:
Mon Apr 06, 2020 9:23 pm

I see... the CAPE is wrong AND the Ph.D's that backtested the CAPE at Vanguard are ALWAYS wrong... oh wait, but wouldn't your argument than make the CAPE right (since their argument was that the CAPE was not a predictor in the short term)? :o

I kid, but seriously...

TLDR, "You have no real evidence to support your claim of 0% predictability of CAPE, and dislike people with PhDs." :)

No judgement... I have no PhD, so I am not offended.
Haha, yeah I'm just a skeptic. While I do tactical investing, I use multiple indicators so I don't put all my eggs in one basket, and I do not use valuation metrics at all. I mean, who even knows what valuations really are? S&P 500 aggregate corporate profits peaked all the way back in 2012, but earnings per share kept going up. How does this happen? Financial engineering. It's companies levering up to do buybacks. I would say the market is more "overvalued" than it appears for this reason.

Re: Golden Butterfly rebalancing bands

Posted: Mon Apr 06, 2020 11:23 pm
by CJ043332
pmward wrote:
Mon Apr 06, 2020 9:46 pm
CJ043332 wrote:
Mon Apr 06, 2020 9:23 pm

I see... the CAPE is wrong AND the Ph.D's that backtested the CAPE at Vanguard are ALWAYS wrong... oh wait, but wouldn't your argument than make the CAPE right (since their argument was that the CAPE was not a predictor in the short term)? :o

I kid, but seriously...

TLDR, "You have no real evidence to support your claim of 0% predictability of CAPE, and dislike people with PhDs." :)

No judgement... I have no PhD, so I am not offended.
Haha, yeah I'm just a skeptic. While I do tactical investing, I use multiple indicators so I don't put all my eggs in one basket, and I do not use valuation metrics at all. I mean, who even knows what valuations really are? S&P 500 aggregate corporate profits peaked all the way back in 2012, but earnings per share kept going up. How does this happen? Financial engineering. It's companies levering up to do buybacks. I would say the market is more "overvalued" than it appears for this reason.
I hear ya. I definitely think that we have really drifted from fundamentals. Fundamentals can really be distorted by psychology and manipulation.

The only metric that is always guaranteed is that almost everyone is chasing a quick buck, making short-term movements impossible to predict.

That Vanguard study showed every single indicator had basically no predictability over 1 year and over 85% of them had practically no correlation over 10 years. The remaining were still not "very strong". So you definitely don't seem wrong in your views.

I view the CAPE as a drunk weatherman that is pretty good at saying it will snow in winter and rain in the spring. He has no clue what the weather will be tomorrow or the next day. In fact, it might not even snow at all during the winter. So although I am not going to migrate to Florida, it doesn't really hurt me to put a snow brush in my car just in case. Just balancing that little extra risk with a little extra preparedness.

If you think about it, it is really logical. As long as some people pay attention to P/E (which any Graham/Buffett follower would do), P/E will continue to at least be a whisper of reason in equity prices. After-all, lot's of people in the world still see a stock with negative P/E as bad luck :) Their choosing not to buy that stock still affects the price, even if others don't care about it. It isn't that fundamentals have NO SAY, it is that they are drowned out by all the other factors. However, those fundamentals are always there acting as a drag to some extent (even if it is a small drag when faced with massive fed intervention and speculation). That's also what the data seems to reflect.

I think it would also explain how both you and those PhD's can both be right in your views :)