The Yen, Flow of Funds and the Stock of JGBs

I've been meaning to put this piece together for a while, but couldn't face ploughing through the Flow of Funds data on the BoJ's website - a traumatising exercise indeed. There has been a lot of noise recently about Japan's Government Debt soaring above 200% of GDP and John Paulson & David Einhorn have been telling anyone that listens that interest rates in Japan are going to rise dramatically, spuring a default and possible currency crisis. They tell a good story that is certainly not outside the realms of possibility and they are both extremely smart individuals, but I think they have got this wrong. As a warning, this is a bit wonkish...

There are several arguments for a sustained sell-off in JGBs. First, the DPJ have announced that they will be ramping up fiscal stimulus (as part of their rebalancing strategy) which will result in a signficant jump in issuance, sending the fiscal deficit to nearly 11% of GDP and the stock of government debt to well over 200% of GDP. Second, the collapse in USDJPY over the past couple of years has led many to argue that the Yen is overvalued - PPP measures support such a conclusion, but as will be detailed below, the reality is more subtle. Third, the tepid rebound in Japanese Industrial Production in the face of dramatic rebounds in, for example, Korea has led many to argue for competitive devaluation. Fourth, net portfolio flows in recent months suggest that foreigners are abandoning Japanese assets, while Japanese continue to accumulate foreign assets (see below chart). And Fifth, that the combination of these factors will mean that financing the deficit will require much higher fiscal premia and the possibility of a fiscal crisis is thus very high.

But these arguments make several assumptions about the structure of Japanese capital markets and neglect to consider the fact that Japan runs a large (although, admittedly smaller than it was a couple of years ago) Current Account surplus. First, the issue of Yen valuation: indeed PPP measures suggest an overvalued currency, however, long-run REER measures (see below JPMorgan's Broad REER Index) suggest that the level of the Yen is about right - the real issue was the severe undervaluation of the Yen in the period 2005-8. An optical focus on the USDJPY rate rather than broader measures is probably responsible for this conclusion.

For many years, Japan has recycled its Current Account surplus by purchasing foreign assets - especially USTs. As can be seen from the below chart from the US Treasury's TIC data (brown line is Total Foreign Holdings of USTs, white line is Japanese holdings), this has accelerated over the past couple of years, presumably as a result of rising risk aversion and a desire to "average in" as the USD fell. But the fact that these purchases have continued at about the same rate as during the periods of extreme risk aversion despite the fact that risky asset markets have recovered much of their lost ground suggests something else is going on.

Usually, these foreign purchases have been left unhedged on an FX basis as the cost of the hedge has historically been quite high (selling USDJPY and funding it day to day or via an FX Forward resulted in paying significant negative carry). But this is no longer the case: since central banks globally have slashed rates, there has been a convergence of short-term interest rates. Specifically, it has never been this cheap to fund a short-USDJPY position: the Tom/Next Yen FX Swap is essentially zero. So I believe that Japanese financial institutions have been playing a carry trade whereby they have bought long-dated USTs and hedged the FX risk by selling USDJPY and are funding it short-term. And it is for this reason that those that have been (fashionably) short the Yen have been repeatedly frustrated this year. Such weakening is unlikely until either (i) long-term UST yields converge sufficiently with JGB yields, (ii) the Fed hikes rates sufficiently to make funding short-USDJPY positions more expensive, (iii) portfolio flows move sufficiently negative as to offset the recycling of the Current Account, or (iv) Japan ceases to run a Current Account surplus. It is clear that there is a long way to go before the former conditions are met, and given foreigners own such a small amount of Japanese financial assets (a mere 6.8% of JGBs are foreign-owned, in stark contrast to 35% of Gilts and 50.3% of USTs!) it is unlikely that the third condition will be met. The final condition is perhaps the only one that seems reasonable - however, if the broad rally in risky assets continues, even in the face of central bank intervention, major competitor currencies (such as the KRW) should appreciate and reduce the pressure on the trade balance.

Regarding the issue of whether increased issuance of JGBs can be financed without recourse to higher interest rates or a slide in the currency, the key issue here is FLOW OF FUNDS not STOCK OF JGBs. And it is here that I think Paulson & Einhorn's arguments fall down because they are talking about the stock of JGBs, which as a proportion of GDP is certainly approaching dizzying levels that have in some cases historically been associated with default, but in others (notably during wars) have not. Those who have read their financial history will note that there have been very few instances of sovereign debt crises where the country in question ran a Current Account surplus. In fact, off the cuff, I can't think of any. There is a reason for this: a Current Account surplus represents an excess of savings, and in Japan's case, the Government's borrowing is more-than offset by domestic savings. The point being that government liabilities can be funded entirely domestically and the currency is only at risk if the domestic population suddenly decide that their currency is worthless. There are plenty of instances of this occuring historically, but they have always been accompanied by hyperinflation, whereas deflation is entrenched in Japan.

_

One of the key arguments that the bears put forward is centred around the idea that the Household's savings rate has declined dramatically as Japan's demographic problems materialise, and thus there will be no household financial surplus to fund the very large 11% government deficit. That household saving will not be able to fund the deficit is certainly true to an extent, however, given the damage done to household balance sheets over the past couple of years, it would be reasonable to expect some increase in saving to repair this damage. But much more importantly, their argument neglects to consider the financial surplus of private sector corporates. Specifically, corporates have been squeezed from all sides: the Yen has rallied sharply against major competitors, the global recession has suppressed external demand and the DPJ's fiscal stimulus has been aimed squarely at consumers, rather than at them (past stimuli have usually been comprised of corporate tax cuts and infrastructure/pork barrel policies). Faced with such headwinds, corporates have no choice but to build their savings - in response to the transition of the Japanese banking crisis into its acute stage in 1997/1998, corporates rapidly ramped their precautionary savings and slashed CapEx (see below chart of Financial Surplus/Deficits from the Flow of Funds data - the BoJ altered the construction of this data in 1999 so the crossover levels may not be perfect, but will not detract from the overall argument). Indeed, given the large amount of spare capacity both domestically and also globally, it is unlikely that corporates will be looking to invest in the near-term. Thus, corporates holdings of JGBs are likely to rise both directly and indirectly (via their bank deposits being recycled into JGBs by the banking system). There is a slightly more subtle corrollary from this effect. An increase in corporate savings is equivalent to a fall in demand for loans from financial institutions. As a result, bank loan-to-deposit ratios will fall further - they are already below 80% in aggregate - and as has historically been the case, financial institutions will increase their share of the stock of JGBs.

To conclude, those who expect the Japanese Government Debt to GDP ratio to result in an imminent crisis are confuing Stock and Flow of Funds. It is undoubtably true that the ratio is approaching unsustainable levels, but it is not clear whether a catalyst will present itself to force the issue. The simple existance of Japan's Current Account surplus and structure of its capital markets (where 93.2% of the stock of JGBs are held domestically, and only 5.2% of these directly by households) make it extremely unlikely that a sovereign debt crisis can occur [UPDATE: As reader GE succinctly points out, Japan will never default on its own people]. Additionally, the low cost of funding external carry trades in, for example, USDJPY means that there is additional support for the the Yen in the form of FX hedging, preventing the exchange rate from acting as a lever to force the issue. Finally, note that this FX hedging by Japanese financial institutions also allows the BoJ to increase its purchases of JGBs without risking weakening the currency (as has been the case with the US and UK versions of Quantitative Easing). To summarise, those attempting to short JGBs and the Yen on the expectation of a sovereign debt and currency crisis are likely to lose money like those before them did on numerous occasions in the 1990s.