Questions about PP from Japan

[MediumTex]

I'm sorry, I feel this has gotten too much into a discussion about general PP theory. Could I ask you to keep it on topic related to how it directly affects individual investors in a Japanese context? Thank you.

I think that part of the solution is going to be holding 50% 10 year Japanese bonds. 

If 25% is in gold or a gold fund, that just leaves the stock piece to sort out.

[stone]

mlyazaki, I totally agree that Clive's objections apply no more to a Japanese PP than they would to any other PP. I don't see why Japan is any different from say the UK (where I live) in terms of being suitable for the PP. I think as Medium Tex says just have 50% 10yearJGB, 25% gold and 25% stocks. I don't see what there is to suggest that at this point Japanese stocks are not sensible. Over the past few years when converted back into Yen, wouldn't Japanese stocks be just as good/bad as say US stocks?

Clive, the crucial point though is that gold is not just any other priced entity. It is a widely accepted, timeless, global, alternative store of value with a huge liquid pool stored up. Its price responds to fiat currency problems in a way that underpins the HBPP.
Under a gold standard system, supply issues could cause inflation such that a pint of milk or whatever would cost you more USD or gold than it did the year before. Silver just did its wacky thing, with the silver price getting flung to and fro in a way that did not link in with whether or not your pint of milk became more expensive. There are fundamental reasons behind why silver has its type of price movements and gold has gold's. Since we are now on a fiat system, the silver price movements get sort of superimposed upon the shifts moving the gold price but I think that is totally different from imagining that silver behaved under a gold standard like gold behaves now.

[MachineGhost]

I'm sorry, I feel this has gotten too much into a discussion about general PP theory. Could I ask you to keep it on topic related to how it directly affects individual investors in a Japanese context? Thank you.

I think that part of the solution is going to be holding 50% 10 year Japanese bonds.  

If 25% is in gold or a gold fund, that just leaves the stock piece to sort out.

I concur.  50% 10-year JGB's should have a duration approaching that of 25% cash and 25% 30-year LT bonds.  I would rollover each year.

The stocks should be be 50% in small cap Japanese dividend-paying stocks and 50% in a cap-weighted world index ex-Japan, preferably hedged against Yen exposure.

[miyazaki]

The thing is, the bonds and cash portion of a Japanese PP are very different from any of the English speaking country's one. I happen to know that in the UK, you can get around 3% interest just from shopping around for a good savings account, whereas in Japan, 0.1% is advertised as a "special high rate savings account", with most accounts being much lower. We have access to inflation rates, but whether those are to be trusted is up for debate. Therefore, we cannot accurately say whether those interest rates are in fact equal in real terms or not... Anyway, I guess it would be psychologically more satisfying to see your cash and bonds rising in numerical terms automatically, rather than having them stay flat. You might say it's foolish, but these kind of psychological factors are what keep people doing irrational things with their money.

As for stocks, there are plenty of index funds for those. That's perhaps the easiest part!

In general, I really strongly agree with the advice being given on this board and in the book. There are just so many unknowns in the world, that we cannot prepare for. Japan may have a fantastic recovery in the next 20 years. It may also default on it's debts. We never know. That's why I started researching the PP more in detail, and I feel questioning it specifically in a Japan context is necessary. If it turns out that it is indeed exactly the same as other countries, then that's great! But we need to question it in order to find out.

[stone]

For all of those saying don't buy Japanese stocks, here is a different view: http://www.investingdaily.com/14541/war ... n-60-years

mlyazaki, if you had a GBP cash account paying 3% per year or whatever and had reinvested that interest over the past decade, then I guess it still would not have kept up with the Yen account paying 0.1%. Also at crucial times such as 2008, Yen would have been the best possible currency for giving you the purchasing power to load up with stocks and gold whilst they were good value. Yen has a very attractive history of getting stronger at just the right time when global assets plunge.

