This week, the ECB announced that it will conduct specific operations to absorb the liquidity injected by the SMP. In this regard, the ECB carried out a quick tender on May 18 to collect one-week fixed-term deposits. This liquidity-absorbing operation was conducted as a variable rate tender with a maximum bid rate of 1%. The ECB announced that it intended to absorb €16.5 billion, an amount corresponding to the size of the SMP taking into account transactions with settlement at or before Friday, May 14. These new fixed-term deposits are also eligible as collateral for the Eurosystem's credit operations. As the ECB already announced that it plans to carry out another liquidity-absorbing operation next week, we would expect that weekly term deposits are likely to become a regular fixture on the ECB's OMO (Open Market Operations) calendar. Going forward, this weekly announcement will also give the market a good idea how big the purchases during the previous week were.
Additional insights can be gleaned from the consolidated balance sheet of the Eurosystem, comprising the ECB and all 16 NCBs. This balance sheet data come out every Tuesday, at 2 pm London time. Here we are looking in particular at the following balance sheet items and how they have changed since last week:
• On the asset side, item 7.1 (securities of the euro area residents denominated in euro - securities held for monetary policy purposes, which so far have consisted of the covered bond purchases).
• On the liability side, item 2.3 (liabilities to euro area credit institutions related to monetary policy operations denominated in euros - fixed-term deposits). We expect both to rise by a similar amount.
Hence, despite the sterilisation of the SMP, the size of the ECB balance sheet is likely to increase. However, the excess reserves of the banking system - which are sitting in the ECB's deposit facility - should not be affected by the SMP once the time deposits are in place. Importantly, the monetary base - defined as currency in circulation plus the minimum reserves that banks are required to hold and any excess reserves they want to hold voluntarily in the deposit facility - should not change as a result of the SMP. This neutrality of the programme with regard to its monetary policy stance was stressed by the ECB from the start. Nonetheless, the SMP triggered a lively debate on whether it would mark the start of QE by the ECB and whether it could be sterilised effectively. Many were quick to point out that the Federal Reserve had also initially planned to sterilise its purchases but eventually had to resort to outright QE, given the size of its purchase programme.
While there is no uniformly agreed definition of QE, the ECB has clearly pursued QE since autumn 2008. An expansion of excess reserves permitted or desired by the central bank is the clearest indication that a QE regime is in place. Dick Berner and David Greenlaw from our US economics team are looking for a continuation of the ongoing sterilization of excess reserves in the US. As this process progresses, there is likely to be less pressure on overnight rates and a lower probability of excess reserves spilling out into the real economy. Joachim Fels and Manoj Pradhan from our global economics team suggest that an enlarged balance sheet provides useful information about the QE regime when excess reserves are being sterilised. This is because sterilisation alters the composition of the liability side of the central bank's balance sheet (and thus the composition of the asset side of the balance sheet of the commercial banking system), but not the overall size of the balance sheet. The total amount of liquid assets (cash in excess reserves, or reverse repos and term deposits that represent an asset for commercial banks) that are present in the banking system because of QE remains unaltered.
It is clear from the detailed comparative work done by Fels/Pradhan on QE strategies in the major industrial countries (see "QE2", Global Monetary Analyst, March 9, 2009) that the ECB started its QE programme in autumn 2008, when it switched its refi tenders to fixed-rate full allotment, in order to provide euro area banks with unlimited liquidity. As a result, both the ESCB's balance sheet and the excess reserves surged. As a consequence, the ECB balance sheet surged from €1.5 trillion to €1.9 trillion and the excess reserves from essentially zero to around €275 billion.
Contrary to the Federal Reserve, the ECB expanded its balance sheet through the liability side (i.e., through its lending). Fels/Pradhan coined this strategy ‘passive QE' to underline the difference with the ‘active QE' pursued by the Federal Reserve and others. It is ‘passive' in several ways. First, it operates via the passive side of the balance sheet. Second, the amount of QE is determined by the bids banks submit at refi operations, which the ECB ‘passively' fills, all under the fixed rate full allotment tenders. The key difference between passive and active QE is:
• that the ‘passive' QE works via the banking system rather than aiming to circumvent it;
• that banks remain owners of the assets pledged to the central bank as collateral and thus benefit on the upside and the downside; and
• that the additional liquidity provided by the ECB needs to be rolled whenever the refi operations expire, making an exit easier to execute.
