gipa69
collegio dei patafisici
Interessanti articoli sulla debolezza dello yen che tiene presente non solo l'aspetto speculativo.
In particolare da una parte il terzo tipo di carry elencato nel primo articolo (quello effettuato dall'estero) e dall'altro il fattore 2 ed il fattore 3 sempre del primo articolo e cioè la copertura che effettuano gli investitori giapponesi che vanno all'estero e degli investitori esteri che vanno in Giappone e le enorme riserve valutarie detenute dai paesi asiatici e dai paesi del medio oriente che vengono diversificate principalmente su sterlina euro e dollaro contribuiscono a quella enorme posizione short contro yen.
Questi tre movimenti confermano in un modo o nell'altro la crescita e la forza dell'economia e quindi questo è il motivo perchè la debolezza dello yen è supportiva alla crescita dei mercati nel loro complesso.
Certamente in ordine crescente di pericolosità per la stabilità dei mercati troviamo il fattore 3, poi il fattore 2 ed infine il terzo tipo di carry.
Currencies: Why Is the JPY So Weak?
Stephen Jen and Luca Bindelli (London)
Summary and conclusions
The JPY remains weak, despite the economic recovery in Japan. In REER terms, the JPY is at its weakest level since the period right before the Plaza Accord in 1985. It is widely presumed that the so-called ‘JPY carry trades’ — powered by the net fixed income outflows from Japan — have become even more popular recently as most other central banks have tightened much more than the BoJ has. It has been argued that unless these Japanese capital outflows abate, the JPY will likely remain weak.
In this note, we challenge the notion that the primary reason behind the weak JPY is increased Japanese capital outflows. Instead, we propose three other factors that may have weighed on the JPY: (1) a sharp curtailment of equity inflows into Japan in 2006; (2) changing currency hedging ratios; and (3) the ‘Global Funneling’ process. All these factors may take time to reverse, to push JPY higher.
Questioning the notion of rising Japanese capital outflows
Investors and commentators talk about ‘JPY carry trades’ without properly defining what they are. In a previous note (Don’t Blame the BoJ for the Bloated Global Asset Prices, June 15, 2006), we suggested three types of JPY carry trade. What we find remarkable is that none of the three types of JPY carry trades has actually increased in size in recent months, contrary to popular presumption.
• Type 1 JPY carry trade: net fixed income outflows from Japan. This is the type of JPY carry trade we presume most investors have in mind: net fixed income outflows propelled by interest rate differentials. Despite the popularity of this notion, data do not definitively support this idea. Since 2H05, net bond outflows have actually declined sharply. In fact, in recent months, there have been net bond inflows, contrary to popular presumption. The recent rally in NZD/USD and NZD/JPY has been widely interpreted as a sharp rise in uridashi flows into New Zealand. This may very well have been the case. However, from the perspective of the net aggregate fixed income outflows, the official data don’t corroborate the notion that this type of flow was behind the weakening in the JPY this year.
• Type 2 JPY carry trade: JPY duration trade. There are also JPY-JPY carry trades whereby Japanese investors fund their long JGB positions with short-term credit/loans. This, in fact, was the mechanism through which the BoJ ensured that the yield curve remained contained with quantitative easing (QE). This is an indirect way through which other types of JPY carry trades can be fuelled, as the whole yield curve in Japan is artificially depressed by JGB purchases by banks. Japanese banks have been more involved with this type of JPY carry trade than other institutions. However, Japanese Banks have not increased their holdings of JGBs in recent years. Also, banks’ holdings of foreign securities have not increased either, consistent with the first point we made above.
• Type 3 JPY carry trade: non-Japanese residents borrow in JPY outside Japan. Data on these activities are not definitive. In the BIS Quarterly Review from June 2006, there was a chapter on this issue. However, as we argued in our note from June 15, 2006, the survey data are unclear on whether there has been a large increase in foreign borrowing in JPY. We highlighted that while the stock of outstanding JPY-denominated claims held by both Japanese and non-Japanese banks rose noticeably in 4Q05, JPY loans from Japanese banks were declining in 2005, reaching US$181 billion by year-end.