[Arturo]

The US/UK did so more gradually and are only now just getting down to where Japan was at in the 1990's. Japan's past 20 years could be an indicator of what the US/UK has yet to endure. Assuming a 20 year phase shift between Japan and US/UK, perhaps a combination of both sets of PP results might be indicative of an overall full cycle PP average.

So you are proposing that in the future (the time that could happen is unknown), UK, USA or EU will have to experience this long deflation stagnation, so PP will get balanced to the values you said (PP return is the combination of Japan and USA).

[Dets]

For those who have looked at this in detail, if we were to include the increase in global purchasing power due to yen appreciation, would that improve the true real return to something a little closer to what you would have earned in a UK or US PP?  Presumably the cost of anything you're importing has dropped, so your true gain isn't really captured by just your domestic Real Return.  Or am I missing something? 

[Gosso]

For those who have looked at this in detail, if we were to include the increase in global purchasing power due to yen appreciation, would that improve the true real return to something a little closer to what you would have earned in a UK or US PP?  Presumably the cost of anything you're importing has dropped, so your true gain isn't really captured by just your domestic Real Return.  Or am I missing something?  

Hello and welcome to the world of tomorrow PP forum!

If the investor stays in Japan then the real return is all that matters (since the inflation rate would factor in the lower cost of imports).  However, if they move to (or visist) another country which has seen it's currency weaken against the Yen, then they would see an increase in purchasing power.

So you are proposing that in the future (the time that could happen is unknown), UK, USA or EU will have to experience this long deflation stagnation, so PP will get balanced to the values you said (PP return is the combination of Japan and USA).

The longer term real reward for cash = 0%, gold = 0%, LTT = 2%, stocks = 6%.

25% weightings of each has an expectancy of ( 0% x 0.25 ) + ( 0% x 0.25 ) + ( 2% x 0.25 ) + ( 6% x 0.25 ) = 2% average.

Since 1972 the respective figures have been 1%, 4.4%, 4%, 5.1% (average 3.6%)

A mean reversion shift of perhaps -1%, -4.4%, 0%, 7% (average 0.4%) is perhaps less unreasonable than that of expecting prior near twice expected rewards being maintained indefinitely.

I'm not into economics, so I've no idea what name might be given to such a combination of averages. My guess is that perhaps too much money looking for forms of safe real returns for the investor to live off weighing down (reducing) those real gains to the lowest common denominator. On the basis that long dated inflation bond yields are presently very low (even negative) that LCD could be pretty low. If nine retired have $1m each and one has $10m, then a 4% real might be the LCD. If nine have $10m and one has $1m, a 0% LCD might be an acceptable level.

Have you looked at applying a 10-month SMA to your long term data?  It would also be interesting to see how a Japanese PP would have done with a SMA trading system.  EDIT:  Nevermind, I just saw Ad's link back on page 1.  Here it is again: http://www.advisorperspectives.com/dsho ... e.php  Very interesting...although too bad they didn't go all the way back to 1972.

[Dets]

Thanks for the response Gosso. It makes sense, it's just that the measurement of Japanese CPI has been called into question for some time now.  In this article the authors estimate the CPI in Japan has been overstated by around 2% a year, which would make a huge difference in real return if true.

http://www.nber.org/papers/w13255

[Gosso]

Thanks for the response Gosso. It makes sense, it's just that the measurement of Japanese CPI has been called into question for some time now.  In this article the authors estimate the CPI in Japan has been overstated by around 2% a year, which would make a huge difference in real return if true.

http://www.nber.org/papers/w13255

Hmmm, that's interesting.  I suppose it's possible they are understating their deflation, possibly with the intent to discourage savings.

[MediumTex]

Thanks for the response Gosso. It makes sense, it's just that the measurement of Japanese CPI has been called into question for some time now.  In this article the authors estimate the CPI in Japan has been overstated by around 2% a year, which would make a huge difference in real return if true.

http://www.nber.org/papers/w13255

Hmmm, that's interesting.  I suppose it's possible they are understating their deflation, possibly with the intent to discourage savings.