In addition, the list of collateral that can be funded at the ECB is much broader than the assets bought by other central banks. Finally, the banks decide which assets to fund at the ECB, not the ECB which ones to buy. To sum up, the ECB has long engaged in successful QE. As such, the SMP does not mark the start of QE in the euro area.
Why sterilise when the ECB's strategy has been one of ‘passive' QE anyway? To be clear, as long as the ECB conducts the bulk of its refinancing operations at fixed-rate tenders with full allotment, its does not control the size of its balance sheet. Nor does it control the size of the excess reserves the banking system holds in the deposit facility. Both crucially depend on the bidding behaviour of euro area banks at the various refi operations, be it the weekly MROs, the monthly SLTRO or the three- and six-month LTROs. In our view, the key difference is the timeline. The purchases under the SMP will likely be held to maturity. So not sterilising these purchases would amount to an almost permanent injection of liquidity for several years. The refi operations all have a sell-buy date of one year at most. The liquidity injection automatically reverses. We will likely see this clearly when the first one-year LTRO rolls off on July 1. This first one-year LTRO saw €442 billion being taken down by the banking system last summer. As a result, excess reserves jumped and the EONIA overnight rate fell even further below the refi rate. While the ECB will provide a bridge for banks to roll this funding into the next MRO, etc., it remains to be seen how much collateral the banks still want to fund at the ECB in the current environment.
In addition, the ECB likely wants to underscore that its SMP is a temporary measure only. This argument was made by a number of Council Members already last week and then included in the formal decision on the SMP published by the ECB on May 14. The temporary (and hence limited) nature of the programme would also make sense, given the dissenting views on the Governing Council, notably by Bundesbank President Axel Weber and reportedly (Der Spiegel, May 17) also by Nout Wellink, the Governor of the Dutch Central Bank, and by Juergen Stark, a member of the ECB Executive Board. It is key to the ECB, we think, to maintain its credibility as an independent inflation fighter. The ECB thus remains very focused on anchoring inflation expectations.
Council Members including ECB President JC Trichet emphasised the delineation of responsibilities between fiscal policy and monetary policy. In the view of the ECB, the sovereign debt crisis is primarily a fiscal policy concern that needs to be addressed by aggressive austerity measures and, if needed, intra-government emergency lending facilities for distressed borrowers set up under the EU Stabilisation Fund. But the ECB is keen to underscore that the SMP is not a monetary policy measure, but targeted market intervention to relieve stress points in individual market segments, similar to the covered bond buying programme launched last summer. The notion is not to add overall liquidity to the system.
Signposts for ECB watchers in the weeks ahead. While the Governing Council meeting this coming Thursday is not a monetary policy meeting, we could nevertheless see an announcement afterwards (not necessarily at the usual time of 12.45pm though, but perhaps later than that) providing further details on the buying programme and its sterilisation over and above what has already been mentioned. In addition, we will mark our calendars for the announcements and execution of the weekly term deposits and the release of the ESCB balance sheet. The first term deposit was heavily over-offered, with the offers amounting to ten times the amount the ECB wanted to drain (see Interest Rate Strategist - ECB Tender Tracker, May 18, 2010). In this draining operation, 223 participants offered €163 billion of funds compared to an allotment fixed at €16.5 billion. The weighted average rate the ECB paid for these funds came in at 0.28%, which is equivalent to EONIA - 6bp, another reaffirmation of the large amount of excess liquidity in the system as the ECB's QE strategy is still in full swing. Note that the amount of excess reserves in the system could potentially step-change when the first one-year tender of €442 billion comes out of the system at the beginning of July. In this case, EONIA could leap higher.
Central banks around the world have different policy objectives - some have dual targets related to both economic growth and inflation (such as the Fed) and others have a more limited focus (such as the ECB, which operates under a price target). All central banks use a variety of different tools to pursue their objectives. Most - including the Fed and the ECB - have been operating under an interest rate target approach for some time. The justification for adopting an interest rate target can get somewhat complex, involving issues such as the relative stability of the money demand function and the definition of money itself, but the basic idea is quite simple: a lower cost of money stimulates the economy more than a higher cost of money and vice versa.
