Cliff # 1

About cliffeconomics

This blog offers original economic thought and policy recommendations on Germany, the euro area, and whatever cliff has on his mind.

Cliff # 3

About cliff

The author is an economist specialized in financial and macroeconomic policy analysis. All posts present a personal opinion, and all analysis is based on publicly available information.

Cliff # 1

About cliff

The author is an economist specialized in financial and macroeconomic policy analysis. All posts present a personal opinion, and all analysis is based on publicly available information.

Friday, September 30, 2011


Privatization is often thought of as below-the-line savior to the debt problems in the euro area. But a crisis may not the right time to privatize, other than to raise liquidity that cannot be obtained else.

From a flow perspective, it would always make sense to hold on to assets that have a higher return than the liabilities, rather than disposing of the asset to pay down debt. From a stock perspective, both assets and liabilities suffer in times of crisis, so depending on the relative drop in value privatization may or may not become more attractive. Usually, prices of equity reacts more strongly than debt to a downturn, but in a situation in which debt is so risky that it resembles equity, the logic may not hold.

For Spain, it makes financially sense to hold off with the Loteria privatization from a flow perspective, and (given ECB’s interventions) also from a stock perspective:
  • From a flow perspective, Loteria’s profitability massively exceeds the cost of debt service. Loteria’s return on equity is about 17 percent. Compare this to average yields of Spanish government bonds of 4-5 percent!
  • From a stock perspective, Loteria’s expected sale value has fallen by 20 percent (from EUR21 billion to EUR17 billion) since December 2010. Assuming that at in December 2010 a privatization-debt-buyback deal would have been optimal (which is doubtful anyway), the sale would still make sense if bonds would have also dropped by 20 percent. This is not the case (the 10 year benchmark dropped by about 6 percent), and would only become the case if 10-year yields would rise from currently 5.2 to more than 7.5 percent.

Thus, governments should be cautious when wishing for higher privatization to fill below-the-line financing gaps. Notwithstanding legal obstacles, a better solution to utilize government assets would be to securitize and bring them onto the balance sheet so they can be used as financial collateral to acquire funding.

Sunday, September 25, 2011

Overlooked factors in the discussion of Eurobonds

Often the argument is brought forward that Eurobonds will be expensive for Germany. The bonds would have higher yields than German Bunds given that  weaker debtors are aggregated into one issuer. Here are two counterarguments. (A third one, higher liquidity, has been made in abundance already.)

First, credit fundamentals do not suggest that the euroarea as a whole is less solvent than Germany (see table). Debt levels, cyclically adjusted primary balances, and long-term contingent liabilities from pension and health are roughly on level with Germany. If pre-crisis convergence growth would be restored, the euroarea as a whole would grow faster than Germany, although current potential growth estimates suggest that growth paths are roughly at level. And with the US Treasury’s AA+-rated and Japans AA--rated bonds yielding on par or less than Germany’s AAA Bunds, a non-prime rating should no longer sound the death knell for eurobonds.

General government debt (percent of GDP, 2010)
Cyclically Adjusted Primary Balance (percent of GDP, 2011)
Growth in public health and pension expenditure (ppt of GDP, 2010-30)
Historic GDP growth (real, 2005-10)
Potential GDP growth (real, 2011-16)

Sources: Eurostat; IMF Fiscal Monitor and WEO.

Second, the counterfactual of yield compression in the eurozone 1995-99 may be overlooked. If yield compression for peripheral issuers rests in part on an implicit guarantee from eurozone membership, so it is likely to be borne by Germany. (I know that substituting whacky monetary policy for Bundesbank policy is another factor.) But partially, if peripheral bond yields declined due to the implicit solvency backstop, German yields must have increased. Isn’t it surprising that Bunds yielded lower than gilts until the onset of the ERM II fixing (see figure)? Also, non-resident holdings, while increasing for all countries, accelerated more for other euroarea countries than Germany (see figure). This supports anecdotal evidence that investment flows after the euro introduction were directed away from Germany towards other euroarea countries, possibly lowering their relative yields. With eurobonds, this demand would be unified into one instrument.

