Monetary Policy

In an earlier post, I added reserve purchases by the world’s major emerging market central banks, Japan and Switzerland to the bonds purchases by the Fed, the BoJ, the ECB and Bank of England. I wanted to highlight that the central banks of the world were buying a lot of U.S. and European bonds before the big central banks started quantitative easing (QE). China and others bought a ton of bonds prior to the global crisis. Emma Smith, an analyst at the Council on Foreign Relations, helped me with the data work for that post; she and I are jointly writing the follow up posts. In addition to looking at the total number of G-4 bonds bought by the world’s central banks—counting bonds bought in large scale asset purchase programs (QE) alongside estimated reserve purchases—it is interesting to look at central bank purchases by currency. QE results in the purchase of your own country’s bonds; reserve purchases mean you need to invest in bonds issued by someone else—e.g. both the Fed and the PBOC have bought large quantities of U.S. Treasuries and Agencies at different times over the last fifteen years. Take central bank purchases of dollar bonds. The chart below relies on the Fed’s data on its purchases, and an estimate of the dollar bond purchases implied by global reserve growth. Before the global crisis, central bank purchases of dollar bonds came from reserve managers. Their accumulation of dollar assets picked up from around 2003—coinciding with the dollar’s depreciation against the euro, the beginning of the rise in China’s current account surplus and a pickup in capital flows to a range of emerging economies. In early 2008, the Fed was actually selling a portion of its bond portfolio—it didn’t want its balance sheet to expand as its lending to the world’s banks rose in the run-up to the global crisis—and after Lehman, reserve managers started to sell. But the Fed soon reversed course, and started purchasing large amounts of Treasuries and Agencies in its QE programs. And emerging economies recovered and resumed large scale intervention of their own—albeit at a lower level than pre-crisis—taking central bank demand to new highs. Post QE3, though, central banks’ direct purchases of dollar bonds collapsed. The Fed stopped buying, and emerging market central banks have on net been selling. The current (low) yields on ten-year Treasuries aren’t a function of a direct central bank bid. The story for euro and yen bonds though is almost the opposite of the story for dollar bonds. Purchases by reserve managers were never as important, and in the past few years, both the ECB and the BoJ have ramped up their direct purchases. Before the global crisis, reserve demand for euro-denominated bonds was around a third of the level of demand for U.S. dollar-denominated bonds. Reserve purchases of euros continued through 2012, and any slowdown in their demand was made up for by the ECB’s initial purchases of bonds (through the SMP). There was a drought in total central bank purchases from 2012 to 2014 (part of the reason why the euro stayed high?). And then, obviously, a spike associated with the ECB’s QE program. The scale of the ECB’s purchases far exceeds the estimated scale of euro-denominated bond purchases by the world’s reserve managers. Net central bank demand for euro assets thus is at a long-term high. Yen and sterling-denominated assets have never played all that large a role in reserve managers’ portfolios. For the yen in particular, central bank demand for bonds has essentially come from the Bank of Japan. And I should note that the estimated central bank reserve purchases are just that—estimates.  Much of the flow comes from an assumption that central banks that do not report the currency composition of their reserves (at least not historically) to the IMF have maintained a constant yen share of their reserves. With the yen’s fall from 2012 to 2015, that required some purchases… Reserve managers were more interested in the pound pre-crisis. Actually until fairly recently! But the scale of reserve inflows was still dwarfed by the Bank of England’s asset purchases in 2009-10 and 2011-12. (Note the scale has changed from previous graphs.) While central bank purchases of sterling disappeared in the past few years, they obviously are now poised to pick up. After the Brexit vote, the Bank of England announced £70 billion ($90 billion) of gilt and corporate bond purchases. Is there a big story here? I suspect so, but it is also a complex story—and a more complex story that I think is widely recognized. A few points. The data here shows the flow of central bank purchases, not the stock. The stock of bonds held by central banks may matter more for rates. And that stock is going up everywhere but for the United States. And while there isn’t a central bank bid for U.S. dollar bonds right now, the stock held by reserve managers and the Fed isn’t going down that much, as the Fed is keeping its holdings constant. U.S. yields are also a function of the market’s expected course for the Fed. And by the echo of the central bank bid outside the United States. In a subsequent post Emma and I will show that the BoJ and the ECB’s purchases far exceed issuance (something that wasn’t true for the US at the peak of “QE”), so the BoJ and ECB are reducing the stock of bonds in the market’s hands.  Some of that feeds into demand for other assets in Japan and Europe, some of it feeds into demand for U.S. bonds. That is what the TIC data in the U.S. is also telling us. Central banks—reserve managers—are selling U.S. bonds. The net inflow that supports the U.S. current account deficit is a function of Americans who previously held European and Japanese bonds selling and bring the proceeds home (and who prefer the yield on Treasuries to the yield on bunds and JGBs) and purchases of U.S. bonds by private investors in Japan and Europe. Think of low rates in Europe and Japan as a requirement, in equilibrium, for European and Japanese savings to leave their home regions and take currency risk abroad. It all adds up, as it must. But it is a slightly different equilibrium than the equilibrium in the pre-crisis world, where emerging market central banks took on a lot of the world’s cross-border currency risk.   Their large scale reserve purchases were a function of both their underlying current account surpluses and net private inflows, as investors were moving money into fast growing emerging economies. Today a lot more of the currency risk associated with the United States ongoing need for financing (the U.S. still runs a current account deficit) is being taken on by private investors. All this said, our goal here is not to put out a fully formed theory—we don’t have one. Our goal is simply to illustrate how to combine the reserves data with data the major central banks report about their own balance sheets. Of course the implicit argument is that the “market distortions” from “QE” likely preceded “QE”, as there isn’t a strong theoretical reason why PBOC purchases that remove U.S. bonds from private market hands have a widely different impact than Fed purchases that do the same thing (abstracting from any signaling about the course of U.S. rates in the Fed’s purchases, of course).

