QE created a surge in excess reserves: In September 2008, central banks opened up liquidity facilities to alleviate the stress from frozen fixed income markets. These operations resulted in the build-up of ‘excess reserves' (ER) and an expansion in central banks' balance sheets. In the past, such a build-up in ER would have been ‘sterilised' by central banks by selling government securities. This time around, however, that was not done for two reasons. First, we have argued that central banks have pursued QE with the intention of increasing the growth of money, given near-zero policy rates, while more ER would push overnight rates lower. Second, the sheer size of the increase in ER relative to the size of government securities held by or available to the Fed that could be used to drain reserves was at least partly responsible for central banks not being able to drain excess reserves (see Keister & McAndrews (2009), Haubrich & Lindner (2009)). Under these extenuating circumstances, the Fed even turned to the Treasury for assistance and the Supplemental Financing Program was created to help drain ER (see Appendix for details on how the SFP worked and how its wind-down will push up excess reserves).
But why have excess reserves stayed so high? Could be a supply-side issue... Tautologically, commercial banks are holding on to excess reserves because they cannot put the funds to better use. However, this does not help us in identifying whether supply or demand is the source of the problem. While reserves may be higher than required by regulators, banks may not consider them ‘excessive' in an economic sense, as argued by Friedman and Schwartz (1963). Commercial banks could be hoarding this cash with a ‘precautionary motive', much as consumers increase savings when uncertainty around their income stream increases. This would be a problem on the supply side of the loanable funds market.
Feedback from our money market desk in London provides some interesting insights into the supply side of the argument. First, the interbank market no longer functions as it used to. Lending up to maturities of 3 or 6 months has been replaced by overnight lending or up to a maximum of 30 days, and that too only among select institutions. Absent an interbank market from days gone by, it is little wonder that banks are electing to hold assets in the form of ER. Second, financial institutions still seem to be wary of risks other than liquidity risk. The risk of a possible downturn in the economy and asset prices thereby adversely affecting their balance sheets cannot be ruled out, while there are also risks that regulation could require these institutions to hold more liquid positions in the near future. Until there is some resolution of uncertainty on both of these fronts, it seems that the precautionary motive for holding ER will remain in place.
...but it could also be the lack of demand for bank credit: It could also, however, be that banks simply do not have enough demand for bank credit - a demand-side problem. Most likely, it is a combination of both supply and demand factors. We have argued before that credit lags the recovery as lending standards slowly relax and borrowers slowly re-enter the market as their income and revenue streams become less uncertain, or show outright improvement.
Reserves earn interest...even at the Fed: Another reason why excess reserves have stayed elevated is that commercial banks earn interest on reserves (even at the Fed, which started this practice in October 2008). Particularly in the US where, prior to October 2008, ER would have received no interest, commercial banks would be strongly incentivised to move those funds elsewhere. Now, ER earn the policy rate, which is more attractive than low front-end rates on government yield curves. Further, commercial banks - having already purchased a lot of government bonds - are possibly reluctant to add to these positions, given an impending economic recovery. Part of the regulatory risk that we mention previously is that banks could be asked to hold more risk-less assets on their balance sheets. This process is already in motion in the UK, where banks are likely to hold more liquid assets (see Regulation-Driven Demand, Laurence Mutkin, October 9, 2009).
Researchers at the Federal Reserve argue that the build-up of ER is almost exclusively the result of central bank policy and should not be seen as a reflection of the lending behaviour of commercial banks (see Keister & McAndrews, Why Are Banks Holding So Many Excess Reserves? July 2009). The implication seems to be that a reversal of central bank policy will result in a painless resolution of the ER issue. If so, the argument is not convincing. If the decision to hold assets in the form of ER is the result of a desire to hoard cash by commercial banks, then an attempt by the central bank to reduce ER without a change in the underlying preferences of commercial banks could lead to a decrease in lending. This was costly for the Fed in the 1930s when higher reserve requirements led commercial banks to cut lending.