Thus, if the eurozone is saved, and countries brought back from crisis, I don’t see a strong argument why eurobonds would yield those 100-200 basis points higher that are cited by others (e.g., ifo).
Source: Bloomberg

Sources: IMF Investor base dataset; CPIS. 
1/ Data include France, Italy, Spain, Greece, Portugal, Ireland. 
2/ Data include all euroarea countries (without Germany) but excludes 
the financial centers of Luxembourg and Ireland

Tuesday, September 20, 2011

Economics 1-ohhh-1: Ways to resolve a debt overhang

Without implication, imagine there is a small country within a currency union. This country, let's call it Olive, suffers from too high debt and a liquidity crisis. Imagine there is another, large and solvent country that is called Oak. And on top of that, assume politicians really want to end the crisis. This sounds like a distant imagination, doesn't it?

What can be done? Debt crises can be overcome in two ways: (i) growth and (ii) transfers, whereby transfers can take a zillion different shapes with different distributional effects, in particular: (a) inflation, (b) grants by Oak to Olive, (c) default by Olive.

Growth is a first-best solution. Instead of producing traditional olives, a coincidental invention allows olive trees to grow gourmet olives with incredibly better taste. The resulting jump in value creation allows Olive to raise more taxes and pay off its debts. Everybody would be better off and there are no obvious distributional consequences. (Although those with savings for their retirement would also prefer to buy gourmet olives from their savings, but have only saved enough to afford traditional olives. In terms of the economics of happiness, tough luck!)

This is where we run out of Wunderland solutions. Gourmet olives remain a dream. Realizing this, politicians go on to examine transfers. Transfers can be seen as tax collected from some and handed out to others.

First, there is the inflation tax. Assume that the central bank has successfully been captured by politicians (another tough fact of life) and helps out by getting inflation going. (Of course, this is not as easy as it sounds.) Prices rise, and so does the current value of (olive) output, while the current value of promised future payments declines. This is nothing else than a tax on future payments, whereby the tax is higher the further the payment lies in the future. Creditors with long-term (fixed rate) savings lose most, while debtors with long-term (fixed rate) debts gain most. Usually, government debt and pension savings have the longest duration. Hence, inflation redistributes wealth from creditors to debtors, and mostly so from pension funds (owned by the working population) to governments. This subtle way of wealth distribution takes place across the entire currency union, not only in Olive! Besides the question of stocks (of savings and debt), there is also a question of flows (of income and consumption): prices rise before incomes do, hitting in particular low wage earners. Finally, as is the case for all taxes, inflation has distortionary effects. In particular, inflation leads to higher interest rates, elevating the cost of investments, and thus lowering productivity growth across the monetary union. But politicians still love inflation: it is a very subtle tax, and the electorate may fail to see the true cost of it--unless the central bank completely loses its credibility and an inflation spiral gets out of control. So, political cost for politicians may be low. But costs from economic distortion in saving and investment as well as credibility costs of the central bank are substantial! Hm, doesn't sound too good, does it?

Second, there are unconditional transfers, or grants. The distributional effects are explicit: taxpayers in Olive win, taxpayers in Oak lose. Full stop. The crisis should be over, and Oak may actually suffer less than thought: To the degree that Olive's creditors are actually residents of Oak, the inter-country transfer turns out to be an intra-Oak-transfer. In other words, Oak redistributes wealth from all taxpayers to some creditors. Thus, to some degree the transfer is self-serving and may be cheaper to Oak than the distortionary inflation tax or the mess that may come out of a default (see below). The problem is not the feared reach into poor-Oak-taxpayers' pockets (figurated in large letters all over tabloid newspapers), but a conceptual one: critics claim that transfers induce "moral hazard": A transfer saves Olive's politicians from the pain of cleaning out their olive orchard, and the problem is more likely to repeat itself. This critique is not entirely true. Any crisis hurts Olive's government no matter what. Therefore, they have the incentive to bring their orchard in order. But they probably won't turn it into an Oak grove. Depending how well the ex-ante disciplining mechanisms are that force each member of the currency union to keep its orchards, groves, or whatever in best order, transfers are the least disruptive solution. This is the idea behind the fiscal union.