When a central bank runs out of room to cut interest rates or when an interest rate approach to monetary policy is perceived to be lacking in some way, the central bank can attempt to achieve its policy objective by manipulating its balance sheet. Typically, an expansion of the balance sheet will lead to the creation of bank reserves and will increase the monetary base. Such balance sheet expansion and creation of excess reserves is commonly referred to as ‘quantitative easing' (or QE) because it involves a change in a quantity variable (reserves and/or the monetary base) as opposed to a change in the interest rate target. A central bank could also merely shift the mix of its balance sheet in order to achieve a policy objective - say by adding one type of asset and selling another (or by adding a liability). The action involved in offsetting the acquisition of an asset by a central bank - either by liquidating a different asset or by adding an equal amount of liabilities - is called sterilization. It should be obvious that when the acquisition of an asset is sterilized, there is no QE because the balance sheet effects are neutralized.
From the perspective of a monetarist, QE is important because it represents growth in the monetary base and - assuming some degree of stability in the money multiplier - an accompanying rise in money supply. This should help to stimulate economic activity, while also carrying some inflation risk. However, from the standpoint of most mainstream macroeconomists, the degree of stimulus and the inflationary consequences of QE are less clear-cut. This reflects the fact that (a) the money multiplier may not be stable, and (b) the determination of inflation is more closely tied to the amount of slack in the economy and inflation expectations than to gyrations in the money supply (which are unobservable in any case). The US experience provides a good case study of the debate over the significance of QE.
The US Experience
The Federal Reserve shifted to quantitative easing in September 2008 when it expanded a number of liquidity programs, including the term auction facility (TAF) and central bank FX swap lines, and ceased its sterilization efforts.
Up to this point, the Fed had been sterilizing the impact of its new support facilities by liquidating Treasuries. For example, the TAF was introduced in late 2007 and was scaled up to $150 billion by May 2008. Over that same interval, the Fed liquidated more than $200 billion of its holdings of Treasuries in order to sterilize the TAF and other special programs.
So, the volume of bank reserves was essentially unchanged during this period, but the mix of balance sheet items shifted. In September 2008 - as financial markets were correcting severely - the Fed gave up trying to sterilize. Excess reveres rose from a normal level of $1.0-1.5 billion to $270 billion in October 2008, as the liquidity support programs continued to expand.
By the end of 2008, excess reserves reached $800 billion and the monetary base had nearly doubled in size. In early 2009, the Fed started to purchase large quantities of MBS and agency debt, and in March it began buying Treasuries. However, it's important to note that the bulk of the QE took place several months before the Fed started buying mortgages and Treasuries. Thus, it is incorrect to simply refer to the Fed's bond purchases as QE. These purchases certainly helped to sustain quantitative easing (because other programs were simultaneously winding down), but the QE was essentially complete before the first MBS was purchased.
In addition, Fed officials are resistant to using the term QE to describe their actions in recent years. They prefer to call the approach ‘credit easing'. This seems somewhat justified because there was no explicit (or even implicit) target for bank reserves or the monetary base. Instead, the Fed's objectives were to restore liquidity to important markets - such as the interbank funding market, commercial paper market, etc. - and to push mortgage rates to artificially low levels in order to support consumer cash flows and boost housing affordability. The Fed succeeded in achieving these goals. The TAF, FX swap lines, and alphabet soup of other liquidity support facilities appeared to play an important role in reining in Libor. Meanwhile, the LSAPs (large-scale asset purchases) helped to drive mortgage rates lower, which provided a significant amount of stimulus to the economy.
Still, despite the Fed's protests, it seems reasonable to refer to central bank actions that expand reserves and the monetary base (far beyond the historical norm) as QE. Our own Japan chief economist Robert Feldman (who obviously has some experience with QE regimes) offers this: "monetary base growth in excess of 30% on a year-on-year basis is a good definition for entry into QE". This criteria was easily met in the US as the monetary base more than doubled in size.
