Follow the Money

I liked the lede of this week’s economist story on the Fed, the dollar and oil: “THE spirit of St Augustine hovered over the Federal Reserve this week. “Oh Lord, let us stop cutting interest rates, but not yet” And I liked the analysis even more – perhaps because it draws on an excellent Goldman Research paper by Jens Nordvig and Jeffrey Currie. The Economist alludes what I think is the most important point: dollar weakness and oil strength and both are manifestations of the fact that global growth has been far stronger than US growth. Weak US growth translates into dollar weakness. Strong growth outside the US – particularly in emerging economies – translates into ongoing growth in demand for oil. All the more so because many of the most rapidly growing economies in the world keep domestic oil prices below world market prices (this is true for the oil exporters like Russia and Saudi Arabia as well as oil importers like China), helping to keep demand growth up. Combine growing economies with stagnant supply and, well, prices have to rise to bring demand in line with supply. The impact of strong global growth on oil – and commodities more generally – is one reason why strong growth outside the US doesn’t clearly help to reduce the overall US trade deficit. Strong growth abroad means it is easier for the US to sell more goods abroad. It also increases the price of the United States’ agricultural exports. Alas, those aren’t all that large a share of total exports anymore, in part because a lot of corn is buying converted into ethanol and burned in American SUVs. But strong growth abroad also increases the price of oil. And the US now imports a lot of oil. That means it spends a ton of money on imported oil. Way more than in the 70s in absolute terms, and even relative to world GDP. The Goldman paper suggests that the United States’ energy inefficiency, its modest exports to the oil-exporting region and a reduced willingness on the part of the oil exporting economies to hold dollars – together with the Fed’s tendency to target core inflation while the ECB targets absolute inflation – explains why the dollar has tended to fall when oil rises. Goldman found that the negative correlation with between the dollar and oil holds even if oil is priced in euros or a basket of global currencies – i.e. a high real oil prices contribute to a weak dollar more than a weak dollar contributes to a high dollar price of oil. The Economist: “So is the weaker dollar driving oil prices up or are high oil prices driving the dollar down? The Goldman analysts argue the latter because oil exporters import more from Europe than America and hold less of their oil revenues in dollars. A second factor lies with central banks. Because the Fed focuses on “core” inflation (which excludes food and fuel), whereas the ECB targets overall inflation, America’s central bank runs a looser policy in response to higher oil prices, thus pushing the dollar down.” This negative correlation obviously makes monetary policy particularly difficult for those oil exporters that insist on pegging their currency to the dollar. They are effectively importing a doubly pro-cyclical policy – currency weakness together with low US rates – at a time when high real oil prices are producing a boom. No wonder inflation is now at or above 10% in almost all the big oil exporters that peg to the dollar (or even a dollar-euro basket). Saudi Arabia just recently joined the club, if 9.6% y/y inflation is rounded up. Though I guess it is possible to argue that 10% inflation isn’t that bad, as Saudi inflation was closer to 30% back in the 70s … I am particularly interested in one part of Goldman’s argument – the claim that oil exporters hold less of their oil revenue now in dollars. If there is one thing I would like to know even more than the dollar share of China’s reserves, it is what fraction of the overall increase in the official assets of the oil exporting economies that is going into dollars. We know that Russia is putting a lower share of its assets into dollar now than in 2005. Norway too – though the shift is more modest. Iran and Venezuela have likely moved in the same direction, though they are rather less transparent than Russia and Norway. But we don’t know what the Gulf has been doing. And at current oil prices, the Gulf really matters. The big Gulf sovereign funds (ADIA, KIA, QIA) seem to hold about the same share of their assets in dollars as Russia, or maybe just a bit less. But recently the growth in the assets of the Gulf’s central banks, including the Saudi Arabian Monetary Agency (SAMA), has been faster than the growth in the assets of the Gulf’s sovereign funds. And they likely still hold most of their reserves in dollars. That is why Rachel Ziemba and I argued that the Gulf as a whole hasn’t been able to diversify away from the dollar, even if individual institutions have. However, our argument depends on a lot of assumptions and inferences. We don’t know for sure. And the answer matters now more than ever. The Saudi central bank added $40b to its foreign assets in the first quarter. If oil stays where it is now, the Saudis could add close to the $200b to their foreign assets this year. That is a lot for a country of maybe 25 million people; only a country of 1.3 billion people will do more …