China’s headline reserves dipped by about $19 billion in September, dropping below $3.2 trillion. Adjust for foreign exchange changes, and the underlying fall is widely estimated to be a bit more—around $25 billion. Press coverage emphasized that the fall “exceeded expectations.” To me that suggests “expectations” on China’s reserves aren’t formed in all that sophisticated a way. $20-30 billion in sales is in line with the change in the PBOC’s balance sheet in July and August (the FX settlement data, the other key proxy for intervention, suggests more modest sales in August). Throw in the September spike in the Hong Kong Inter-bank Offered Rate (HIBOR) —which suggested a rise in depreciation pressure on the CNY and CNH —and $25 billion in sales is if anything a bit smaller than I personally expected.* Of course, some of the sales could be coming through the state banks; time will tell. Even if the pace of sales did not pick up in September, there is is an interesting story in the Chinese data. The $75 billion a quarter and $300 billion a year pace of sales implied by the July-September monthly data aren’t anything like the pace of sales at the peak of pressure on China’s currency. But $75 billion a quarter is a still bit higher than the underlying pace of sales in Q2. The balance of payments data show Q2 reserve sales of about $35 billion (the change in the PBOC’s balance sheet reserves was $31 billion). But other parts of China’s state added to their foreign assets in Q2. In fact, counting shadow intervention (foreign exchange purchases by state banks and other state actors), I actually think the government of China’s total foreign assets may have increased a bit in the second quarter. There are a couple of line items in the balance of payments that seem to me to be under the control of the state and state actors. Most obviously, the line item that corresponds with the PBOC’s other foreign assets ("other, other, assets" in balance of payments speak: up $12 billion in q2, after a bigger rise in q1). But most portfolio outflows are likely from state-controlled institutions (portfolio debt historically has been the state banks, portfolio equity historically has been the China Investment Corporation and the state retirement funds in large part). If these flows are netted against reserve sales, there wasn’t much of a change in q2. In my view, shifts in assets within the state should be viewed differently than the sale of state assets to truly private actors. To get a positive number in q2, though, you need to add in the buildup of foreign assets associated with the increased foreign lending of the state banks (this adjustment is the most debatable). I suspect that the bulk of the China Development Bank’s outward loans are in the banking data, and thus the loan outflow should be viewed as a policy variable (China for example looks to have shut down this channel in q4 2015). Offshore loans were up about $25 billion in q2—a bit over the five year (2011 on) average of around $15 billion a quarter. That pushes my estimate for the total accumulation of foreign assets by China’s state, counting policy lending, into positive territory. Q2’s balance of payments data paints a picture of relative stability. I suspect that my broader metric for Q3 will show a fall in q3. And if that fall is eventually confirmed,** there is a question of why pressure picked up. China’s trade accounts show a substantial surplus (a very substantial surplus on the goods side, and a decent surplus on goods plus travel and tourism—the non-tourism service account is close to balanced). In volume terms, Chinese export growth has picked up—with y/y growth since April on average of 5 percent.*** That is better than the overall expansion of global trade. The pressure is all from the financial account. Interest rate differentials have shrunk, but are still in China’s favor. But the interest rate differential now can easily be dwarfed by exchange rate expectations. Over the past 14 months, the yuan has fallen by 8 percent against the dollar. My guess is that expectations for further depreciation picked up over the course of q3. The yuan appreciated a bit against the dollar from February through May (while depreciating against the basket). But the yuan depreciated against the dollar after the Brexit vote —and ticked down again a bit in late August. That, in my view, contributed to the expectations that China’s authorities are looking to continue the yuan’s depreciation—at least against a basket—after a temporary pause around the G-20 Hangzhou summit and the final SDR decision. Moves against the dollar still seem to have a disproportionate impact on expectations. Note: This chart has been updated to reflect data through 10/13/2016 It is not unreasonable for the market to think that the yuan’s future moves against the dollar will be asymmetric. If the dollar weakens against the major floating currencies, China may want to follow the dollar down. And if the dollar strengthens, maintaining stability against the basket—let alone maintaining a depreciating trend against the basket—implies a further depreciation in the yuan against the dollar. The implication of this view is that the market is (still) betting on where it thinks Chinese policy makers want the currency to go. As long as the market thinks China wants to depreciate one way or another against a basket after Hangzhou and the SDR decision, outflow pressure will continue. The alternative view is that Chinese residents want to get out of Chinese assets independently of the expected path of the yuan, and that the controls put in place in the spring are starting to show a few more leaks. The weaker fix on Monday doesn’t really