To me it's all about housing prices and rent levels.

If the cost of housing is falling it takes a LOT of increased milk, egg and gas prices to offset that.

[Arturo]

Thanks for the response Gosso. It makes sense, it's just that the measurement of Japanese CPI has been called into question for some time now.  In this article the authors estimate the CPI in Japan has been overstated by around 2% a year, which would make a huge difference in real return if true.

http://www.nber.org/papers/w13255

Hi Dets,

if the article states that inflation has been overstated by around 2%, and Japan had around 1,2% official inflation during the 90's, it means that they had negative inflation, it means, real deflation of around -2%. So according to this statement, in stead of having a PP return during 90's of 0,5%, they could had around 2.5%, beating inflation/deflation. Its not a super return like the 9,8% in USA, but at least, it could preserved your savings against the toughtest economic conditions in history (from a japanese citizen point of view).

[stone]

Slotline, the currency hedged stocks certainly are not part of a classic HBPP. The classic HBPP has domestic stocks and no hedging. Currency risk exposure is what the gold part of the HBPP is for.

[Dets]

Hi Arturo, I agree, if that number is true it would make things look alot better.  The other thing that stands out to me is how much the strength of the yen hurt the gold portion of the PP return from a Japanese investor's perspective. 

If you look at Gold as a put on your currency though it makes sense that you wouldn't do as well in periods like they just had.

[Gosso]

For reference here is the CAGR for the Nikkei and Dow (dividends not included):

                  Nikkei      Dow (Yen)        Dow (USD)
1972-2012    +3.5%        +3.0%            +6.9%
1980-1990    +18.6%        +6.1%            +11.5%
1990-2000    -5.6%        +11.6%            +15.3%
2000-2012    -6.3%        -1.0%              +1.4%

[D1984]

Having raw S&P500 exposure in Japan through 2008-onwards was probably a bad idea.  But you can't predict things like that. 

A non currency hedged S&P 500 position did do worse from 2008-early 2009 as the S&P's 45%+ loss through early 2009 got compounded in Yen terms by a rising Yen versus the USD. However, by late 2009 the S&P 500 position had closed the gap and then reversed it. Today 100,000 Yen worth of an S&P 500 index fund bought at the start of 2008 is worth just under 78,000 Yen; a similar Yen-denominated investment made at exactly the same time in a TOPIX index fund is worth around 52,000 Yen. The Japanese market crashed hard (like pretty much every other stock market) in late 2008 and early 2009 but unlike every other market it never really recovered. Combine this with the earthquake and the nuclear reactor disaster and it turns out that by mid-2012 (probably actually earlier than that) a Japanese investor would actually have been better off to have invested in the S&P 500 (or for that matter, in emerging markets or in the EAFE-ex Japan) than in his/her home country's equities.

[Gosso]

Slotine,

Currency hedging would have been a good gamble for the Japanese investor.  However, for this to persist the Yen needs to continue to strengthen...it might happen, there is momentum behind it, but in my mind it is a 50/50 chance (others may disagree).  Personally I would have been very happy simply holding the non-hedged Dow (or simply an ex-Japan stock ETF) since it nicely offset the Nikkei's collapse in the 90's and even helped in the 00's.

I was searching for the tracking error on XSP but it appears iShares removed the chart (they must have gotten lots of nasty questions), but IIRC the tracking error was over 1% annually, which included the MER, so the cost to hedge was around 0.80%.  Not worth it, IMO...especially when there's a 50/50 chance of it working...better to go with the low cost option.

Canada is in the same boat as Japan where everyone assumes the loonie will continue to climb against the USD (it's almost impossible to buy a non-hedged S&P 500 ETF denominated in CAD...although Vanguard Canada seems to be rolling one out soon).