The US likely needs to attract a net capital inflow of roughly $65b a month to finance its current account deficit. The increase in the Fed’s custodial holdings for foreign central banks between March 5 and April 2: $69.8b ($29.2 Treasuries, $40.6b Agencies) The increase in the Fed’s custodial holdings from April 2 to April 30: $66.8b ($40.7b Treasuries, $26.1b Agencies) Over the last week of April alone, central banks added $27.4b to their US custodial holdings, including $18.7b of Treasuries. Pick how you want to do the math. $68.3b in average monthly purchases works out to around $820b a year. $17.1b in average weekly purchases (over the last 8 weeks) works out to more like $890b annually. Either way, it is more than enough to finance the (expected) US current account deficit if US investors don’t add to their foreign portfolios and existing foreign investors don’t abandon the US. Incidentally, the $8.7b in average weekly purchases of Treasuries over the last 8 weeks would – if sustained -- be enough to finance a $454b budget deficit without selling a single Treasury bond to private investors. Sometimes I think the US should drop the façade of auctioning off Treasuries and just negotiate private placements with the People’s Bank of China and the Saudi Monetary Agency. What’s more, all this financing was provided more or less unconditionally, with the United States creditors taking on the risk of future dollar depreciation. Further dollar depreciation against the euro – and, perhaps more importantly, the risk of further dollar depreciation against their own currencies. It goes without saying that this flow is far, far larger than the $30b or so sovereign funds committed to troubled US financial institutions in December and January ($40b if UBS is considered a US financial institution). Yet it has attracted far less attention. For all the rhetoric that sovereigns are stable long-term investors able to take the long-view and buy beat down financial assts (super-senior tranches of CDOs based on mortgage-backed securities anyone), the enormous growth in central banks custodial holdings with the Fed suggest that central banks are buying the expensive, safe financial assets everyone else now wants – not taking a long-view and looking for value in the debt market. The world of sovereign investors is now quite big. Almost unimaginably big – with China and the oil the exporters leading the way. Total sovereign asset growth likely topped $300b in the first quarter. It includes a lot of different institutions pursuing a lot of different strategies. Some clearly have invested in risky assets than sovereigns have typically invested in – whether US financial institutions or European oil companies. But I would bet that if we had real time aggregate data on sovereign flows, the real story of the past six months has been a flight by sovereign investors away from risky assets. That certainly seems to be the story of the last two months, judging by the enormous growth of the Fed’s custodial accounts. But if an informed reader has a different view, I am all ears … Update: For the sake of comparison, I went back and looked at the growth in the Fed’s custodial holdings in the first six months of 2004, a period marked by unprecedented Japanese intervention in the foreign exchange market (at the end of 2003 and in early 2004) that fueled very large purchases of US Treasuries (in the first half of 2004).   The average monthly growth in the Fed’s custodial holdings in the first six months of 2004 was just a bit under $28b -- well under half the monthly growth in the the last two months. There is a real story here.

The BIS computes the real exchange rate for a broad range of countries (if only the data extended backwards a bit further). The BIS data shows that China’s real exchange rate tracked the US real exchange rate pretty close through 2005. That isn’t much of a surprise: China pegged to the dollar. It actually tracked the US real exchange rate quite closely through most of 2006, even after formally moving off the peg. The initial move off the dollar peg was exceptionally timid. Too timid in my view. Over the past year though, China’s real exchange rate path has diverged from the US path. The pace of RMB appreciation against the dollar picked up. And Chinese inflation also picked up. The two have combined to produce a meaningful real appreciation against the dollar -- though not, it has to be emphasized, a meaningful real appreciation against the euro. Chinese exporters have taken notice. Michael Pettis reports that exporters’ complaints are starting to influence the policy debate in Beijing. The influential Li Yang (a former PBoC advisor now at the Chinese Academy of Social Sciences) is now calling for a halt to RMB appreciation against the dollar. Keith Bradsher of the New York Times reports that Chinese firms exporting to Europe now want to be paid in euros -- and Chinese firms exporting to the US want to be paid quickly. Some firms have apparently gotten burnt promising to deliver goods for dollars in the future only to see the RMB appreciate against the dollar before the goods were delivered, cutting into their profits. Fair enough. These stories explain why exchange rate moves do have an impact on trade flows, despite often made arguments to the contrary. If the rise that the RMB will rise against the dollar (or for that matter the euro) makes Chinese producers more inclined to sell their goods for RMB in the domestic market, all the better. That is what should happen: domestic demand should displace external demand as the driver of Chinese growth. Such complaints though strike me as premature. The graph above suggests that China’s real appreciation has only just begun. China’s real exchange rate has tracked the real exchange rates of Japan and the US -- two of the countries with the worst performing currencies over the last five years.   Japan and the US have much weaker economies than China, so the fact that China’s real exchange rate has tracked their real exchange rates down until recently is in some deep sense puzzling.  