Will managing excess reserves prove difficult? Central banks can drain/reduce ER in a variety of ways. The old-fashioned tool of requiring banks to hold a higher proportion of their demand deposits as reserves is unlikely to be used. The disastrous policy of raising the Required Reserve Ratio by the Fed in 1936 and 1937 was "an important contributing factor and perhaps even the most important factor" that led to the recession of 1937-38 (Mayer & Cargill (2004)). Rather, our US team expects that draining reserves in the US will likely be executed through a combination of reverse repo operations , term deposits and incentives via the interest rate on excess reserves. Please again see Fedspeak: Roadmap for the Exit for a discussion of the nuances of these operations. At the ECB, control of liquidity and reserves would likely be done through reverse tenders (at the traditional short maturity or for longer ones) or by issuing debt certificates (i.e., central bank bonds). The latter remains an option at the ECB's disposal, though it has not been used yet. The BoE has arguably already taken steps to cut ‘excess' reserves by lower offerings in weekly open market operations. Going forward, the BoE could reduce offerings in its long-term open market operations or even issue short-term BoE bills.
Interest on reserves helps central banks manage reserves... Keister & McAndrews and others have argued that the important step of paying interest on reserves allows the Fed to separate its interest rate policy from its policy objectives regarding the monetary base (reserves plus currency in circulation). In simple terms, if central banks kept raising the interest rate paid on reserves in line with policy rates, commercial banks would be less inclined to move ER elsewhere. In theory, central banks could well carry a large quantity of ER on their balance sheets even as policy rates are raised, without these reserves spilling over into the real economy.
...but the argument is far from watertight: There is clearly significant merit to this argument, but it does not seem to be as comprehensive as its proponents would suggest. For one thing, banks have traditionally earned profits via the maturity mismatch between short-term liabilities and long-term assets with the benefit of an upward-sloping yield curve. Why they should then prefer to keep over US$1 trillion in the form of risk-free assets when (probably) better risk-adjusted returns are available in a recovering economy is not clear to us. Federal Reserve staff themselves believe that commercial banks could re-deploy some of the cash on their balance sheets into securities or loans (as per the FOMC September minutes). Further, non-price objectives (e.g., market share) are likely quite strong in an environment where the Great Recession and events in the housing and credit markets have created very distinct groups of winners and losers among commercial banks. In a nutshell, central banks may have all the tools they need at their disposal, but will need to count on the one factor that is not under their control: the willingness of commercial banks to play along.
Do ER then pose any risk to policy or the macroeconomy? If central banks manage the process of draining ER efficiently or set the right incentives to control the movement of ER, then clearly even a large stock of reserves poses little risk to the macroeconomy. However, like many of the dilemmas facing central banks today, such a balanced exit is far from a foregone conclusion. In an earlier note (see "Growing Pains", The Global Monetary Analyst, September 23, 2009), we highlighted that central banks have looked through the initial stages of the recovery, waiting until the recovery is sustainable and inflation expectations turn up before they hike policy rates. A sustainable recovery is also when borrowers are likely to return to the market, providing commercial banks with alternative lending opportunities for the stockpile of ER. The risk is that commercial banks could seize these opportunities, expanding credit at a time when central banks are trying to reign in their über-expansionary policies somewhat, thereby rendering their efforts less effective.
Second, the spillover of these ER into the real economy via lending will likely lead to an increase in the money supply, pushing inflation risks higher. As Keister & McAndrews themselves note, "while lending by banks does not change the total level of reserves in the banking system, it does affect the composition of that total between required and excess reserves". As lending out of ER increases, deposit creation and money creation will also be stepped up. The increase in demand deposits and money supply means that the level of required reserves that banks have to hold will rise. Thus, required reserves will increase at the expense of ER. As money supply has continued to increase during the recent past, so have required reserves. Lending out of ER will exacerbate both trends. Meltzer (2009) and Feldstein (2009) have also voiced concerns about these risks to inflation from inflated ER.
Of all the policy dilemmas that confront central banks, reining in excess reserves likely worries them the least, given their experience and control in managing their own balance sheet. However, the smooth transition that they likely assume is not a done deal, in our view. We have suggested that the biggest risk going forward is stronger-than-expected growth in the next few quarters (see "‘Up' with ‘Swing'", The Global Monetary Analyst, September 16, 2009). Both the supply of and demand for bank credit in such a scenario could conspire to lay to rest the best laid plans of central banks.