Other reasons why the JPY is weak
We are in no way disputing that there are large short-JPY positions in the market, in search of higher nominal yields outside Japan. What we question is whether this is the dominant factor behind the weak JPY, particularly when data are not unambiguously supportive of this popular notion. Specifically, we believe that three other factors could also be at least as important as the ‘JPY carry trades’. Correctly identifying the factors weighing on the JPY is important, as a prospective rally in the JPY would require a reversal of these factors.
• Factor 1. A collapse in net equity inflows into Japan. We suspect that the change in net equity flows may have been at least as important as ‘JPY carry trades’ in keeping the JPY weak this year. Net equity inflows averaged around ¥0.88 trillion (or around US$7.7 billion) a month during mid-2003 to end-2005. However, so far this year, net equity inflows have averaged only ¥0.48 trillion (or US$4.2 billion). In fact, in July, the 3MMA of the net equity flows actually turned negative.
• Factor 2. Changing currency hedge ratios. Data on currency hedging are scant. However, we believe that the reason why, in general, short-term nominal rates have become so influential for currencies is that cross-border holdings of assets have risen sharply in recent years: countries simply hold much more of each others’ assets than ever. The bloating of the international balance sheets elevates the role of currency hedging. Since the cost of hedging is determined by short-term nominal interest rates, central bank policies have become even more important for exchange rates than before. Further, while tracking capital flows is helpful in thinking about USD/JPY, hedging applies to the entire stock of assets outstanding. In other words, in theory, the adjustments in the hedge ratios by Japanese and non-Japanese investors apply to the entire stock of assets being held, not just the flows. As the nominal cash yield differentials between the US and Japan stay wide, it pays for Japanese investors to reduce their hedge ratios on their USD assets and for non-Japanese investors to raise their hedge ratios on their Japanese assets holdings.
• Factor 3. The ‘Global Funneling Hypothesis’. Back in August, we proposed a structural explanation for why there has been a steady upward bias in EUR/JPY and GBP/JPY. As long as excess savings from oil exporters and Asian exporters continue to be funneled into primarily USD, EUR and GBP assets, there will be a constant upward bias to USD/Asia, EUR/Asia and GBP/Asia. The fact that JPY is now behind GBP as a reserve currency is very telling.
As long as the funds under management by oil-exporting countries (which we estimate to be around US$1.2 trillion: about US$350 billion in the form of official reserves and another US$850 billion on ‘sovereign wealth funds’), and by the Asian central banks (with US$2.6 trillion in reserves, and another US$500 billion in ‘sovereign wealth funds’) continue to rise, this process will remain an important factor depressing the JPY, and prevent it from rallying too hard too fast.
What we are not yet certain about is whether the level of the rate of change of these excess savings affects the strength of this funneling process.
From ‘nominal’ to ‘real’
Nominal variables continue to drive exchange rates. This is problematic, both from a conceptual and a fair value perspective. If a country’s inflation rate is kept low because of productivity growth, its currency would be weak if nominal factors dominate, but the JPY should be strong if real factors dominate. This is precisely what we are witnessing in Japan. According to our calculations, some two-thirds of Japan’s growth in recent years may have come from total factor productivity (TFP) growth. This means that both aggregate supply and aggregate demand have risen in tandem, creating muted inflationary pressures. From a valuation perspective, high TFP growth should lead to a strong currency. But if inflation remains low, the BoJ could afford to be slow in normalizing rates, and the JPY would stay weak as a result. This is perverse, and ultimately a breaking point will be reached where the real fundamentals will start to dominate, in our view.
From ‘bond’ to ‘equity culture’
This distinction between ‘nominal’ and ‘real’ is also related to the distinction between ‘bond’ and ‘equity’ culture. In the example we proposed above, where high TFP growth depresses inflation and therefore interest rates, both equities and bonds should do well in Japan. If ‘bond culture’ dominates, fixed income flows would be more important in driving USD/JPY. But if ‘equity culture’ returns, equity flows would drive USD/JPY lower. Thus, for USD/JPY and EUR/JPY to trade lower, we need to see a return to ‘equity culture’ and a resumption of net equity flows into Japan.