settle this debate; the fix was below the symbolically important 6.7 level against the dollar, but was also broadly consistent with maintaining stability against the CFETS basket. There isn’t yet enough information to determine if China’s current policy goal is stability, or a stable pace of depreciation. * We don’t know the currency composition of China’s reserves, or the precise way changes in the value of China’s bond portfolio enter into headline reserves. The noise in headline reserves goes up when the expected change is small v the size of the stock, given all the other moves that can impact the stock. Plus or minus $10 billion in headline to me is noise. I prefer the proxies for intervention, which are less influenced by valuation. ** I am waiting for broader measures of sales for September; all analysis for now is contingent on confirmation by subsequent data releases. *** The simple average of monthly y/y changes in export volumes for 2016 is just below 4 percent; a bit higher than than the simple average of monthly changes in import volumes. Data is for goods only, and the y/y changes are distorted in q1 by the lunar new year. Export volumes are up even with softness in U.S. imports from China (setting finished autos aside)

Folks in the market like to talk about what is happening to China’s forward book. Some think that China (or its state banks) sold dollars forward last fall, and, well, to many the (modest) disclosed short position that China reports in the IMF’s SDDS reserves template isn’t all that convincing. In part because the disclosed forward book never changes much.* But China also quite clearly isn’t the only country in Asia with a forward book. "Shadow intervention" is actually rather common, in both directions. At the end of August, Korea had bought about $48 billion in dollars forward, up from just under $45 billion in July.** Technically, the forward book may be the forward leg of a swap contract.*** No matter—the rise in the forward book clearly reflects the central bank’s activities in the market. Adding in the forward book shows the true scale of Korea’s intervention in August. The balance of payments reserve outflow was just over $3 billion. The balance of payments number should track valuation-adjusted headline reserves. The forward book rose by a bit more $3.1 billion. I like to watch government deposits and government bond purchases too; they are up $1.7 billion (with a big increase in government deposits abroad). Korea’s intervention hasn’t always only appeared on the central banks’ balance sheet (though some of the portfolio debt comes from Korea’s National Pension Service). Sum it up, and Korea’s government could have bought as much as $8 billion in the market in August. Annualized, that is a 7 percent of GDP pace of purchases (China’s annualized August sales, judging from the intervention proxies, ran 2.5-3 percent of GDP). It is also the highest pace of purchases since the spring of 2014 (Korea did sell at a faster pace last August, and also clearly was selling earlier in the year). Adding in the National Pension Service’s ongoing purchases of foreign equities would bring the pace of government purchases of foreign assets to over $9 billion, and 8-9 percent of GDP. Obviously it is only one month, and the monthly numbers are volatile. But the scale of the monthly purchases in August also matters; it shows commitment. Reuters reported earlier this week: "traders remained on guard against possible intervention by authorities to stem further gains [in the won]." And I suspect Korea has made its point to the market (1090 is strong enough, thank you very much), and September’s intervention will be a bit smaller. Headline reserves for September should be out soon. Update: Korea’s reserves rose $2.3 billion in September, to $377.8 billion. Valuation changes shouldn’t be a big factor so, based on headline reserves, Korea looks to have continued its reserve purchases. But it is impossible, given Korea’s tendency to also add to its forward book and to increase the foreign assets held in other parts of the government, to know Korea’s full activities in the market from the headline change in central bank reserves alone. See the August data! * The short position has been a constant $28.9 billion since February (there is no data before that). ** Call me a grumpy old man if you want, but in my opinion the redesigned web page for rediffusion of the national data in the SDDS reserve template on the IMF’s web page has made it more difficult to find (or to find quickly) the national data on forwards (Korea is here, more up-to-date data comes from Korea’s own web page). The old format did not allow online cross-country comparison, but it generally was easy to use for my purposes—the national details were readily available. Now a lot of the data seems only to be available in a cross-country table by indicator. The new presentation of the national data shows headlines reserves and a few graphs (though the time series on the graphs is far too short) —but the real value of the SDDS disclosure comes from the disclosure of forwards and other details, not the headline numbers (which most countries present on their own). It would be a great help if there was a simple way to expand the view on the national pages to include the reserves detail and to download the full historical time series (available in the past as a .csv file—old fashioned, but it worked, without requiring a complex query to the data base). *** Think spot purchases, with the dollars then swapped for won with the banking system. The reported forward is the forward leg of the swap. The central bank retains the foreign currency risk. The banks get dollar funding.