Appendix: The Coming Surge in US Excess Reserves
Our US economics team expects excess reserves on the Fed's balance sheet (B/S) to expand from US$987 billion now to about US$1.2 trillion by early 2010. With passive QE programmes already ratcheted down to a significant extent, the US$350 billion of active QE yet to come along with the wind-down of the Treasury's Supplementary Financing Program (SFP) from its November 2008 peak of US$589 billion to US$15 billion are the reasons for this expected surge. In this box, we try to explain how QE creates excess reserves and how the wind-down of the SFP facility means that excess reserves are likely to increase.
The Fed's passive and active QE policies since September 2008 and early 2009, respectively have resulted in a massive expansion of its balance sheet. In the process of taking on illiquid assets via liquidity programmes (passive QE) or MBS/agency debt/USTs through outright purchase (active QE - see "QE2", The Global Monetary Analyst, March 4, 2009, for more on this distinction), the Fed issued ‘cash' by crediting the reserve account of commercial banks. Since the increase in reserves was over and above that required by regulation, the so-called ‘excess reserves' account of commercial banks increased due to these programmes. The process of creation of excess reserves is best explained by the T-accounts below:
In mid-September 2008, the US Treasury created the SFP "at the request of the Federal Reserve", under which the Treasury can auction Treasury bills that will "provide cash for use in the Federal Reserve initiatives". Under the SFP, the Treasury sold T-Bills at a special auction open to institutions already accessing the Fed's liquidity facilities. Proceeds from the auction were paid out of excess reserves and into the Treasury's account at the Fed. On the Fed's balance sheet, this simply amounts to a transfer of funds between excess reserves and the Treasury's deposit.
On a net basis, however, it is almost like the Fed and the Treasury jointly financed the liquidity programmes started in September 2008. The SFP effectively helps the Fed to drain excess reserves by the amount of the T-Bills auctioned.
The severity of the downturn, however, led to the Fed to pursue a US$1.75 trillion active QE programme starting January 2009, which made the SFP contribution relatively less vital, given the size of this programme. On the fiscal side, issuance needs for the massive US$787 billion fiscal package and the dramatic cyclical fall in tax revenues led the Treasury to wind down the SFP facility to US$15 billion in order to avoid breaching the debt ceiling. In the event that Congress raises the debt ceiling, the Treasury could re-issue T-Bills to assist the Fed in mopping up excess reserves.
Effectively, this would mean that, as T-Bills mature, the funds from redemption are placed back in excess reserves. As of now, there are US$65 billion of SFP bills left, US$50 billion of which will be wound down. The SFP wind-down has thus already contributed to an increase in excess reserves and will continue to do so until the desired US$15 billion size for the programme is achieved. The much bigger impact on reserves going forward will be the US$291 billion of MBS securities and US$58 billion of agency debt that the Fed will purchase as it completes its active QE programme.
Upping the Fiscal Stimulus Further ...
The incoming coalition government between Chancellor Merkel's Christian Democrats and her new junior coalition partner, the Free Democrats, has agreed on additional fiscal stimulus in 2010 and, according to present projections, also in 2011. On top of the €14 billion (equivalent to 0.6% of GDP) already decided on by the previous coalition, the new administration is planning to cut taxes by a further €7 billion in January 2010 (about 0.3% of GDP). These additional tax cuts range from abolishing specific limitations to what constitutes tax-deductible business expenses under the corporate tax code, to a reduction in inheritance tax and in the VAT on hotel room charges, to higher child tax credits and/or child benefits. In addition, the coalition treaty foresees a further tax relief of up to €24 billion under a fundamental overhaul of the personal income tax schedule starting in 2011. Essentially, the idea of this reform of income taxes paid by individuals as well as many small companies is to simplify the income tax to just a few brackets and to reduce to progressive impact of the current tax schedule. The net tax relief aims to benefit low to middle-income earners primarily, as well as families.
... Ahead of a Key Regional Election Next May ...