Bottom line
We failed to find definitive proof that ‘JPY carry trades’ have risen this year, keeping the JPY weak, despite the strong real fundamentals of Japan. Instead, we suspect that three other factors have been at least as powerful drivers for the JPY: (1) the collapse in foreign equity inflows; (2) changing currency hedge ratios; and (3) the ‘Global Funneling’ process. Therefore, for the JPY to rally, we need to see a resumption of foreign equity inflows, the BoJ normalizing rates and oil prices staying low. It may take some time for these three factors to turn around. Given the circumstances, the JPY may stay on the weak side for longer than we had expected.
Currencies: GPIF Outflows Still a Headwind for JPY
Stephen L. Jen (from Tokyo)
Summary and conclusions
The JPY remains weak, relative to both history and Japanese economic fundamentals. Nominal cash yield differentials are an important headwind for the JPY, and the ‘Global Funneling’ process may also have undermined the JPY. In this note, I argue that an important structural factor — continued diversification by the Government Pension Investment Fund (GPIF) — is yet another force investors should keep in mind when thinking about the JPY.
Though I continue to believe that, over the coming months, the JPY is more likely to trade meaningfully stronger as current pricing is already very stretched in valuation terms, I recognize that risk-reward for the near term is in favor of JPY shorts. I see the low 120s in USD/JPY and 150-155 for EUR/JPY as the multi-year peaks for these two crosses.
The basic story
I first wrote about this issue on August 5, 2004 in GPIF Outflows as a Medium-Term Support for USD/JPY. In that note, I argued that the GPIF, which manages some US$1.4 trillion in assets, was on track to invest more than US$100 billion overseas between 2004 and March 2009.
In March 2005, the end-2009 targeted foreign asset holdings were raised from 15% of total assets to 17%. At the current exchange rate, this translates into cumulative outflows of close to US$77 billion between April 2006 to March 2009, for average annual outflows of close to US$26 billion over the next three years.
This GPIF argument is totally separate from the other arguments I’ve presented in recent weeks on USD/JPY and EUR/JPY. It is a multi-year plan that has somehow not attracted as much attention from investors as it should have, in my opinion.
In March 2001, as a part of FILP (Financial Investment and Loan Program) Reform, or the ‘Zaito Reform’, the way in which government-managed pensions (or the pension reserves) were invested was significantly altered. Before the reform, all pension reserves had to be deposited with the MoF’s FILP account. Of these funds, most was then directly lent to various public works projects (called the FILP projects), while the rest (about 15% of the total assets as of end-March 2001, or about ¥10.7 trillion, including investments in Japanese and foreign equities and foreign bonds) was invested in non-FILP assets (e.g., stocks and bonds). The non-FILP investment fund is the Nempuku Fund.
After the Zaito Reform, the Nempuku Fund was reorganized into the GPIF. The key aim was to sever the hard financing link between the pension reserves and the public works projects under the FILP. From 2001 onward, public pension funds would enjoy a great deal more flexibility in the assets it can hold, while the public works projects (the FILP projects) would compete for financing in the open capital markets, like any other project or business, but with government guarantees in some cases. In other words, in theory, the FILP projects would need to, directly or indirectly, issue a large amount of their own bonds in the market, while the public pension funds would be relatively free to decide whether or not to invest in these bonds.
To minimize the ‘shock’ impact of this change — keep in mind that the FILP projects were worth a huge ¥147 trillion (or about US$1.2 trillion) — this transition in the financing scheme was to be spread out over eight years. The aim was that, by end-March 2009, the publicly managed pensions would have no direct financing links with the FILP projects.
The basic ‘benchmark portfolio’ targeted for 2009 includes a healthy size of foreign portfolio holdings. Between April 2006 and March 2009, I calculate that some ¥9.2 trillion (equivalent to close to US$76.7 billion) in new foreign bonds and stocks will be bought, mostly unhedged, which translates into about US$26 billion in outflows in each of the coming three years. In FY2004 and FY2005, net outflows from the GPIF were a bit higher, at some US$35 billion each year. These planned outflows are of a significant size — equivalent to about a third of Japan’s total trade surplus — and should help keep the JPY undervalued.
