article

Why do the Japanese keep ¥36 trillion under mattresses and why are government bond yields so low?

Yesterday CNBC published a piece saying that Japanese households could be holding around ¥36 trillion in physical cash. They suggest that part of the reason for this is to avoid tax, and in a society where the deceased are still collecting pensions, it certainly seems that there could be some problems with fraud.

But there is something more important going on here than welfare fraud. The same piece put this ‘mattress money’ into context: last year the Japanese Finance Minister indicated that ¥880 trillion was also being held as cash in the bank.

Together, these statistics simply reveal an economy where there is not much incentive to invest in the future. Individuals and companies are happy to earn close to zero in the bank (or actually zero under their beds) rather than lend.

In fact, the returns you receive for lending to the Japanese government have fallen over the last five years in spite of a massive increase in government debt and an explicit objective to increase inflation.

episode1

Another article yesterday suggested that the Japanese debt to GDP ratio (currently 240% according to the IMF) is set to worsen and even though such forecasts should be treated with the same faith as many of the other articles published on April Fool’s day, now is timely to take a closer look at Japan.

Why aren’t markets more worried about Japanese default?

As the Eurozone crisis rumbles on a common question is why we are so worried about Greece, or Portugal, or Italy, and not Japan, when Japan’s debt to GDP ratio is so much higher (see figure 2).

episode2

The reality is that debt to GDP has not been a good indicator of likely defaults or of where yields are going in developed economies (in the US yields has fallen as debt/GDP has increased for example).  Most cases of sovereign defaults are in emerging markets where the ability to print currency to pay debt has been constrained, either because the country has issued bonds which pay in a foreign currency and/or because the currency has been pegged.

In contrast, because the vast majority of Japanese debt is issued in Yen, it can theoretically continue to meet any debt it has to pay forever. It simply ‘prints’ more Yen to make any payments. For countries that have this ability, default therefore becomes more about willingness to pay than ability to pay; is the country willing to tolerate the costs that come with ‘printing more money’ to pay their debts?

These costs are generally inflation and currency devaluation. And for Japan today, these costs are not nearly beginning to bite. In fact, Japan is still in the position where it is trying its hardest to create inflation and benefits when its currency weakens.

For this reason the market at present is comfortable that the government will meet its obligations, even if this comfort makes some sectors of the investment industry angry. The question remains however: even if existing holders of Japanese debt are comfortable that they will get paid, why are they willing to accept being paid so little?

Why are Japanese bond yields so low?

This is a debate that has been rumbling on for years. There are various arguments but they tend to centre on the fact that growth is weak (in part due to demographics) and so there are few alternatives to generate a better return than even these low yields. Because companies are unwilling to invest and/or banks unwilling to lend as a result of these low growth opportunities banks are left holding large amounts of government bonds.

Other, related, arguments focus on who owns Japanese government debt. Data in a piece from the Wall St. Journal highlighted two aspects of this argument: the small amount of foreign ownership by non-Japanese investors (which our colleagues at Bond Vigilantes discussed in 2012), and the fact that the Bank of Japan has now become the largest holder of government bonds.

episode3

It is the latter fact that has drawn most attention of late as the Bank of Japan’s stimulus programme has led to it buying up over 70% of all new issuance of sovereign debt each month. With the ECB embarking on its own bond buying spree this year, double digit returns from European sovereign bonds, and Ben Bernanke’s recent blog on the ability of central banks to influence rates, the role of central bank bond buying in influencing yields is something the team will discuss in a blog later this month.

Irrespective of these debates however, the question for investors is always the same: how much can I make, how much can I lose, and what are the probabilities attached to this? With Japanese bond yields so low (see figure 4) these probabilities certainly seem asymmetric: even allowing for negative rates the scope for yields to fall appears limited relative to the extent that they can rise.

episode4

The current environment is one where it does not take much of a leap to foresee scenarios that would pressure rates. ‘Abenomics’ is a determined effort to increase inflation and growth, with an inflation target of 2%. We have commented before on how difficult this will be, but if they even come close, a yield of 0.3% on 10 Year Treasuries could look very unappealing. There are other areas too which could be positive for growth: the fall in the oil price has dampened inflation in the short term but could ultimately be strong driver of growth, a proposed VAT increase has been postponed, while exports are showing signs of picking up in response to the fall in the Yen (see figure 5).

episode5

Of course all these forces remain to be seen and many have lost money betting against Japanese government bonds over the years. But, if you do have some spare money under your mattress at home it seems likely that there are far more attractive opportunities out there today than Japanese government debt.


The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested. Past performance is not a guide to future performance.