International ETF's

VXUS is total international ETF
VWO is emerging markets

Japan won't be in emerging, so you'd have to use something like VEA with VWO if you splice and dice

Oh, and I would definitely be more keen to invest in India than Japan. INDA looks like it's only appreciated by about 33% over the last 5 years, and doing something like EWJ for Japan seems too dangerous to me.

India has an economy growing faster than China's, and will have the largest population in the world in a few decades. Japan, by contrast, is an extremely low growth economy with an elderly society on its last legs. Moreover, Japan is entering an experimental zone with monetary policy, and nobody knows how it will end up. It could get very, very bad over there.

Schwab has a good one.

I've been overweight on India for several years. PIN.

I recently liquidated some emerging markets and moved it to Japan. HJPX. I did so after reading some analysis put out by Stringer Asset Management.

Vanguard VEA and VWO (Emerging and Developed Markets) ETF's have been doing well this year. I own both and they have done a few % better than all my US investments

For the sake of efficiency, allow me to use the made-up term "overdispersed" for Hussman's notion of unfavorable market internals with high dispersion, which he believes is a bearish indicator.

quote:

---

a counter hypothesis could be that bond yield to stock ratios indicate that, in terms of that trade off, stock are properly valued to what corporate bonds are yielding.

---

I don't think that Hussman's metrics for overdispersion relate to the bond yield relative to stock ratios. Indeed, he makes the point repeatedly that interest rates have very little to do with projecting future returns on equities.

When market internals are overdispersed, that doesn't mean that bond yields are low relative to stocks. Rather, it means (if I'm understanding him correctly) that low-grade bonds are not well differentiated from high-grade bonds, that covenant-light bonds are not well differentiated from covenant-heavy bonds, etc.

It's hard to know the exact concrete details of what goes into his metrics overdispersion, because it's proprietary, and he only discusses it generally, but I am guessing that it also takes into account dispersion within a single asset class (and not just bonds). So if all junk bonds (or S&P 500 stocks) move in tandem, then there would not be much intra-class dispersion. But if 15% of junk bonds (or S&P 500 stocks) suddenly drop below or rise below the rest of the pack, then you have greater dispersion--i.e., greater risk discrimination by investors, who are being careful about researching the specifics of individual securities, rather than simply passively investing in certain classes of assets.

quote:

---

Hussman stated that quantitative easing was a reason for why his valuations were not completely appropriate, I think in the 2009 time frame.

---

No. I think I may have been trying to explain too much at one time with my earlier post, and didn't communicate these concepts as well as I could have done.

Hussman's valuations have never been wrong. That's his bread-and-butter stuff, the statistic (non-financial market capitalization / corporate gross-value-added) that models nominal returns over the next 12 years. I'm pretty sure that whatever his statistic was for 2005 has predicted returns to the present day fairly well.

But like I wrote in my earlier post, "a market being overvalued doesn't tell you very much about what it's going to do over the next 1-2 years." But Hussman runs money with actual funds (he now does it as non-profits), so he has to attempt to give shorter-term market outlooks than 10-12-year projections, and that's where things get difficult.

(By the way, he didn't really "call a market crash" in 2014; that was just me being lazy with my words and using sloppy language to convey that he had a "hard negative" market outlook at that time. As he's often explained, when he gives negative market outlooks, that doesn't really predict investor returns over the next day, or even the next 2-3 years. It only means that the probability distribution for daily returns starts to have very fat left-tail events in it, and that over the completion of a 12-year cycle, valuations look poor.)

Prior to 2014, this (in general terms, since a lot of the details are proprietary secrets) is how Hussman came up with his negative market outloooks:

overvalued + overbought + overbullish = hard negative

There was nothing at all wrong with the "overvalued" part. As far we know, the 10-12-year projections from 2014 will prove just as correct in the year 2026 as they always have. I doubt the "overbullish" was the problem either. It was the "overbought" part that seems to me to have been the source of his problems... relying on too much T.A. mumbo jumbo about Bollinger bands and candlestick charts and shapes and the like. (I don't even know what particular metrics he uses for his technical analysis, because he never really discusses it in any detail... other than to say that it used to work for Depression Era historicals, but doesn't work in a near-ZIRP environment.)

So after his blunder in 2014, he came back with new requirements before taking a hard negative market outlook. The formula now looks like this:

overvalued + overbought + overbullish + overdispersed = hard negative

And, really, like I mentioned above, I think the "overbought" stuff is mostly mumbo jumbo, so he might as well use the following:

overvalued + overdispersed + overbullish = hard negative

That probably does a good job at informing investors regarding what percentage of their assets should be allocated to U.S. equities, based upon their own particular situations and preferences. In theory, there is no lower or upper limit to this stuff. It can go into negative territory if you're short, and if you are leveraged long (which Hussman has recommended at certain points in time), then you can push beyond the 100% barrier.

Hope that helps. This stuff can be confusing at first, like stepping into a whole other universe.

For the most part. I would use more careful language, and say that when assets within a class are moving all over the place, that's typically a precondition for a big drop (or at least greatly increases the probability of left-tail events).

If you're holding a security over a 10-12-year period, of course, then it doesn't matter whether there is overdispersion--only the valuation metrics matter.

Going by the famous Shiller CAPE (use the link in the 2nd paragraph at Shiller's Online Data to download his Excel file), it's now at about 30.49, just a tad below the September 1929 peak of 32.56. Now some might say, "Well it might still have room to run, because it got even higher in December 1999 when it hit 44.20."

However, that doesn't take into account several factors that can improve upon the Shiller CAPE metric, and under the improved metric from Hussman, things look much, much worse. Hussman's metric takes into account the corporate profit levels, which was famously noted to be "the most mean-reverting series in finance" by Jeremy Grantham in November 2014. The following graph from his Monday, July 24, weekly report tells you all you need to know:



We are now (as of 7/31/17) at 178% above the norm, and completions of the market cycle have always brought the metric under 50% above the norm... so there's a lot of room to drop. As Hussman wrote on 7/24/17...

quote:

---

On that note, it’s worth noting that the Shiller CAPE reached 30 last week, which places current valuations among the highest 3% of historical observations, on that particular measure. In reality, however, the situation is much more extreme. While some observers take solace in the fact that the CAPE reached a much higher level of 44 at the 2000 bubble peak, that historical comparison ignores the embedded margin. See, at the 2000 peak, the embedded margin of the CAPE was just 5.1%, compared with a historical norm of 5.4%. At present, because of depressed wages and high profit margins in recent years (which are now reversing given a 4.4% unemployment rate), the CAPE quietly embeds a profit margin of 7.2%. Put another way, investors are currently paying a very high multiple of a very high earnings number that quietly benefits from cyclically elevated profit margins. On the basis of normalized profit margins (an adjustment that improves historical reliability), the normalized CAPE would presently be 40. There is only a single week in history where the normalized CAPE was higher. That was the week of March 24, 2000, when the S&P 500 hit its final bubble peak, at a normalized CAPE of 41.

---

Caveat emptor.

Since I sold my last shares of Mexican stocks on July 31.

I have been completely out of U.S. equities since 2015.

My plan is to let the S&P 500 make its drop, then wait for the 50d SMA to cross back above the 200d SMA (don't try to get the bottom timing exactly correct and risk catching a falling knife), and then leverage up to the gills with something like TNA for the long-haul. Hopefully that will allow me to sit back and leave all this market watching crap alone for a very long time (i.e., around 10 years or so).