A bit more detail on the updated calculations
As of March 2001, when the diversification program began, GPIF had ¥173 trillion in total assets, 85% of which was directly invested in the various FILP projects, with another 8.5% invested in JGBs. Not counting ‘short-term assets’, only 6.1% of total assets (or ¥10.7 trillion — the figure I mentioned earlier) was invested in Japanese stocks or foreign stocks and bonds.
As of March 2006, GPIF had ¥161.9 trillion in total assets. Foreign asset holdings had risen to 11.3% (from 2.5% in 2001, and against a target of 11.0%) of total assets. The absolute value of foreign security holdings was ¥18.3 trillion.
For March 2009, the targeted foreign asset holdings are 8.0% in foreign bonds and 9.0% in foreign stocks. This means that, from an initial share of 2.5% of total assets held in foreign securities back in 2001, the aim is to raise these holdings to 17% by 2009. In absolute terms, this corresponds to an increase in foreign security holdings of ¥23.2 trillion (or close to US$200 billion) over the eight-year transitional period.
My thoughts
I have the following thoughts:
1. JPY carry trades over-rated. I am in no way disputing the existence of JPY carry trades. However, I think investors place too much emphasis on this idea, and may miss other factors that need to be considered in thinking about the JPY. In my September 28 piece, Why Is the JPY So Weak?, I pointed out that bond outflows from Japan have not increased with the widening in cash yield differentials vis-à-vis the US and Europe. I also suggested that currency hedging, the ‘Global Funneling Hypothesis’, and now GPIF outflows are also important factors to consider. The prospective EUR/JPY and USD/JPY sell-off I envisage for the coming years will be a gradual one, due to these headwinds for JPY.
2. Fed-BoJ discussion is critical for USD/JPY. Given the importance of the cash yield differentials, what the Fed and the BoJ do is clearly very important. Our official view is that the Fed is likely to hike one more time in early 2007 to take the FFR to 5.50%. For the BoJ, the outlook is at least as uncertain. There will be some technical factors that will perturb CPI in the coming months: changes in cell phone charges may add 0.25% to CPI; the rise in medical costs may add another 0.10% to CPI; but the fall in gasoline prices could reduce CPI by 0.20%. What this means is that, when the BoJ meets in December, CPI (ex-fresh food) may be running at 0.2-0.4%. December seems to be the first opportunity for the BoJ to consider another rate hike, as the next Tankan and the Tokyo November CPI will be out by then. My belief is that the BoJ has a good chance of taking the policy rate to 0.50% at its December meeting. Of course, the next Outlook Report, to be released at the end of this month, will provide important guidance on the BoJ’s policy path. Specifically, the discussion on consumption and wages will be key, as muted wage pressures is the number one economic puzzle in Japan right now.
3. Declining ‘home bias’ a real issue for the JPY over the medium term. In general, declining ‘home bias’ has been a common theme in the world in recent years. I myself have written on the fall in the Feldstein-Horioka S-I coefficient, and this trend is one explanation for why global imbalances have not posed nearly as acute a problem for the USD, the AUD and the NZD as traditionalists may have thought. Even in Japan — a country known for exceptionally high ‘home bias’ — there has been a broad-based increase in interest in and appetite for foreign assets.
4. GPIF outflows do not dictate USD/JPY’s trend. It is important to keep in mind that GPIF outflows are one of several factors to consider. In no way do these outflows dictate the trajectory of USD/JPY. For example, in 2002-04, USD/JPY drifted lower, despite some US$65-70 billion worth of capital outflows from the GPIF. Similarly, US$26 billion in annual outflows from the GPIF should be seen as a headwind for the JPY and not something that will necessarily sink it.
Bottom line
GPIF will continue to invest US$26 billion in each of the next three years in foreign securities. I believe that this will pose yet another headwind for the JPY, cushioning the eventual fall in USD/JPY and EUR/JPY.