Unless the federal government announces major spending cuts to rein in the rapidly rising budget deficit after the key regional election in North Rhine-Westphalia next May, the budget deficit will widen noticeably and the debt burden will pile up more quickly. Even though a structural reform of the German tax and social security system is very welcome, further fiscal stimulus in 2011 could be risky, in our view. Not only would it likely be a pro-cyclical policy, further fuelling the recovery underway in Europe's largest economy, but it could also cause consumers and companies to anticipate much tighter fiscal policy in the future, thus making them less inclined to spend the additional take-home pay. This danger is genuine as the newly introduced debt brake, which under the German constitution limits federal deficits to 0.35% of GDP from 2016 and no longer allows regional states to run deficits from 2020 onwards, would require a rather drastic fiscal consolidation between 2012 and 2016 if it is to pass its first reality test.
... Would Be Risky from an Economic Standpoint
In addition, the extra fiscal stimulus could have undesirable repercussions elsewhere in the euro area. Several other governments might be tempted to follow the German example and embark on further fiscal stimulus. At first sight, these preliminary fiscal plans outlined in the coalition treaty don't seem to conform with an agreement on exit strategies reached among European finance ministers only last week, and which is about to be approved by the heads of government at the EU Council meeting later this week. At the last Ecofin meeting, finance and economics ministers had agreed that fiscal consolidation would start in 2011 at the latest and would at least entail a discretionary tightening of 0.5% of GDP per annum. Furthermore, the additional fiscal stimulus would put the German government potentially on a collision course with the ECB, which has argued for a while that fiscal consolidation should start as early as next year. Some ECB Council Members have even argued that a lack of fiscal effort on the part of euro area governments could force the ECB to tighten earlier than it would do otherwise. As the largest economy in the euro area and as a perceived stability anchor, Germany has an important role to play in this respect.
We Still Believe That Fiscal Policy Will Be Tightened in 2011
On balance, we believe that spending cuts will be implemented in 2011 to fund income tax cuts and also to rein in the budget deficit. Recent comments made by the new German Finance Minister, Wolfgang Schäuble, appear to support our view. But these spending cuts will probably only be given more serious consideration after the regional election in Germany's most populous state, North Rhine-Westphalia, which goes to the polls next May. A fresh set of the bi-annual tax estimates that is traditionally due in May could provide a trigger. The new centre-right government cannot afford to lose its current majority in what historically has been a stronghold of the Social Democrats, because this would mean losing the majority in the upper house of parliament, the Bundesrat, which needs to approve many legislative initiatives (see Germany Economics: Mending Europe's Largest Economy, September 27, 2009).
While our base case is that there will be some serious fiscal consolidation effort in 2011, we cannot completely dismiss the possibility of a fundamental shift towards the left either. We had noted such a shift already before the election (see for instance German Elections: Not a Done Deal Yet, September 8, 2009). To an economist this might be a reason for concern. But to a political scientist this might just represent clever tactics in a situation where the current disarray in both left-of-centre parties, the Social Democrats and the Left Party, offers an opportunity to capture more of the hard-fought-over middle ground in politics.
Sifting Through the Small Print of the Corporate Tax Code
One of the key initiatives with respect to corporates is to do away with a number of restrictions to tax-deductible business expenses, which were introduced as part of a corporate tax reform during the last parliament (see Strategy and Economics - Germany: A Compromise on the Corporate Tax Reform, November 7, 2006). Abolishing these specific rules will likely cause tax revenue shortfalls of around €2-2.5 billion next year, which at the margin is likely to boost corporate profits by about 0.5%. The specific rules in the corporate tax code that will be modified or completely abolished include the limits to interest expense deductions, to restructuring cost deductions and some cross-border taxation rules, as well as the limits to carry forward losses after taking over a loss-making company (please see the aforementioned note for details of exactly what these entail). Apart from the rule limiting the tax-deductibility of interest expenses, which will likely bring a welcome relief for a wide range of companies in a situation where access to financing is still an issue, the impact of the other legislative changes will likely be company-specific. An additional relief for family-owned companies that are in generational transition phase comes in the form of the amended rules for companies that are passed on and continue to be operated by the heirs (which enjoy a favourable tax treatment). In addition, legislators want to look into the tax treatment of holding companies at some point during the new parliament.