   

The proxies that provide the best estimates of China’s actual intervention in the foreign currency market in August are out, and they in no way hint at the stress that emerged in Hong Kong’s interbank market in September. The PBOC’s balance sheet shows foreign currency sales of between $25 and $30 billion (depending on whether you use the number for foreign currency reserves or for foreign assets). A decent sum, but also a sum that is consistent with the pace of sales in July. SAFE’s data on foreign exchange settlement, which in my view is the single best indicator of true intervention even though (or in part because) it aggregates the activities of the PBOC and the state banks, actually indicates a fall-off in pressure in August. The FX settlement suggests sales of around $5 billion in August. Even after adjusting for reported changes in forwards (the dashed line above). All this said, there is no doubt something changed in September. The cost of borrowing yuan offshore spiked even though the exchange rate has been quite stable against the dollar and generally stable against the CFETS basket. Two theories. One is that China that the market thinks China will find a way to resume the yuan’s slow slide against a basket of currencies of its major trading partners after the G20 summit, even if that means additional weakness against the dollar. There was a widespread belief in the market—and among analysts who watch such things—that China would not allow a significant move in the market before the G20 summit. Now all bets are off, or will come off after the yuan is formally included in the SDR in early October.* The spike in offshore yuan interest rates thus reflects a true rise in speculative pressure against the yuan, one that the PBOC is resisting. Saumya Vaishampayan and Lingling Wei of the Wall Street Journal: "Suspected intervention by Chinese banks in what’s known as Hong Kong’s “offshore” market has led to a surge in the cost for banks in the territory to borrow yuan from each other. Investors and analysts believe the intervention—which they say has likely come at the behest of China’s central bank—is aimed at thwarting bets against the Chinese currency, also known as the renminbi. The suspected heavy buying by Chinese banks has helped squeeze a market China had tried to foster just a few years ago as it looked to promote the yuan as a major international trading currency." The other is that the PBOC has tightened offshore yuan liquidity for reasons of its own (not necessarily in response to a rise in speculative pressure), in part by putting pressure on the state banks not to roll over maturing forward contracts.* Only the PBOC, and perhaps the the Bank of China, knows for sure. I do though suspect that China is likely to have to show a bit more of its hand in the foreign currency market relatively soon. Does China manage for stability against a basket, or manage for a depreciation against the basket? Has the CNY depreciated by enough, or do the Chinese authorities want a larger move? How much weight does it put on the dollar versus the basket when push comes to shove? As many have noted, the broad effective value of China’s currency has slid pretty steadily this year—though there was a bit of a pause in August. Until recently, that slide was consistent with a yuan that was only a bit weaker against the dollar than in January (the move from 6.6 to 6.7 came in the face of the Brexit shock; it didn’t appear to be a unilateral Chinese move). Call it the Chinese currency version of Goldilocks Now, well, a further depreciation of the yuan against the basket might mean testing the post-Brexit lows against the dollar. And the yuan is getting to be within shouting distance to its level against the dollar during the 2008-2009 repeg. It doesn’t take all that much imagination to realize that reversing 8 years of appreciation against the dollar could matter politically as well as economically. The trade "fundamentals" to my mind do not provide a strong case for further weakness in the yuan against a basket of currencies. Even with weak Chinese exports to the United States, the Chinese data on export volumes shows modest year-over-year growth. August export volumes appear to be in line with recent trends; a reasonable estimate suggests Chinese exports continue to grow a bit faster than would be implied if China’s exports were growing with global trade.** I personally do not think China can expect to go back to the days when Chinese export growth significantly exceeded global trade growth (see Box 2 of this ECB paper). Of course, the trade data also isn’t the only factor that drives currency markets. * A small technical point. The Mexican peso plays a much bigger role in the U.S. dollar’s broad exchange rate than it plays in China’s basket. And the peso is weak right now. That is one reason why the CFETS index might diverge from the dollar index. The dollar/euro and dollar/yen rates have been relatively stable in September, though of course that could change. ** I am focusing on the data on volumes, not the nominal number. If export prices did not change in August, August export volumes increased year-over-year. The official Chinese data will be out in a few more days.