In particolare da una parte il terzo tipo di carry elencato nel primo articolo (quello effettuato dall'estero) e dall'altro il fattore 2 ed il fattore 3 sempre del primo articolo e cioè la copertura che effettuano gli investitori giapponesi che vanno all'estero e degli investitori esteri che vanno in Giappone e le enorme riserve valutarie detenute dai paesi asiatici e dai paesi del medio oriente che vengono diversificate principalmente su sterlina euro e dollaro contribuiscono a quella enorme posizione short contro yen.
Questi tre movimenti confermano in un modo o nell'altro la crescita e la forza dell'economia e quindi questo è il motivo perchè la debolezza dello yen è supportiva alla crescita dei mercati nel loro complesso.
Certamente in ordine crescente di pericolosità per la stabilità dei mercati troviamo il fattore 3, poi il fattore 2 ed infine il terzo tipo di carry.
Currencies: Why Is the JPY So Weak?
Stephen Jen and Luca Bindelli (London)
Summary and conclusions
The JPY remains weak, despite the economic recovery in Japan. In REER terms, the JPY is at its weakest level since the period right before the Plaza Accord in 1985. It is widely presumed that the so-called ‘JPY carry trades’ — powered by the net fixed income outflows from Japan — have become even more popular recently as most other central banks have tightened much more than the BoJ has. It has been argued that unless these Japanese capital outflows abate, the JPY will likely remain weak.
In this note, we challenge the notion that the primary reason behind the weak JPY is increased Japanese capital outflows. Instead, we propose three other factors that may have weighed on the JPY: (1) a sharp curtailment of equity inflows into Japan in 2006; (2) changing currency hedging ratios; and (3) the ‘Global Funneling’ process. All these factors may take time to reverse, to push JPY higher.
Questioning the notion of rising Japanese capital outflows
Investors and commentators talk about ‘JPY carry trades’ without properly defining what they are. In a previous note (Don’t Blame the BoJ for the Bloated Global Asset Prices, June 15, 2006), we suggested three types of JPY carry trade. What we find remarkable is that none of the three types of JPY carry trades has actually increased in size in recent months, contrary to popular presumption.
• Type 1 JPY carry trade: net fixed income outflows from Japan. This is the type of JPY carry trade we presume most investors have in mind: net fixed income outflows propelled by interest rate differentials. Despite the popularity of this notion, data do not definitively support this idea. Since 2H05, net bond outflows have actually declined sharply. In fact, in recent months, there have been net bond inflows, contrary to popular presumption. The recent rally in NZD/USD and NZD/JPY has been widely interpreted as a sharp rise in uridashi flows into New Zealand. This may very well have been the case. However, from the perspective of the net aggregate fixed income outflows, the official data don’t corroborate the notion that this type of flow was behind the weakening in the JPY this year.
• Type 2 JPY carry trade: JPY duration trade. There are also JPY-JPY carry trades whereby Japanese investors fund their long JGB positions with short-term credit/loans. This, in fact, was the mechanism through which the BoJ ensured that the yield curve remained contained with quantitative easing (QE). This is an indirect way through which other types of JPY carry trades can be fuelled, as the whole yield curve in Japan is artificially depressed by JGB purchases by banks. Japanese banks have been more involved with this type of JPY carry trade than other institutions. However, Japanese Banks have not increased their holdings of JGBs in recent years. Also, banks’ holdings of foreign securities have not increased either, consistent with the first point we made above.
• Type 3 JPY carry trade: non-Japanese residents borrow in JPY outside Japan. Data on these activities are not definitive. In the BIS Quarterly Review from June 2006, there was a chapter on this issue. However, as we argued in our note from June 15, 2006, the survey data are unclear on whether there has been a large increase in foreign borrowing in JPY. We highlighted that while the stock of outstanding JPY-denominated claims held by both Japanese and non-Japanese banks rose noticeably in 4Q05, JPY loans from Japanese banks were declining in 2005, reaching US$181 billion by year-end.