[D1984]

From what we've seen of the Japanese PP so far most of the poor performance of the past 22+ years has been due to the bursting of the stock market bubble and resulting balance sheet recession and deflation (and accompanying strengthening of the Yen). Japanese stocks managed their second worst ever long-term returns during these years (the worst ever was if you'd bought Japanese equities during the latter part of WWII; losing the war and the postwar inflation wiped out 95% of the real value of your stocks in less than six years and you didn't even reach breakeven again-much less show any gains-until the early 80s). Even Japanese SCV stocks never provided positive returns at any time from 1990 to the present. This got me wondering if there was any investment strategy (that didn't involve market timing or "the trend is your friend" buy and sell signals, or that didn't involve stock picking...since no actively managed Japanese mutual fund, ETF, unit trust, or toshin that I know of has been able to provide positive results-in Yen terms, not USD terms-either during the last twenty-two years) that involved owning Japanese equities that DIDN'T have negative returns for this period.

As it turns out, there was (and is). I remembered that during the best previous example of a multi-year depressionary/deflationary crash we have to go by (the Great Depression) that high-dividend stocks (assuming you reinvested the dividends) recovered to 1929 levels within a few years whereas the lower-paying (and non-payers) recovered much more slowly (too until late 40s/early 50s to get back to pre-crash levels). I recalled what MachineGhost and AgAuMoney were doing with stable, blue-chip dividend stocks in their PPs/VPs and looked to see if anyone had tried a similar strategy for Japan and how such a stratgey had fared since 1990. Well, someone (Credit Suisse) did. They performed a study (first in 2006 for the US, then in 2009 and 2011 across multiple countries...including one for Japan) of investing in high-dividend, low-payout ratio stocks vs other categories of dividend payers (and vs non-payers and vs the total market indices for those countries).

The strategy was as follows:

1. Take all the stocks listed on a country's broadest stock market index (in Japan's case this would be the TOPIX...I believe it has maybe 1500-1600 stocks).

2. List them in order by market cap from largest to smallest.

3. Eliminate the smallest 15% by market cap.

4. Of the remaining 85%, throw out any non-dividend payers.

5. When you have done that, sort the remaining dividend-paying stocks by yield from greatest to least.

6. Throw out the lowest yielding half of those.

7. Take the stocks you have left (the highest yielding half of the dividend payers) and sort them by payout ratio from greatest to least

8. Throw out the stocks with the highest payout ratios.

9. You now have a quasi-index of high-yielding, low-payout ratio stocks. Equal weight them (buy an equal amount of each) and repeat the process above every quarter to rebalance.


The fact that you own the (relatively) high yielding, low payout stocks means more dividends to compound via reinvestment so you still build wealth even in a falling market. Low payout ratios tend to mean the dividend is sustainable and if it won't be raised at least it won't be cut.

As I recall this strategy (high-yield, low-payout) produced respectable gains in every market and beat the broad index (and the non-payers) in every country; it beat the other categories of dividend payers in all but Hong Kong and Germany; in those two markets high-dividend high-payout stocks won....in Hong Kong apparently dividends are preferred over capital gains as unlike in most major countries dividends are totally tax-free; in Germany high payout ratios are accepted because German corporations-unlike American ones, for instance-tend to let the dividend fluctuate with earnings...so when business is good you get a much higher yield than you'd get on a comparable stock in most other countries but when it's bad the dividend gets cut (and then re-raised when things are going well again); also, German companies tend not to retain as much of their earnings and many times prefer to rely on bank loans (rather than retained earnings) for capital when needed; these two factors perhaps mean that high-payout ratios are not in and of themselves an especially dangerous sign for German stocks and this is possible why the payout ratio screen was less meaningful for Germany. Even with all of the above, high-yield low-payout stocks still produced decent (market-beating) gains in these two countries; they just didn't do quite as well as their high-payout cousins.