As mentioned earlier, the monetarist view is that QE represents an important event because the expansion of the monetary base is likely to be accompanied by growth in the money supply. However, this was not really the case in the US. While the base doubled, growth in narrow money experienced only a modest acceleration, as the money multiplier plummeted. This reflected the fact that the excess reserves created by the Fed were parked in cash. Thus, the monetarist assessment of the US QE experience would be that it didn't work. Of course, the Fed's view - and the mainstream view - is different. The form of QE matters, and it's not just about pushing up the monetary base and the money supply. Indeed, as noted earlier, the objectives of the Fed's version of QE were largely achieved. Moreover, if the Fed can now engineer a successful exit from QE, any inflation consequences and market distortions should largely evaporate.
Can the US version of QE be successfully unwound? Fed officials have publicly outlined the mechanisms that will be utilized (in particular, see "Federal Reserve's Exit Strategy", Ben Bernanke's testimony to the House Financial Services Committee on February 10, 2010). First, exit will occur as some of the assets that were acquired prepay and/or mature. Second, the liquidity support facilities - which led us into QE - have already essentially been wound down. This is one reason why the recent agreement to resuscitate the FX swap lines was not necessarily a slam-dunk for the Fed. It is actually a small step backward from the standpoint of restoring normalcy to the balance sheet - especially since it appears that the Fed will not sterilize the balance sheet impact. But it is only a small step because the ECB's dollar lending program is structured in a way (i.e., wide spreads, high haircuts) that should limit the volume of euros acquired by the Fed.
Finally, the Fed plans to use term deposits, reverse RPs and asset sales to unwind QE. The asset sale option has generated a lot of interest recently and appears to have gained unanimous acceptance among Fed officials. But the sequencing still appears to be reverse RPs and term deposits first, followed by asset sales later on. From our standpoint, the speed at which QE is unwound depends, in large part, on the effectiveness of the interest on reserves (IOR) program. We have serious concerns regarding the effectiveness of IOR and thus believe that exit from QE in the US could be very messy.
Some background may be helpful. The IOR program was introduced in autumn 2008 and was aimed at allowing the Fed to put a floor under the fed funds rate despite the massive excess reserve position. The logic behind the IOR program is simple: why would an institution sell fed funds at a rate below the rate they could earn by parking these funds with the Federal Reserve? However, the IOR program did not prevent fed funds from trading well below target in late 2008. The problem appears to be related to the actions of the GSEs - particularly, the Home Loan Banks - who are major net lenders in the fed funds market. The GSEs are not depository institutions and thus are ineligible for interest on reserves. So, their actions can put downward pressure on the fed funds rate when the excess reserve position is high.
The Fed argues that the banking sector should arbitrage away any fed funds market distortions related to the GSEs, but couldn't do so during the most intense portion of the financial crisis due to balance sheet constraints. We are skeptical that banks will engage in such arbitrage behavior (even in a more normal balance sheet environment) and thus are concerned that the Fed may not be able to hike the fed funds rate when the time comes, unless it is willing to drain away a sizeable portion of the excess reserve position.
As Chairman Bernanke indicated in his February Congressional testimony: "As the time for the removal of policy accommodation draws near, those operations [reverse RPs and term deposits] could be scaled up to drain more significant volumes of reserve balances to provide tighter control over short-term interest rates. The actual firming of policy would then be implemented through an increase in the interest rate paid on reserves. If economic and financial developments were to require a more rapid exit from the current highly accommodative policy, however, the Federal Reserve could increase the interest rate paid on reserves at about the same time it commences significant draining operations."
So, Bernanke is saying that the Fed will try to engineer a gradual exit from QE, but could be forced into a more rapid exit. It should be obvious that the process of draining $1 trillion or more of excess reserves in a short period of time is fraught with potential market risks. As a comparison, throughout its history, the Fed has from time to time drained reserves from the system via reverse RPs. But with the exception of a brief period in 2008 when a special $25 billion operation was in place, the largest reverse RPs ever conducted were less than $10 billion. By comparison, these days we are talking about an eventual drain need in excess of $1 trillion.
Of course, this is merely the flip side of entering into QE. The Fed (quite appropriately, in our own view) threw everything but the kitchen sink at the financial crisis. The clean-up could entail significant volatility in front-end rates, given the enormity of the task. We view such volatility and short-run market distortion as the likely costs associated with exit from the US version of QE.
Question 1: How large is the Chinese economy's exposure to the Euroland economy in general and the peripheral European countries that are experiencing fiscal pressures in particular?