Limiting Increases in Non-Wage Labour Costs
Another important relief measures benefitting corporates as well as private households is the government's intention to prevent contributions to social security schemes (pension, health, long-term care and unemployment) from rising markedly in the wake of the shortfalls caused by the financial crisis. These contributions are roughly shared in equal proportions between employers and employees and should not exceed 40% of gross wages, according to the coalition treaty. For 2010, an expected shortfall of revenues versus spending totalling about €20 billion (0.8% of GDP) that would normally need to be covered by higher contribution rates will be met by additional subsidies out of the federal budget. Hence, from the corporate perspective there won't be an increase in non-wage labour costs. Looking further ahead, the new government aims to freeze non-wage labour costs, notably contributions to unemployment insurance and statutory healthcare insurance, at current levels. Such a cap on non-wage labour costs is an important step in maintaining Germany's hard-earned cost-competiveness.
In addition to the further subsidies to the social security systems in 2010, the government plans to cap employers' contributions to the statutory healthcare insurance scheme at the current level of 7%. Hence, any future increases in healthcare costs will likely be borne by private households. From 2011 onwards, a fundamental healthcare reform could see their contributions become independent of income. Instead, a fixed fee per capita might be charged (with tax subsidies for those who cannot afford to pay it out of their own pocket). Eventually, however, the government will also have to address the demographic pressures on the pension system that have been reinforced by some of the decisions taken by the previous government.
Bring in Some Flexibility at the Fringe of the Labour Market
Unfortunately, the long-standing demands of the Free Democrats to reduce employment protection have not been successful in the coalition talks. However, some measures will introduce some flexibility at the fringes of the labour market. In particular, fixed-term employment will be liberalised a bit further. In addition, the implementation of any further minimum wages in individual sectors of the economy will have to be approved by Cabinet (rather than just the Labour Minister) to become effective. They would also have to be supported by the majority of trade representatives and employers' associations in the sector. Hence, it will be more difficult to put the straightjacket that those minimum wages constitute with respect to flexibility at the company level onto new sectors. However, the straightjacket won't be removed in any of those sectors that already had minimum wages approved under the previous government. As we have argued before, the minimum wage legislation is largely a measure that reduces the opportunity for new competitors to take on incumbents (who are typically covered by the sector-wide wage contract). In a further attempt to defuse the public's perceived fears that the new government will bring in a new era of labour market liberalisation and deregulation, new legislation will be introduced on so-called immorally low wages, which will prevent wages from falling more than one-third below the sector average.
We continue to believe that the labour market will be key in shaping the recovery as well as determining the size and the duration of the hit to Germany's growth potential. Let's hope that the labour market will weather this downturn better than previous ones and avoid a permanent ratcheting-up in the unemployment rate that characterised so many post-war recessions in Germany.
Summary
South Africa's 2009 Medium-Term Budget Policy Statement (MTBPS) shows that the government expects the country's 2009 fiscal deficit to deteriorate from an initially budgeted 3.8% of GDP in February this year to 7.6% of GDP, and confirmed that the overall public sector borrowing requirement (PSBR) will indeed deteriorate significantly. The National Treasury also introduced a number of measures to further relax foreign exchange controls, and indicated that the Minister of Finance is in agreement with the governor of the South African Reserve Bank (SARB) that monetary policy should be designed to support balanced and sustainable economic growth. In a Q&A session post MTBPS delivery, the Minister indicated that the National Treasury "sees no reason to change the inflation target", but that the global inflation targeting debate is being monitored.
Revenue Estimates Appear Somewhat Conservative
Details of the MTBPS show that the deterioration in the deficit was due to the combination of a sharp revenue undershoot, as well as a modest expenditure over-run. In line with our forecasts, most of the revenue undershoot is expected to come from much lower VAT collections, corporate and personal income taxes. But while the expected shortfall in personal income and other taxes is in line with our estimates, the anticipated undershoots in VAT (R31 billion) and corporate taxes (R21 billion) are, however, larger than our forecasts of R24 billion and R10 billion shortfalls, respectively. We believe that the difference is largely driven by what appears to be a relatively bearish outlook on corporate profitability by the National Treasury. Thanks largely to the conservative growth outlook priced in by the Treasury, its overall tax revenue shortfall of R70.5 billion is some R12 billion larger than our estimate of R58.5 billion.
The Treasury left its 2009 forecast of fiscal transfers to the Southern African Customs Union (SACU) unchanged - although it has made downward adjustments to its 2010-12 forecasts. While we agree that these SACU transfers are generally backward-looking, we believe that the Treasury is being overly cautious here. To be clear, our analysis suggests that these transfers should fall by some R3 billion this year, allowing the government to generate additional savings of a similar magnitude.