Even in Japan's 22-year bear market, the high-yield low-payout strategy produced an average (simple interest) quarterly gain of around 0.60% IIRC. This may seem laughable when compared to American, EM, or non-Japan EAFE returns for the same period but it means that Japanese high-yield low-payout stocks are up (as of the end of 2011) about 50% over their January 1st, 1990 levels. No other non-market timing strategy I know of involving Japanese stocks (other than just flat out shorting them) has produced that kind of returns during what is arguably the longest deflationary spell in modern (fiat money) history.

Screenshots, graphs, and links to sources/files in PDF format can be provided if anyone is interested.

[Ad Orientem]

This got me wondering if there was any investment strategy (that didn't involve market timing or "the trend is your friend" buy and sell signals, or that didn't involve stock picking...since no actively managed Japanese mutual fund, ETF, unit trust, or toshin that I know of has been able to provide positive results-in Yen terms, not USD terms-either during the last twenty-two years) that involved owning Japanese equities that DIDN'T have negative returns for this period.

Just for the record the Japanese version of the PP did not produce negative returns, either real or nominal. The nominal returns were extremely low (to be expected in a deflationary environment) but they were positive. And real returns adjusting for the deflation were also positive, albeit on the low end of the usual 3-6% range.

[D1984]

This got me wondering if there was any investment strategy (that didn't involve market timing or "the trend is your friend" buy and sell signals, or that didn't involve stock picking...since no actively managed Japanese mutual fund, ETF, unit trust, or toshin that I know of has been able to provide positive results-in Yen terms, not USD terms-either during the last twenty-two years) that involved owning Japanese equities that DIDN'T have negative returns for this period.

Just for the record the Japanese version of the PP did not produce negative returns, either real or nominal. The nominal returns were extremely low (to be expected in a deflationary environment) but they were positive. And real returns adjusting for the deflation were also positive, albeit on the low end of the usual 3-6% range.

My bad; I should have been clearer...by "any investment strategy that involved owning Japanese equities" I meant "any strategy that involved investing all or a majority of one's capital only Japanese equities." Obviously the PP does not qualify as per the above as it holds only 25% equities so it is actually a very light on stocks compared to most investment allocations (I would also add that if the Japanese government's inflation/deflation numbers are to be believed then even the Japanese PP had an eleven-year period from 1990-2000 where it earned less for the whole period than was needed to keep up with inflation....although if the government is deliberately understating deflation then the Jap PP did OK-not great but OK-during those years...although must admit I find explanations of positive or negative real returns being obscured by government conspiracies to understate deflation to be rather Shadowstats-ish in a Bizarro World sort of way). For that matter a portolio holding 1% Japanese equities and 99% 10-year JGBs (said portfolio did fairly well from 1990-2011) would technically qualify as "an investment strategy that involved owning Japanese stocks" since it did own SOME stocks but I don't think most people would consider such a portfolio to really be a stock-based investment strategy. Just IMO.

[D1984]

BTW, the Japan PP with 30yr treasury bonds (synthetic history) fairs a bit better by reducing those two drawdowns in the early 90's

How did you construct synethtic 30-yr JGB yields and total return prior to 1999? 20-year yields but on a 30-year bond and using a bond value calcuator to see how a rise or fall in yields would effect the return each year? Did you derive the yield from longer-term corporates based on the spread between 20-year corporates and 20-year JGBs? Or did you do something else entirely (and if so, what was it)?

[MachineGhost]

Even in Japan's 22-year bear market, the high-yield low-payout strategy produced an average (simple interest) quarterly gain of around 0.60% IIRC. This may seem laughable when compared to American, EM, or non-Japan EAFE returns for the same period but it means that Japanese high-yield low-payout stocks are up (as of the end of 2011) about 50% over their January 1st, 1990 levels. No other non-market timing strategy I know of involving Japanese stocks (other than just flat out shorting them) has produced that kind of returns during what is arguably the longest deflationary spell in modern (fiat money) history.

Screenshots, graphs, and links to sources/files in PDF format can be provided if anyone is interested.