A: The negative impact on the real economy due to China's trade exposure to Euroland will likely be quite manageable, especially in the short run.
In 2009, the Euroland economy constituted the market of destination for about 20% of China's exports in total, of which only about 3% is accounted for by the four European countries - Portugal, Italy, Greece and Spain - that are experiencing fiscal pressures.
Moreover, the impact of the sovereign debt crisis on real economic activity in Euroland in general and demand for goods and services imported from other countries in particular will likely spread over time, as the contractionary impact of fiscal consolidation efforts in peripheral European countries will likely take time to show. Our colleague, Elga Bartsch, Morgan Stanley Chief European Economist, cut her forecast of EMU GDP growth for 2010 to 0.9% from 1.2% and her forecast for growth in 2011 is 1.1% (see European Economics: Bumping Along Below Par, March 11, 2010).
In the near term, the effect of a cyclical upswing due to a global recovery will likely dominate the growth-dampening effect of the ongoing sovereign debt crisis, in our view. We expect China's export growth to continue to be quite robust over the coming months.
Question 2: What is the risk of as sharp a decline in China's exports as was the case in the wake of Lehman's bankruptcy in September 2008?
A: The probability of a sharp contraction in global trade flows in general and China's exports in particular is very low, in our view.
The collapse of global trade in the immediate aftermath of Lehman's bankruptcy reflected a sharp contraction in underlying real demand as much as a financial market phenomenon: the drying-up of trade financing when major financial institutions carried out significant deleveraging.
The risk of disorderly deleveraging by banks in Europe seems quite low at this juncture. First, liquidity risk in the eurozone appears to have been largely contained after the €750 billion stabilization fund was set up by the Euroland authorities. Second, the structure of debtors and creditors involved in the sovereign debt crisis is rather simple and transparent, and this should help to substantially reduce the counterparty risk that contributed to the disorderly deleveraging by sophisticated financial institutions in late 2008 and early 2009.
Question 3: What is the impact of a substantially weakened euro on the Chinese renminbi exchange rate policy?
A: Barring persistent and high EUR/USD volatility, the renminbi is still on track to exit from its current de facto peg against the US dollar, in our view. We continue to believe that June/July represents the most likely ‘window of opportunity' for such a move.
Under the de facto peg, euro weakness translates into renminbi appreciation on a trade-weighted basis. Many market observers therefore start to attach a low probability to a renminbi de-peg from the USD in the near term, given that the renminbi has appreciated passively on a trade-weighted basis.
The renminbi has appreciated on a trade-weighted basis since the beginning of the year. However, despite the appreciation, at its current level (as of May 19), the renminbi NEER has nearly reached the same level as about year ago.
Moreover, the currencies of many of China's EM peers have appreciated much more than the renminbi has on a trade-weighted basis. This is because almost all major EM currencies appreciated significantly against the euro. And China is by no means an exception.
These developments suggest that to the extent that China is competing with its EM peers in third markets (e.g., Euroland), China has not lost competitiveness as a result of its passive appreciation on a trade-weighted basis. In this context, expectation of a delay in renminbi exit from a USD peg and revaluation due to euro weakness is not well founded, in our view.
We are actually of the view that maintaining a competitive renminbi exchange rate on a trade-weighted basis is not the most important consideration in the Chinese authorities' decision to peg or de-peg the renminbi against the USD. Instead, we believe that injecting a sense of stability and confidence by maintaining a ‘nominal anchor' amid a financial crisis is the key factor. In absence of high market volatility, the benefit of maintaining a more flexible exchange rate regime likely outweighs the need to maintain a nominal anchor.
Looking ahead, we would like to reiterate that renminbi revaluation remains a policy option despite recent euro weakness. While we do not expect an imminent renminbi move in the run-up to the US-China Strategic Dialogue that is to take place on May 24, we stick to our view that June/July is the most likely ‘window of opportunity' for such a move (see China Economics: Renminbi Exit from USD Peg: Whether, Why, When, How, April 4, 2010). In particular, a move in the run-up to the G20 summit that is due to take place on June 26-27 in Toronto would make a lot of sense to us. And from a tactical point of view, it makes sense to us to make the move when market expectation of a renminbi revaluation is low as it is now.
Read Full Article »