But Expenditures Will Likely Overshoot
Apart from the ‘goods and services' budget category, where expenditure is now budgeted to fall sharply in the coming years, there were no major surprises in the expenditure budget. A number of sub-categories were shifted around; however, the overall impact was broadly unchanged, relative to our forecast. For example, while allocations to public corporations were raised from R53 billion to R57 billion (we had expected a contraction to R48 billion), the increased allocation here was funded from internal resources (a draw-down on the contingent reserve line), leaving the net impact on the overall expenditure budget broadly unchanged.
Transfers to households came in above their February estimates too, as applications for old age and child support grants rose during the economic downturn. According to the Treasury, the social grants program drew in an additional 600,000 people in 1H09, mainly new recipients of the child support (475,000) and older persons' grants (86,000).
The real surprise was in the ‘goods and services' budget (usually government consumables and other expenditure items that help lubricate the civil service), where the Treasury cut its 2009/10 estimates by some R10 billion, without indicating how it aims to achieve this. To be fair, the MTBPS does indicate that the government intends to reduce spending on non-core functions, shift resources from administrative to frontline service provision, weed out poorly performing programs, reduce wastage, prevent extravagant spending, reform the procurement systems, etc. However, this appears to be more of a wish-list of medium-to-long-term objectives that are unlikely to be achieved in the current fiscal year, in our view. For the record, we do not believe that 2009/10 expenditures will be capped at R841 billion. Instead, we stick to our forecast of R849 billion. Elsewhere, the Treasury expects other payments - mainly interest on higher levels of government debt - to rise ahead of our estimates over the Medium-Term Expenditure Framework, reaching R108 billion by 2012/13.
What's more, the MTBPS makes additional resources of R78 billion available over the next three years, half of which is expected to be paid out to provincial government for the funding of higher personnel costs in the main, as well as education, health and housing expenses. However, it is interesting to note that this additional resource rollout is significantly back-loaded, and will only affect the overall spending budget in 2012 (R31 billion) and 2013 (41 billion). In the near term (i.e., 2009-11), an increased allocation to wages is expected to be funded by a shrinking ‘goods and services' budget. As discussed earlier, we believe that the anticipated savings the government hopes to make to its ‘goods and services' budget are overly ambitious and unlikely to be attained.
We also noticed that the Treasury has cut back on its 2009-11 infrastructure allocations to public-private partnerships, but has raised allocations to state-owned enterprises (SOEs). We suspect that the increased outlays on public enterprises are an indication that the government may be willing to allocate yet more resources to Eskom in particular, were its request for massive tariff increases not to be approved by the Regulator. On the whole, the infrastructure budget for non-financial public enterprises has been raised by some R30 billion over the 2009-11 period. We did not expect increased allocations here.
Double-Digit Public Sector Borrowing Requirement
With a revised revenue estimate of R657.5 billion and an expenditure target of R841.4 billion, the 2009/10 main budget shortfall of R181.6 billion (7.6% of GDP) turns out to be some R8 billion higher than our estimate of R174 billion (7.3% of GDP). The higher-than-expected fiscal gap results in an overall PSBR of R285 billion (11.8% of GDP) - also higher than our estimate of R275 billion (11.5% of GDP). This divergence is almost entirely explained by the Treasury's conservative corporate income tax estimate. We find it surprising that the Treasury expects the PSBR to remain high in 2010-12. This may again have to do with conservative accounting and the Treasury's relatively bearish growth (and, as a corollary, tax revenue) outlook.
Yield Curve to Steepen as Issuance Soars
With respect to funding, the Treasury appears to have taken advantage of the strong commercial bank appetite for Treasury bills and hopes to triple its 2009 Treasury bill issuance from R15.6 billion in the Feb-09 Budget to some R50 billion. This is broadly in line with our estimate of R55 billion. Technically, higher funding in short-dated instruments is often synonymous with rising rollover risk. It is important to note that more than half (i.e., R31 billion) of the revised T-bill budget has already been issued in the first five months of the fiscal year, with less than R20 billion more to go.
Read Full Article »