Just before the global financial crisis, I wrote a paper on the geostrategic implications of the United States’ growing external debt—and specifically about the fact that the U.S.’s main external creditors were increasingly the reserve managers of other states, not private investors. Yes, there were large two-way gross private flows in the run up to the crisis; think U.S. money market funds lending to the offshore arms of European banks who in turn bought longer-term U.S. securities. But, on net, the inflows needed to sustain the United States’ external deficit from 2003 on mostly came from the world’s big holders of reserves and oil exporters who stashed funds away in sovereign wealth funds. With hindsight, I, and the others who speculated about how China’s Treasury holdings might be used for political leverage over-egged the pudding, as Dan Drezner, among others, has pointed out. Greece’s indebtedness to private bond holders and banks proved a bigger constraint on its economic sovereignty than the debt the United States owes to the PBOC and other official investors. Germany was the creditor country that ended up with the leverage, not China. And thinking back even further, Britain’s geostrategic vulnerability to the withdrawal of U.S. financing in the Suez crisis derived from its commitment to maintaining the pound’s external value. Letting the pound float was inconceivable at the time. That as much as anything gave the U.S. leverage over Britain. Worth remembering. I could argue that the global crisis reduced the United States’ need for all kinds of external financing significantly, which is true—and that the leverage that comes from the perception that China could rattle markets in times of stress has not entirely gone away. But it is also true that before the crisis I underestimated the Fed’s ability to influence term premia and the path of long-term U.S. interest rates independently of inflows from foreign creditors.* Call it the geostrategic impact of QE. Yet some of the more subtle aspects of my argument about the strategic consequences of the rise of new large state creditors have, I think, stood up to the test of time. One argument was that a major creditor could have an impact on the U.S. without actually selling dollars, just by moving their dollar portfolio around. And it is quite clear from former Treasury Secretary Hank Paulson’s memoir that the risk of a Chinese/ Russian funding strike of the Agencies in the summer of 2008 was something that worried U.S. policy-makers.** Robert Preston reported back in 2014: "I [Paulson] was talking to them [Chinese ministers and officials] regularly because I didn’t want them to dump the securities on the market and precipitate a bigger crisis ... I was meeting with someone … This person told me that the Chinese had received a message from the Russians which was, ’Hey let’s join together and sell Fannie and Freddie securities on the market.’ The Chinese weren’t going to do that but again, it just, it just drove home to me how vulnerable I felt until we had put Fannie and Freddie into conservatorship ..."** Another argument was that rising reserves would give the world’s new group of creditors “soft” financial power. I wrote: “Today, emerging economies ... not only do not need the IMF; they increasingly are in a position to compete with it. .... Chinese development financing provides an alternative to World Bank lending. Asia is exploring the creation of a reserve pool that could serve as a precursor to a regional monetary fund. If a small emerging economy got into trouble now, it undoubtedly would seek regional financing on more generous terms than those offered by the IMF." That doesn’t quite describe the Asian Infrastructure and Investment Bank. But it isn’t that far off either. Substitute "development finance" for regional financing and "World Bank" for the IMF. The reality is that the world can form financial coalitions of the willing without the participation of the U.S.. Even with the fall in China’s reserves and the strain that low commodity prices have placed on many commodity exporters. Scott Morris of the Center for Global Development has written a new CFR discussion paper on how the U.S. should respond to China’s success in setting up the Asian Infrastructure and Investment Bank. One of his conclusions is straight-forward. If the U.S won’t support an expansion of the balance sheet of the institutions where it has the most influence and weight—institutions like the World Bank—then the world will likely proceed without the United States. And the current “core” development institutions will over time be surpassed by new institutions where the U.S. has less influence. * I imagined a more convoluted path to stability, drawing a bit on Dooley, Garber and Folkerts-Landau. The direct path would be for the Fed to buy what others were selling. I though focused on an indirect path: European central banks, concerned about a rise in their currencies, would intervene and recycle funds back into the U.S. bond market. ** Technically, the Fed could—and in the end did—offset the loss of Chinese demand for the bonds of the Agencies. The Treasury though needed to step in to provide the Agencies with the capital needed to absorb losses on their mortgage portfolio. Dan Drezner argues China’s influence on the Agencies wasn’t decisive—which is fair. The Agencies had sold a lot of bonds to domestic investors as well. But it is also quite clear that China’s holdings weighed heavily on the minds of the relevant decision makers.

I got my start, so to speak, tracking global reserve growth and then trying to map global reserve flows to the TIC data. So I have long thought that large scale central bank purchases of U.S. Treasuries and Agencies, and German bunds, and JGBs didn’t start with large scale asset purchase programs (the academic name for “QE”) by the Fed, the ECB and the BoJ. Before the crisis, back in the days when China’s true intervention (counting the growth in its shadow reserves) topped $500 billion a year, many in the market (and Ben Bernanke, judging from this paper) believed that Chinese purchases were holding down U.S. yields, even if not all academics agreed. The argument was that Chinese purchases of Treasuries and Agencies reduced the supply of these assets in private hands, and in the process reduced the term/risk premia on these bonds. Bernanke, Bertaut, Pounder DeMarco, and Kamin wrote: "...observers have come to attribute at least part of the weakness of long-term bond yields to heavy purchases of securities by emerging market economies running current account surpluses, particularly emerging Asia and the oil exporters .... acquisitions of U.S. Treasuries and Agencies took these assets off the market, creating a notional scarcity that boosted their price and reduced their yield. Because [such] investments were for purposes of reserve accumulation and guided by considerations of safety and liquidity, those countries continued to concentrate their holdings in Treasuries and Agencies even as the yields on those securities declined. However, other investors were now induced to demand more of assets considered substitutable with Treasuries and Agencies, putting downward pressure on interest rates on these private assets as well." I always thought the mechanics for how the Fed’s QE impacts the U.S. economy—setting aside the signaling aspect*—should be fairly similar. Both reserve purchases and QE salt “safe assets" away on central banks’ balance sheets.