Sunday, July 17, 2011

Why Japanese government bond yields are so low

 Paul Krugman poses a puzzle:
A question (to which I don’t have the full answer): why are the interest rates on Italian and Japanese debt so different? As of right now, 10-year Japanese bonds are yielding 1.09%; 10-year Italian bonds 5.76%.

I ask this because in a number of ways the two countries look similar. Both have high debt levels, although Japan’s is higher. Both have awful demography. In other respects, the numbers if anything favor Italy, which has a much smaller current deficit as a percentage of GDP.
I don't really know anything about Italy, but I do know a bit about Japan, and today I asked several Japanese economists why Japanese government bond yields are so low when Japanese government debt is so high. Their answers were pretty much in agreement.

It comes down to three things: 1) financial repression, 2) home bias, and 3) dysfunctional equity markets.

Although many of Japan's legendary bureaucratic ministries have dramatically weakened in power and prestige since the 1980s, the Ministry of Finance (MOF) is still extremely powerful. Japan's financial system, like many financial systems in Asia, is dominated by large banks. In order to finance the government's huge deficits, the MOF puts pressure on the big banks to buy lots and lots and lots of Japanese government bonds (JGBs). 

This holds JGB interest rates down to very low levels. It also reduces deposit rates on Japanese households' savings accounts to zero, since the big banks need to maintain their spreads. So households see JGBs as fairly attractive investments compared to savings accounts. This is less of a factor these days, since Japanese net household saving is actually quite low now, but one professor I spoke with told me that households have apparently been shifting money from savings accounts to bonds.  

This is the same kind of financial repression that Michael Pettis has talked about with regards to China. Ultra-low interest rates on household savings act as a stealth tax on households. In China this stealth tax has apparently been used to pay off non-performing loans at big banks; in Japan, it is going to fund government spending. 

Of course, in China, the government can dictate the return on household savings because China has capital controls. Japan does not. So why don't Japanese households (and pension funds, another big purchaser of JGBs) simply defect and purchase foreign assets? Apparently, home bias is a big part of it. Japanese investors are scared of exchange rate risk; only recently have companies began offering households global index funds matched with currency swaps in order to limit this risk. Also, xenophobia appears to play a role as well; Japanese investors still tend to see foreign assets as inherently risky, ignoring the fact that diversification can dramatically lower that risk.

What about the Japanese domestic stock market? The experience since the bubble and crash in the 80s have convinced many Japanese investors that stocks go down in the long run, not up (as Americans tend to believe). And though no one I talked to mentioned this, I suspect that continued prevalence of cross-shareholdings and all-insider boards of directors have convinced investors that Japanese stock markets are institutionally dysfunctional, and hence a bad bet. I can't say I really disagree.

So I think that solves the Japan part of the puzzle. Bureaucratic power over a bank-centric financial system, enabled by home bias, allows the government to hold down its borrowing costs at the expense of households. As for Italy, though I don't know anything about it, I suspect that it has a far more "normal" financial system.

As a final note, it seems to me that the Japanese government's success in holding down its borrowing costs has probably had big negative effects on the economy. Banks that are forced to buy JGBs can't lend as much to firms, which seems like it would depresses economic activity, holds down growth, and probably contribute to deflation via suppressed wages. Households have been squeezed and squeezed by falling incomes until their savings rates have gone negative, yet they are still earning nothing on their savings. As a result, households are dead set against tax hikes to close the budget deficit, but politically powerful farmers and construction companies won't allow spending to fall either. It's hard to see how this vicious cycle will be broken, until Japan is forced to default from the sheer weight of interest payments. That would be the biggest and most catastrophic sovereign default in world history, unless of course the U.S. manages to pull one off first.

Update: A friend at the Ministry of Finance points out that much of the pressure on banks to buy JGBs is regulatory rather than informal. Banks are required by law to keep a certain amount of their capital in "safe assets," which in practice means JGBs only. Foreign-country bonds do not count, even if exchange rate risk is hedged via currency swaps. He also told me that much of the demand for JGBs is expected to come from small and regional banks in the near future, while big banks are hedging their bets by shifting into shorter-term JGBs.


  1. Doc at the Radar Station11:24 AM

    Excellent post! I also wonder if the Japanese economy is "trapped" in a policy of actively maintaining low-interest on bonds to keep their interest payments low. If the economy should start growing significantly the rates would rise. But, rising rates would boost interest payments, so they stay low and the economy remains mired in a contained depression.