What a fantastic find!  This gives me more confidence that focusing on dividend yields provides deflation protection.  Did they compare to value, momentum or quality?

[D1984]

Even in Japan's 22-year bear market, the high-yield low-payout strategy produced an average (simple interest) quarterly gain of around 0.60% IIRC. This may seem laughable when compared to American, EM, or non-Japan EAFE returns for the same period but it means that Japanese high-yield low-payout stocks are up (as of the end of 2011) about 50% over their January 1st, 1990 levels. No other non-market timing strategy I know of involving Japanese stocks (other than just flat out shorting them) has produced that kind of returns during what is arguably the longest deflationary spell in modern (fiat money) history.

Screenshots, graphs, and links to sources/files in PDF format can be provided if anyone is interested.

What a fantastic find!  This gives me more confidence that focusing on dividend yields provides deflation protection.  Did they compare to value, momentum or quality?

The papers did not compare to value, momentum, or quality (Credit Suisse tends to be highly subject-specific in this kind of research, and the paper was about an international dividend strategy). I myself have looked at SCV and regular value (using the Russell-Nomura indices) and neither outperformed this strategy (although both did bet the market as a whole; they both still lost money from 1990-2011 whereas the HYLP dividend strategy-counting reinvested dividends-was at breakeven to 1990 levels by late 1999/early 2000). The only research I have seen on stock selection for Japan using quality (assuming by "quality" you mean earnings quality/stability of earnings) is the Joachim Klement paper in the JFP from a year or two ago (the title is "Earnings Stability as A Source of Alpha" ); the appendix to this paper shows that selecting stocks with high earnings quality (i.e. high earnings stability) beat the benchmark (Nikkei 225 index) by 5.5% per year on average. They didn't break it down into anualized returns or show a total return chart so I don't know if it would've beat the HYLP strategy or not. I have no idea how a momentum strategy would have done versus the HYLP one.

The original Credit Suisse research on the HYLP as imeplemnted in various non-US countries can be obtained at http://www.mhinvest.com/pdfs_mhi/divide ... _Patel.pdf

Updated results for December 31, 2011 for this strategy are available in David Ruff's "Global Equities: Can Dividends Signal Growth?" and in Nuveen Santa Barbara Asset Management's "Why Global Dividend Companies"; I can post copies of these on a file hoster if needed if they are not freely available online anymore.

[miyazaki]

I have another question about the PP. Excuse the very basic question. I just want to check the meaning of the word. As far as I can tell, the word "bond" means the government debt that individuals can buy, right? But I guess it has a different meaning here? I mean, in the PP book, it's written that holding bonds directly with the Treasury, or in my case, the Ministry of Finance, is the best option. If that is the case, then the bond/cash (lumped together due to a maximum of 10 year bond availability) part is essentially a fixed income at 1% interest? Unless buying on the secondary market, it is impossible to get returns like the 30% gain or something I've seen written elsewhere. Or does "bond" mean a secondary market etf only? If it means the former, then that would mean the only portfolio volitility will come from stocks and gold? That doesn't seem right, so I'm sure I misunderstood. Actually, there are no long term bond ETFs currently available in Japan, only mutual funds. These are split into 2 groups, ones with dividends but higher management fees, and ones with no loads but zero dividends... What kind of product is suitable for this part of the allocation? Thank you for your patience with my questions.

[stone]

miyazaki, I bought some UK treasury bonds at £0.91 that pay £0.04 every year until they mature in 2060. They now can be sold on the secondary market for £1.20. If I was to keep them for the full 50years, then I would get 4.3% per year or whatever it is but with the PP you sell them on the secondary market before they get <15years from maturity. Those UK treasury bonds bonds increased in price to £1.27 at one point. Selling them then would have given a greater >30% profit. Ofcourse the price also falls at times. You can track the prices of JGB on the internet I guess if you want to get an idea of how the prices move around. Hope I understood the question.