** There are now a number of charts illustrating how the ECB and BoJ have kept central bank bond purchases high globally even after the Fed finished tapering. Emma Smith of the Council’s Geoeconomics Center and I thought it would be interesting to add reserve purchases to these charts and to look at total purchases of U.S., European, Japanese and British assets by the world’s central banks over the last fifteen years—adding the emerging market (and Japanese and Swiss) purchases of G-4 bonds for their reserves to the bonds that the Fed, the ECB, the Bank of Japan and the Bank of England bought. Obviously there are important differences between balance sheet expansion done through the purchase of foreign assets (reserve buildup) and balance sheet expansion done though the purchase of domestic assets (quantitative easing).  One is aimed at the exchange rate, the other at the domestic economy. But if portfolio balance theories are right, the direct impact on bond prices from foreign central bank purchases of bonds and from domestic central bank purchases of bonds should I think be at least somewhat similar. As the chart above makes clear, large-scale central bank asset purchases in some sense preceded the global crisis.  Pre-crisis, the purchases came from emerging market central banks. China, counting its shadow intervention, bought about 15% of its GDP in reserve assets from mid 2007 to mid 2008; as a share of GDP, that is a pace of asset purchases equal to the BoJ’s current pace. More recently, reserve managers have quite obviously been selling. The reserves the PBOC bought from 2012 to 2014 were sold in 2015 and 2016. Yet with three trillion plus in reserves, and more in its shadow reserves, the PBOC hasn’t yet sold anything it bought prior to the global crisis. And more importantly, recent reserve sales have come at a time when the BoJ and ECB have been ramping up their purchases. Aggregate central bank demand for safe assets has remained elevated. There has been a significant shift in the currency composition of large scale central bank assets purchases in the past couple of years. In fact, adding reserve purchases to QE purchases magnifies the recent global shift in central bank demand away from dollars. EM reserves managers have been selling dollar assets. At the same time the Fed isn’t buying assets anymore. And the G-7 currency agreement has limited the ECB and (especially) the BoJ to the purchase of their own assets, so there has been a surge in central bank purchases of euro and yen denominated bonds. Prior to the crisis, I was never completely convinced by arguments that the currency composition of the reserves held by central banks didn’t matter. Partially that was self-interest of course; I had spent a lot of time developing my reserve tracking technology. But I also didn’t think private investors were completely indifferent to the currency composition of their portfolio of safe assets. If the central banks accumulating reserves wanted euros rather than dollars, the euro would need to rise relative to the dollar to make "safe" dollar assets cheaper and thus more appealing to private investors. Treasuries and German bunds were not perfect substitutes. Nor for that matter are Agencies and French government bonds. I still think that is the case. It matters that the ECB conducts its asset purchases by buying euro assets, not by buying Treasuries. At the same time, with ECB purchases exceeding euro area government net issuance and with BoJ purchases far exceeding net new issuance of JGBs, it is hard to argue that some investors in European and Japanese debt haven’t been pushed into U.S. Treasuries and Agencies. (This Banque de France working paper found that foreign holders of euro-denominated bonds were the most likely to sell to the ECB.) And thus there should be some impact on the U.S. yield curve from the actions of other central banks even when those central banks aren’t buying dollars. All this said, the indirect impact on Treasuries from the purchase of JGBs and Bunds is likely to be smaller than the direct impact of central bank purchases of Treasuries.** Krishnamurthy and Vissing-Jorgensen’s 2013 Jackson Hole paper argued that Treasuries and Agencies weren’t perfect substitutes, which is why the Fed’s purchases of Agency MBS had an impact.  I have long thought the QE-Agency MBS-lower mortgage payments channel provided an important offset to the fiscal tightening the U.S. did in 2013, as mortgage refinancing put cash directly in the pockets of many households. And the current, relatively low level of direct central bank demand for U.S. Treasuries reinforces Larry Summers’ argument that low Treasury rates now aren’t just a reflections of central banks purchases... In subsequent posts, Emma and I plan to look into these points in a bit more detail, by disaggregating central bank flows by currency and comparing central bank demand (the sum of "foreign” reserve demand and “home” monetary policy demand) to government issuance. A note on methodology. Asset purchases by the Fed, ECB, BoJ and BoE are easy to track directly. Reserve purchases come from summing the foreign exchange reserves of around 60 countries—a broad sample that replicates the IMF COFER data. China’s other foreign assets are added to the total (it doesn’t matter much, but it is a point of pride for me—these are assets on the PBOC’s balance sheet that walk, talk and quack like reserves). The dollar share of reporting economies is assumed to be replicated across the full sample (broadly speaking this produces the same result as assuming a constant 60 or 65% dollar share over time). * Signalling here means signalling a credible commitment to hold policy rates low for a long time, i.e. enhancing the credibility of forward guidance. Michael Woodford, for example argued back in 2013 that this was the transmission channel that mattered. ** Technically, a central bank buying bonds creates bank reserves when it credits the account of the seller of the bonds with cash. QE reduces the supply of safe longer-dated bonds in the market (raising their price, and reducing their yield in theory) while adding to the supply of safe short-term assets (claims on the central bank).