    With respect to the financial repression due to fat spreads between savings and loans... what would happen if (in order to fully reflate the economy) the Fed started paying an interest *subsidy* on demand deposits to make up the loss of interest income? Wouldn't this satisfy the heightened Demand for Money (thinking of Nick Rowe here), and boost spending, enhance deleveraging, etc.?

  2. These factors may be important, too, but I think the ability to control your own national currency is an ultimate trump card in any case.

    Your argument seems to be that Japan's debt is in fact as risky as Italy's (if not more so) but Japanese buy it anyway. But home bias (given Japan's high savings rate) should result in the yen being overvalued. If the yen is overvalued, then why does Japan have a large trade surplus?

    I would argue that the yen is actually undervalued -- in a sort of semi-permanent overshoot where a disparity in purchasing power serves to offset the risk of eventual inflation if the government should need to be bailed out. (If Japan ever has a soft landing, presumably the value of the yen will rise so as to erase the trade surplus, and this expected appreciation compensates for the inflation risk in the event of a hard landing.)

    It also should be noted that, in the absence of a sovereign debt crisis, the BOJ seems to be pursuing (probably unintentionally) a much lower inflation target than the ECB. The ECB's desideratum is just under 2%. The BOJ seems to tighten whenever the inflation rate gets positive, and it doesn't loosen aggressively enough (maybe because it isn't possible) to offset this when the inflation rate gets negative, so the effective target is just under 0%. (Again, the critical issue is having ones own currency: Japan's deflation can flow through to its yields, but Italy's is swamped by inflation in the core countries.) This explains about 200 of the 467 basis points that need to be explained, leaving only 267 for you and me to argue over.

  3. john newman9:41 PM

    An alternate view with some prescriptions at the end:

  4. Anonymous3:36 AM

    Perhaps, unlike the US, Japanese financial repression means the GVT has a vested interest in keeping any local equities from gaining too fast, and inducing withdrawal of money from JGBs. If that's the case then it's one thing the Japanese GVT have managed exceptionally well.

  5. @Andy: I certainly think that if Japanese household home bias were to decrease, the yen would fall, all else being equal (a stagnant economy should have a weak currency). But household home bias is not the only thing determining Japan's trade surplus. There are many other factors. And Japanese bonds aren't as risky as Italian bonds, because of the govt.'s ability to tax households via financial repression. But that could change...

    @Anonymous: That thought has indeed crossed my mind...

  6. Nick R, Kyoto, Japan5:58 AM

    Hi Noah,

    Japanese rates are not low. They are among the highest real yields on the planet.

    Here's a piece I recently wrote that helps explain the "sticker shock" conundrum.

  7. Anonymous7:23 AM

    Unless you are including it in your characterization of "big banks", you are forgetting about Japan Post Holdings which is a wholly owned subsidiary of the Ministry of Finance. The Post Office has been the most reliable and voracious consumer of JGBs and one might speculate that the planned privatization of the entity was halted because shareholders who are not the MoF would be more likely to encourage the Post Office to diversify its portfolio; currently very heavily weighted in JGBs.

    Consider then, that the MoF is in fact its own biggest financer. However, Post Office savings accounts are not growing as quickly and reliably as in the past and probably can be expected to shrink in the near future as Japan's large cadre of retired persons begins to draw down the massive savings they have in Post Office accounts.

    What happens then?

  8. @Anonymous: Japan's gross debt is over 200% of GDP; much of this is held by government agencies such as Japan Post. But Japan's net debt, none of which is owned by government agencies, is still 125% of GDP, which is still big.

  9. Benedict@Large8:33 PM

    Japan's interest rates on its bonds are so low because they do not incur any additional liability when they borrow. This is true for all fiat currencies. The liability is incurred when the borrowed currency is created, NOT when it is borrowed. The borrowing is just window dressing.

  10. M-M-T1:04 AM

    It's called being able to print the sovereign unit of account. Traditional default risk is eliminated, leaving only inflation/currency risk. This is the major difference.

  11. Anonymous1:43 AM

    Interesting piece although it might be more credible and more convincing with some supporting data.

    More importantly, the xenophobia comment is bordering on irresponsible. It is unfair to stereotype the Japanese people this way. This comment also reflects badly on you, so do everyone a favor and kindly remove it.