China’s economy has been stronger than Europe’s, so the ongoing nominal and real depreciation of the RMB against most European currencies but especially the euro is a bit hard to understand on a fundamental basis. Moreover, China’s real exchange rate remains below its level in early 1998. That doesn’t make much sense. China’s economy has been the best performing in the world over the past decade. That should translate into a real appreciation. It strikes me as reasonable to assume that the pace of real appreciation that China experienced in the 1990s should have been expected to continue in this decade, which suggests a large gap between China’s current real exchange rate and its natural rate. The ongoing pressure for real appreciation though is creating very serious policy challenges for the PBoC. Michael Pettis has this right. Frankly, China’s past timidity has put it in a position where it has few good options. MORE FOLLOWS One option is to allow the RMB to appreciate in a gradual way against the dollar. This policy has the advantage of allowing firms time to adjust. They can avoid long-term contracts in dollars, or start hedging their dollar exposure in the forward market. A stronger RMB also will tend to lower import prices -- though this effect is muted if the RMB is rising at a slower pace against the dollar than the dollar is falling, and if commodity prices are still rising in RMB terms. It has the disadvantage of creating unambiguous incentives for anyone who can to hold RMB rather than dollars, particularly when Chinese interest rates are higher than US interest rates. The resulting hot money inflows can fuel rapid money growth and thus create inflationary pressures. This incidentally is a difficult concept to explain. The stronger RMB helps lower the price of imports and thus reduces inflation (at least if the RMB were rising against something other than the dollar). But if too much money comes in to bet on further rises in the RMB, the influx of money can be a source of inflation. The overall impact is ambiguous. One option is to avoid allowing the RMB to appreciate and to hold Chinese rates below US rates. A close variant is a policy of not letting the RMB appreciate by more than the difference between US rates and Chinese rates. This was the policy that China followed immediately after it "depegged" in 2005. But it was a far easier policy to follow back when US rates were higher than Chinese rates. Right now, such a policy would imply cutting Chinese rates so that they are below US rates even as Chinese inflation has picked up -- and limiting the RMB’s appreciation against the dollar even as the dollar depreciates against other currencies. Effectively, it amounts to a policy of allowing real appreciation through inflation. Central bankers in the UAE and Argentina could even provide China with lessons. Another option -- one that many in China seem to be gravitating toward -- is to hold the RMB constant and rely on other policy tools to fight inflation. The PBoC could increase domestic Chinese interest rates. At least to me, PBoC governor Zhou’s remarks recently seem to hint at such an approach. Or China could try to limit inflation by, in effect, controlling prices. China right now is holding domestic petroleum prices below world prices (and in the process encouraging Chinese demand growth) -- a policy that requires China’s state oil companies use their profits from domestic production to offset losses on imported oil. Pettis reports that Xu Zhiman of the NDRC has said that "the government will not increase the price of refined oil or electricity until inflation is brought under control." Such a policy means continued distortions -- with administered prices being only the most obvious. Moreover, it might not work. At least not in the sense of creating incentives to hold dollar rather than RMB. So long as Chinese interest rates are above US interest rates, there is an incentive to hold RMB rather than dollars even if the RMB is stable against the dollar. Another option is a big move in the RMB -- one large enough to end expectations of further moves. But such a move would need to be large in order to really end expectations of further moves. Just think how far the euro has moved against the dollar over the last give years. Any move large enough to be effective would also be disruptive. A large share of China’s economy grew up around an undervalued exchange rate, and could well have trouble competing at market exchange rates. Isn’t that, after all, what all the complaints about the RMB’s still modest appreciation against the dollar imply? My core conclusion is that China should have moved much earlier, as further dollar weakness and a period of low US interest rates were imminently predictable. For too long Chinese policy makers avoided making hard decisions, in part because they hoped that if they waited long enough the dollar would recover. Their problem now is that the dollar in the end only slid further. As they waited, the gap between China’s managed exchange rate and a true market exchange rate only got bigger. China’s policy makers are often said to focus on the long-term. Here though it seems at least to me that a desire to avoid short-term pain led to a series of decisions that created a bigger long-term problem. But the past is the past. Right now, I am in same camp as Wang Tao, Frank Gong and Michael Pettis: the least bad option now is a large one-off revaluation, probably in conjunction with efforts to tighten China’s capital controls. I increasingly though suspect that China will opt for a slower pace of appreciation and price controls.