Foreign exchange intervention has long been one of those things that works better in practice than in theory.* Emerging markets worried about currency appreciation certainly seem to believe it works, even if the IMF doesn’t.** Korea a few weeks back, for example. Korea reportedly intervened—in scale and fairly visibly—when the won reached 1090 against the dollar in mid-August: "Traders said South Korean foreign exchange authorities were spotted weakening the won "aggressively," causing them to rush to unwind bets on further appreciation. On Wednesday (August 10), according to the traders, authorities intervened and spent an estimated $2 billion when the won hit a near 15-month high of 1,091.8." And, guess what, the won subsequently has remained weaker than 1090, in part because of expectations that the government will intervene again. And of course the Fed. And that is how I suspect intervention can have an impact in practice. Intervention sets a cap on how much a currency is likely to appreciate. At certain levels, the government will resist appreciation, strongly—while happily staying out of the market if the currency depreciates. That changes the payoff in the market from bets on the currency. At the level of expected intervention; appreciation becomes less likely, and depreciation more likely.*** 1090 won-to-the-dollar incidentally is still a pretty weak level for the won, even if the Koreans do not think so. The won rose to around 900 before the crisis, and back in 2014, it got to 1050 and then 1000 before hitting a block in the market. In the first seven months of 2016, the won’s value, in real terms, against a broad basket of currencies was about 15 percent lower than it was on average from 2005 to 2007. My guess is that Korea’s practice of intervening to cap the won’s appreciation at certain levels—and systematically trying to keep the won weaker than it was from 2005 to 2008 is part of the reason why Korea runs such a large current account surplus. Just a hunch. Korea’s tight fiscal policy no doubt contributes to Korea’s large external surplus as well (Korea’s social security fund runs persistent surpluses, surpluses that generally have more than offset the government’s small headline deficit). Intervention and tight fiscal sort of work together. The general government surplus acts as a restraint on domestic demand. And intervention helps sustain a large export sector that offsets weakness in internal demand. [*] More specifically sterilized intervention (sterilized intervention means the government buys foreign currency, but offsets the domestic monetary impact of its foreign exchange purchases—whether by raising reserve requirements, issuing central bank paper, or by selling domestic assets). Unsterilized purchases of foreign exchange are a monetary expansion, and thus always should have an exchange rate impact. This San Francisco Fed note is a bit old, but it gives a clear summary of the standard argument while arguing that intervention can have an impact; this Cleveland Fed paper is typical of the view that sterilized intervention doesn’t have an impact. [**] The IMF formally believes—if belief is defined by what is in its workhorse current account model—that foreign exchange intervention only has an impact when a country’s financial account is largely closed. For example, even extremely large scale intervention by say Japan technically would have no impact on Japan’s current account, as Japan’s financial account is considered fully open, and intervention is interacted with the coefficient for openness. Korea’s capital account isn’t fully open, but it is pretty open—so 1 percent of GDP in intervention would, in the IMF’s model estimates, I think have a maximum impact of around 5 basis points (0.05 percent of GDP) on Korea’s current account (the 0.45 coefficient on reserve growth times the 0.13 coefficient on the capital account). However, intervention is only judged to be policy relevant by IMF if there is a gap between the country’s level of reserves and the optimal level of reserves, and if there is also a gap between the country’s controls and the optimal level of controls. As the IMF doesn’t think Korea should throw open its financial account just yet, Korea’s intervention is effectively zeroed out. Joe Gagnon, Taim Bayoumi and Christian Saborowski found a bigger impact from intervention in their 2014 paper: “each dollar of net official flows raises the current account 18 cents with high capital mobility and 66 cents with low capital mobility, for an average effect of 42 cents.” The impact of an incremental 1 percent of GDP in reserve purchases by Korea is thus about 20 basis points on the current account, plus an additional impact from the legacy of past intervention. A high stock of reserves has an impact in the Bayoumi, Gagnon and Sabrowski model on the current account balance of high capital mobility countries. Korea has well over $400 billion in reserves if you include its $44 billion forward book, or very roughly around 30 percent of GDP; with a stock coefficient of .03 that gives a stock impact of around 1 percent of GDP. Other Gagnon estimates point to a slightly bigger impact of intervention. My guess is that Korea’s long record of intervening to cap appreciation together with its less than complete financial openness combine so that Korea’s actions have a bigger impact than implied by the Gagnon, Bayoumi, and Sabrowski estimates, as past intervention makes Korea’s action in the market more credible (e.g. the market believes that if Korea really wants to stop the won from appreciating it will buy in scale, so it doesn’t test the authorities too much once they show their hand).