Other reasons why the JPY is weak
We are in no way disputing that there are large short-JPY positions in the market, in search of higher nominal yields outside Japan. What we question is whether this is the dominant factor behind the weak JPY, particularly when data are not unambiguously supportive of this popular notion. Specifically, we believe that three other factors could also be at least as important as the ‘JPY carry trades’. Correctly identifying the factors weighing on the JPY is important, as a prospective rally in the JPY would require a reversal of these factors.
• Factor 1. A collapse in net equity inflows into Japan. We suspect that the change in net equity flows may have been at least as important as ‘JPY carry trades’ in keeping the JPY weak this year. Net equity inflows averaged around ¥0.88 trillion (or around US$7.7 billion) a month during mid-2003 to end-2005. However, so far this year, net equity inflows have averaged only ¥0.48 trillion (or US$4.2 billion). In fact, in July, the 3MMA of the net equity flows actually turned negative.
• Factor 2. Changing currency hedge ratios. Data on currency hedging are scant. However, we believe that the reason why, in general, short-term nominal rates have become so influential for currencies is that cross-border holdings of assets have risen sharply in recent years: countries simply hold much more of each others’ assets than ever. The bloating of the international balance sheets elevates the role of currency hedging. Since the cost of hedging is determined by short-term nominal interest rates, central bank policies have become even more important for exchange rates than before. Further, while tracking capital flows is helpful in thinking about USD/JPY, hedging applies to the entire stock of assets outstanding. In other words, in theory, the adjustments in the hedge ratios by Japanese and non-Japanese investors apply to the entire stock of assets being held, not just the flows. As the nominal cash yield differentials between the US and Japan stay wide, it pays for Japanese investors to reduce their hedge ratios on their USD assets and for non-Japanese investors to raise their hedge ratios on their Japanese assets holdings.
• Factor 3. The ‘Global Funneling Hypothesis’. Back in August, we proposed a structural explanation for why there has been a steady upward bias in EUR/JPY and GBP/JPY. As long as excess savings from oil exporters and Asian exporters continue to be funneled into primarily USD, EUR and GBP assets, there will be a constant upward bias to USD/Asia, EUR/Asia and GBP/Asia. The fact that JPY is now behind GBP as a reserve currency is very telling.
As long as the funds under management by oil-exporting countries (which we estimate to be around US$1.2 trillion: about US$350 billion in the form of official reserves and another US$850 billion on ‘sovereign wealth funds’), and by the Asian central banks (with US$2.6 trillion in reserves, and another US$500 billion in ‘sovereign wealth funds’) continue to rise, this process will remain an important factor depressing the JPY, and prevent it from rallying too hard too fast.
What we are not yet certain about is whether the level of the rate of change of these excess savings affects the strength of this funneling process.
From ‘nominal’ to ‘real’
Nominal variables continue to drive exchange rates. This is problematic, both from a conceptual and a fair value perspective. If a country’s inflation rate is kept low because of productivity growth, its currency would be weak if nominal factors dominate, but the JPY should be strong if real factors dominate. This is precisely what we are witnessing in Japan. According to our calculations, some two-thirds of Japan’s growth in recent years may have come from total factor productivity (TFP) growth. This means that both aggregate supply and aggregate demand have risen in tandem, creating muted inflationary pressures. From a valuation perspective, high TFP growth should lead to a strong currency. But if inflation remains low, the BoJ could afford to be slow in normalizing rates, and the JPY would stay weak as a result. This is perverse, and ultimately a breaking point will be reached where the real fundamentals will start to dominate, in our view.
From ‘bond’ to ‘equity culture’
This distinction between ‘nominal’ and ‘real’ is also related to the distinction between ‘bond’ and ‘equity’ culture. In the example we proposed above, where high TFP growth depresses inflation and therefore interest rates, both equities and bonds should do well in Japan. If ‘bond culture’ dominates, fixed income flows would be more important in driving USD/JPY. But if ‘equity culture’ returns, equity flows would drive USD/JPY lower. Thus, for USD/JPY and EUR/JPY to trade lower, we need to see a return to ‘equity culture’ and a resumption of net equity flows into Japan.