    1. I am sorry, as a resident of Japan, being fluent in japan, understanding the culture and being married to a Japanese, Japan is generally xenophobic. It is not a negative comment, just a culture difference that changes with time.
      The author is not saying every single Japanese person is xenophobic so it is not a negative commen.t

  12. @Anonymous (most recent):

    You're right that more data would be good, though there is some in the links. Not always possible to publish a full paper in a blog post.

    As for "xenophobia," that's just the term that the Japanese profs I talked to used. I realize it's a common stereotype of Japan, and in the main it's an untrue stereotype, but sometimes the word fits.

    Also, I assume from your aggro tone that you're the regular troll...ever think about getting a handle, so that we can distinguish you from the other Anonymouses? You may hate my blog, but that's still no reason to reflect badly on other commenters...

  13. Anonymous11:43 AM

    genauer says: if you think this from a (retired) investor point of view, it is very simple:
    a) Japan public pensions cover about 41 % of your former income for the average Joe (vs 74 % in Italy, according to the OECD), living for 20 more years, requires 33 % * 20 * 25 % of population * 60 % fraction of salaries to GDP (guess) = 100 % GDP in additional savings. Nippon stocks were a brutal disappointment after the 1990 bubble. Even if Mrs Watanabe puts 50 % of it in the carry trades, 50 % GDP must be invested for the next 5 years of consumption into Yen based assets, because of currency fluctuations. You and I would just do the same rational thing (this is not new: Nobel Modigliani Life Cycle Hypothesis, 1952)

  14. Any analysis of Japanese currency and financial markets that fails to note the overwhelming use of the Yen in the eye-waveringly massive global carry trade misses the plot entirely. The frantic quest for yield in an over-leveraged loose-money world lives on.

    You might have started this piece by noting that the trade weighted Yen actually increased sharply following the devastating tsumai and Fukushima nuclear fiasco. I'd absolutely love to hear an orthodox economist explain how that one fits the theory. Likewise the hardening CW that treasuries will be the best place to be in the event treasuries default.

    Ultimately, as long as 'stable disequilibrium' holds and all of the financial dysfunction on which it rests, there is no conventional theory that supports these behaviors. Japan's government debt load is utterly and incontrovertibly unsustainable, but then that differentiates it not in the least from so much of the globe's positively absurd mountain of financial assets.

    The trick is not to let the speculators spot the kick me sign (Italy's neighbors and lack of printing press make that far more difficult). As Keyenes so astutely noticed it's not what people think, but what people anticipate others will. A massive part of that is convention and institutional arrangement. None of that is a tailwind for the Italians. Then again, this next crisis is fry likely to be the last hurrah for the current global monetary regime, so this could simply be a competition for who gets to play the part of subprime.

  15. Alex Sinclair4:46 PM

    Interest costs to the government consume 22.4% of budget revenues at a blended 1.4% cost. A credit downgrade or two, some inflation or a declining yen could push rates higher. Could interest absorb 50% of revenue? The retirement fund is liquidating $78 billion of JGB's this year. Japan Post, the largest financial institution in the world, is no longer a net buyer and the banks own too much as deposits no longer really grow as the older savers have been retiring--withdrawing funds for living expenses. Young people have become contract workers and generally are having trouble saving money. The real issue for Japan is: Who is going to buy their debt in the future other than the central bank? Too much buying by the central bank could ignite a rapid decline in the yen and an end to what has worked for years.

  16. Anonymous9:09 AM

    "Banks that are forced to buy JGBs can't lend as much to firms"

    and given the corporate sector surplus, it's clear that those companies are starved of funds due to the lack of bank loans...

  17. I like this post very much.Keep doing like this

  18. Please could you get an RSS feed for your site, So I can incorporate it into Reader....

  19. Please could you get an RSS feed for your site, So I can incorporate it into Reader....

  20. Anonymous7:37 AM

    I do not understand all the complicated economics talked above. When I read, I seem to undesratand but then it becomes confusing. Someone said "Great Engineering is simple Engineering". "If something is complex, then it is also wrong". I think that Japan inc is alive and well because they have trade surpluses, they have work ethics, probably they care about their poor, they respect their elders, they have the technology, they have the discipline ... all things fundamental to success than simple thories of economics.

  21. Anonymous7:09 AM

    The BOJ sets interest rates, and there's no inflationary pressure. That's all it is you MORONS.