China is running a persistent current account surplus, one that could be larger than officially reported (the huge tourism deficit looks a bit suspicious). If China paid off all its external debt, it would still have around $2 trillion in reserves.* If it paid off all its short-term debt, it would have $2.5 trillion in reserves. And China has a very low level of domestic liability dollarization (3 percent of total deposits are in foreign currency) True, $3.2 trillion ($3.3 trillion if you include the PBOC’s other foreign assets, as you should, and as much as $3.5 trillion if you include the China Investment Corporation’s foreign portfolio, which is more debatable) isn’t $4 trillion.** But much of the fall in reserves over the last 18 months has stemmed from the use of reserves to repay China’s short-term external debt. The IMF projects that China’s short-term external debt will have fallen from $1.3 trillion in 2014 to just over $700 billion by the end of this year. Reserves are down, but—from an external standpoint—China’s need for reserves is also down. The two year fall in short-term debt is actually about equal to projected drop in reserves. The Fund though sees things a bit differently. Buffers, according to the Fund’s staff report, are now low, and need to be rebuilt. Some in the market agree. And that gets at a critical issue for China, and a critical issue for assessing reserve adequacy more generally. Just how many reserves do countries like China, need? For China, two “traditional” indicators of reserve adequacy—reserves to short-term debt and reserves to broad money—point in completely different directions. Reserves are over 400 percent of short-term debt (way more than enough). *** Reserves are now “only” 15 percent of broad money (not enough; 20 percent of M2 is the traditional norm). You cannot really fudge the difference; you have to tilt one way or the other. (Hat tip to Emma Smith of the Council on Foreign Relations’ Greenberg Center for Geoeconomics, who prepared the charts). I personally put more stock on balance sheet indicators, and reserves to short-term external debt is the most important. From a balance sheet point of view, there is also a strong case for paying attention to reserves relative to domestic sight deposits (a measure of liability dollarization). The Fund though sees things differently; in the design of the new reserve metric the Fund leaned strongly against balance sheet indicators of reserve need (see the discussion of liabliity dollarization here), and instead went with a composite of the three traditional indicators (short-term debt, m2, and imports—though the Fund prefers using exports), with an additional factor for longer-term external liabilities. The Fund’s reserve metric effectively says China needs to hold more reserves, relative to its GDP, than a typical emerging economy. Especially if China opens its financial account before its currency floats freely. And that is the case even though China has much less external debt than a typical emerging economy, and also has less liability dollarization of its financial system. Personally, I think foreign currency deposits pose more risks than domestic currency deposits, and, to the extent a country should hold reserves against the risk of domestic capital flight, those countries with more domestic foreign currency deposits should hold proportionately more reserves. The Fund’s metric, though, explicitly doesn’t adjust for liability dollarization. As a result, the Fund believes China needs to hold far more reserves against the risk of domestic capital flight than other emerging economies, including emerging economies with a lot more domestic FX deposits. The gap between the reserves China needs to hold on the Fund’s metric and the reserves other emerging economies need to hold is even more extreme in dollar terms; as the Fund’s metric requires the biggest emerging economy to hold more reserves than other emerging economies—remember that 20 percent of M2 for China is 40 percent of China’s GDP, or well over $4 trillion.**** One hundred percent of short-term debt by contrast requires reserves of $750 to $900 billion now. I will though give the Fund credit for now recognizing one obvious implication of its analysis of reserve adequacy: China needs to proceed cautiously on financial account liberalization, and the pace of financial account liberalization needs to be in synch with the process of domestic balance sheet repayment and bank/shadow bank recapitalization. *$3.5 trillion in reserves at the end of 2015, v $1.5 trillion in external debt, see table 2 of the IMF’s staff report. ** I will have more to say on this in an another post. There is a debate about the liquidity of China’s $3.2 trillion in reserves ($3.3 trillion counting other foreign assets). The key issue here is how China accounts for the foreign assets of China Ex-Im and China Development Bank (CDB). They shouldn’t technically be counted in reserves, but China’s hasn’t been completely clear on this point. The bulk of the evidence though suggests that China Ex-Im and CDB loans show up in the net international investment position in “loans” not as reserves, and that, to the extent these loans have been financed by entrusted reserves, those entrusted reserves are counted as “other foreign assets” in China’s SDDS disclosure of its foreign exchange reserves. In other words, the illiquid loans are not currently counted as part of China’s “foreign exchange reserves,” but rather appear elsewhere. China really should clarify this though. The IMF Article IV unfortunately did not shed new light on this issue. *** Data is from tables 2 and 3 of the IMF’s staff report, pp 40 and 41. http://www.imf.org/external/pubs/ft/scr/2016/cr16270.pdf China holds roughly three times more reserves than any other country. **** The IMF generally requires 5 percent of M2 in its composite metric, or 25 percent of the standard M2 to GDP norm. That rises to 10 percent of M2 for countries with fixed exchange rates, so long as the financial account is open. With M2 to GDP running around 200 percent thanks to China’s high savings rate, and with China’s GDP projected to rise above $15 trillion over the next few years, this soon won’t be an academic debate.