Bottom line
We failed to find definitive proof that ‘JPY carry trades’ have risen this year, keeping the JPY weak, despite the strong real fundamentals of Japan. Instead, we suspect that three other factors have been at least as powerful drivers for the JPY: (1) the collapse in foreign equity inflows; (2) changing currency hedge ratios; and (3) the ‘Global Funneling’ process. Therefore, for the JPY to rally, we need to see a resumption of foreign equity inflows, the BoJ normalizing rates and oil prices staying low. It may take some time for these three factors to turn around. Given the circumstances, the JPY may stay on the weak side for longer than we had expected.
Currencies: GPIF Outflows Still a Headwind for JPY
Stephen L. Jen (from Tokyo)
Summary and conclusions
The JPY remains weak, relative to both history and Japanese economic fundamentals. Nominal cash yield differentials are an important headwind for the JPY, and the ‘Global Funneling’ process may also have undermined the JPY. In this note, I argue that an important structural factor — continued diversification by the Government Pension Investment Fund (GPIF) — is yet another force investors should keep in mind when thinking about the JPY.
Though I continue to believe that, over the coming months, the JPY is more likely to trade meaningfully stronger as current pricing is already very stretched in valuation terms, I recognize that risk-reward for the near term is in favor of JPY shorts. I see the low 120s in USD/JPY and 150-155 for EUR/JPY as the multi-year peaks for these two crosses.
The basic story
I first wrote about this issue on August 5, 2004 in GPIF Outflows as a Medium-Term Support for USD/JPY. In that note, I argued that the GPIF, which manages some US$1.4 trillion in assets, was on track to invest more than US$100 billion overseas between 2004 and March 2009.
In March 2005, the end-2009 targeted foreign asset holdings were raised from 15% of total assets to 17%. At the current exchange rate, this translates into cumulative outflows of close to US$77 billion between April 2006 to March 2009, for average annual outflows of close to US$26 billion over the next three years.
This GPIF argument is totally separate from the other arguments I’ve presented in recent weeks on USD/JPY and EUR/JPY. It is a multi-year plan that has somehow not attracted as much attention from investors as it should have, in my opinion.
In March 2001, as a part of FILP (Financial Investment and Loan Program) Reform, or the ‘Zaito Reform’, the way in which government-managed pensions (or the pension reserves) were invested was significantly altered. Before the reform, all pension reserves had to be deposited with the MoF’s FILP account. Of these funds, most was then directly lent to various public works projects (called the FILP projects), while the rest (about 15% of the total assets as of end-March 2001, or about ¥10.7 trillion, including investments in Japanese and foreign equities and foreign bonds) was invested in non-FILP assets (e.g., stocks and bonds). The non-FILP investment fund is the Nempuku Fund.
After the Zaito Reform, the Nempuku Fund was reorganized into the GPIF. The key aim was to sever the hard financing link between the pension reserves and the public works projects under the FILP. From 2001 onward, public pension funds would enjoy a great deal more flexibility in the assets it can hold, while the public works projects (the FILP projects) would compete for financing in the open capital markets, like any other project or business, but with government guarantees in some cases. In other words, in theory, the FILP projects would need to, directly or indirectly, issue a large amount of their own bonds in the market, while the public pension funds would be relatively free to decide whether or not to invest in these bonds.
To minimize the ‘shock’ impact of this change — keep in mind that the FILP projects were worth a huge ¥147 trillion (or about US$1.2 trillion) — this transition in the financing scheme was to be spread out over eight years. The aim was that, by end-March 2009, the publicly managed pensions would have no direct financing links with the FILP projects.
The basic ‘benchmark portfolio’ targeted for 2009 includes a healthy size of foreign portfolio holdings. Between April 2006 and March 2009, I calculate that some ¥9.2 trillion (equivalent to close to US$76.7 billion) in new foreign bonds and stocks will be bought, mostly unhedged, which translates into about US$26 billion in outflows in each of the coming three years. In FY2004 and FY2005, net outflows from the GPIF were a bit higher, at some US$35 billion each year. These planned outflows are of a significant size — equivalent to about a third of Japan’s total trade surplus — and should help keep the JPY undervalued.
A bit more detail on the updated calculations
As of March 2001, when the diversification program began, GPIF had ¥173 trillion in total assets, 85% of which was directly invested in the various FILP projects, with another 8.5% invested in JGBs. Not counting ‘short-term assets’, only 6.1% of total assets (or ¥10.7 trillion — the figure I mentioned earlier) was invested in Japanese stocks or foreign stocks and bonds.
As of March 2006, GPIF had ¥161.9 trillion in total assets. Foreign asset holdings had risen to 11.3% (from 2.5% in 2001, and against a target of 11.0%) of total assets. The absolute value of foreign security holdings was ¥18.3 trillion.
For March 2009, the targeted foreign asset holdings are 8.0% in foreign bonds and 9.0% in foreign stocks. This means that, from an initial share of 2.5% of total assets held in foreign securities back in 2001, the aim is to raise these holdings to 17% by 2009. In absolute terms, this corresponds to an increase in foreign security holdings of ¥23.2 trillion (or close to US$200 billion) over the eight-year transitional period.
My thoughts
I have the following thoughts:
1. JPY carry trades over-rated. I am in no way disputing the existence of JPY carry trades. However, I think investors place too much emphasis on this idea, and may miss other factors that need to be considered in thinking about the JPY. In my September 28 piece, Why Is the JPY So Weak?, I pointed out that bond outflows from Japan have not increased with the widening in cash yield differentials vis-à-vis the US and Europe. I also suggested that currency hedging, the ‘Global Funneling Hypothesis’, and now GPIF outflows are also important factors to consider. The prospective EUR/JPY and USD/JPY sell-off I envisage for the coming years will be a gradual one, due to these headwinds for JPY.
2. Fed-BoJ discussion is critical for USD/JPY. Given the importance of the cash yield differentials, what the Fed and the BoJ do is clearly very important. Our official view is that the Fed is likely to hike one more time in early 2007 to take the FFR to 5.50%. For the BoJ, the outlook is at least as uncertain. There will be some technical factors that will perturb CPI in the coming months: changes in cell phone charges may add 0.25% to CPI; the rise in medical costs may add another 0.10% to CPI; but the fall in gasoline prices could reduce CPI by 0.20%. What this means is that, when the BoJ meets in December, CPI (ex-fresh food) may be running at 0.2-0.4%. December seems to be the first opportunity for the BoJ to consider another rate hike, as the next Tankan and the Tokyo November CPI will be out by then. My belief is that the BoJ has a good chance of taking the policy rate to 0.50% at its December meeting. Of course, the next Outlook Report, to be released at the end of this month, will provide important guidance on the BoJ’s policy path. Specifically, the discussion on consumption and wages will be key, as muted wage pressures is the number one economic puzzle in Japan right now.
3. Declining ‘home bias’ a real issue for the JPY over the medium term. In general, declining ‘home bias’ has been a common theme in the world in recent years. I myself have written on the fall in the Feldstein-Horioka S-I coefficient, and this trend is one explanation for why global imbalances have not posed nearly as acute a problem for the USD, the AUD and the NZD as traditionalists may have thought. Even in Japan — a country known for exceptionally high ‘home bias’ — there has been a broad-based increase in interest in and appetite for foreign assets.
4. GPIF outflows do not dictate USD/JPY’s trend. It is important to keep in mind that GPIF outflows are one of several factors to consider. In no way do these outflows dictate the trajectory of USD/JPY. For example, in 2002-04, USD/JPY drifted lower, despite some US$65-70 billion worth of capital outflows from the GPIF. Similarly, US$26 billion in annual outflows from the GPIF should be seen as a headwind for the JPY and not something that will necessarily sink it.
Bottom line
GPIF will continue to invest US$26 billion in each of the next three years in foreign securities. I believe that this will pose yet another headwind for the JPY, cushioning the eventual fall in USD/JPY and EUR/JPY.