Third, there is default. Olive would just walk away from its debts, keeping all olives for herself, in fact imposing a tax on all creditors. If it is known who the creditors are, it is easy to identify the direct distributional effect of this wealth transfer: the creditors are often banks and pension funds. Ergo, wealth is being transfered from depositors and pension savers (who partly may not be residents of Olive, but rather Oak) to the government. The solution does not sound so different from transfer by inflation except that the transfer is much more direct and has less repercussions across the entire monetary union... hang on! Really? Conventional wisdom is that default is extremely costly because it sends a shock wave through the entire domestic economy. Contagion occurs because the government defaults on banks and pension funds, these default on depositors and pension claimholders, these in turn default on their mortgages, and so on. Why does the same not happen with the inflation tax? Because inflation shrinks everybody's assets and liabilities, distributing the burden very evenly. With default, eventually the burden will also be distributed through the system, but by means of small explosions that go off here and there. Key for avoiding this chain reaction is to identify the first line of vulnerable entities and protect them. For instance, Oak gives Olive a bridge loan to "buy time" and allow the banks to build warchests (e.g., capital buffers for banks). Or, Olive builds a central reserve to help the victims (e.g., a bank recapitalization fund financed by, er, well, probably Oak). This sounds difficult, and it surely is. The politics of it are messy, and this is about where we are right now. Ex ante it is not clear what the distributional consequences and deadweight losses are. Default taxes tend to shake up the political elite, and this may actually be the nasty truth of the political economy because backbench politicians often gain from shake-ups. (Just read "Freakonomics" to learn why small drug dealers love to kick off gang wars.) Unless well planned, the ultimate distortionary effect of a default can be substantial: incomplete information about the contagion channels ("who is an emperor without clothes?") can lead to a general loss in confidence (not only in Olive, but across the entire currency union). This will disrupt financial intermediation or, in other words, discourage saving and investment, and lower productivity growth. Avoiding this requires strong political will and as much excellent technical planning as execution. In an idealized world, the distributional consequences could be very limited and direct. The inconvenient truth is that in most cases it turns messy at some point.

Assuming that there are no gourmet olives, there will be a tax on Oak: Grants like default like inflation are all taxes. Whatever the tabloids try to convey to Oak's taxpayers, there will be some costs to preserve the currency union. And just because the costs of a grant from Oak to Olive is so obvious, there is no good reason to believe that the hidden costs from default or inflation are any lower. Grants are not a bad choice if the disciplining devices avoid moral hazard. Default is not a bad choice if the defense shield against contagion works. And inflation... well, unlikely to solve the problem anytime soon.

Tuesday, September 13, 2011

Number of the day: Greece deficit

Projected fiscal deficit of Greece in percent of Germany's GDP: 0.8. Second bailout package in percent of Germany's GDP: 4.6. Memorandum: Marshall Plan in percent of US' GDP: 5.0.

ECB's Securities Market Program is fiscal action in disguise

The ECB's Securities Market Program (SMP) is, as shown by last weeks resignation of J├╝rgen Stark, controversial. One striking argument in favor of intervening in sovereign markets is the analogy to a lender of last resort (LOLR), a liquidity backstop central banks provide to banks (see Paul Grauwe's VOX contribution). Yet, I think a stringend application of the analogy points to something else than the SMP as currently implemented. Here are my points:

(i) Prevent self-fulfilling runs. The LOLR function is intended to prevent the collapse of systemic institutions that are vulnerable to self-fulfilling prophecies. In other words, banks are given a backstop because they are believed to be vulnerable to irrational bank runs. Does the analogy hold to self-fulfilling sovereign debt crises? Secondary market purchases do not provide liquidity to the issuer, but keep their secondary market yields low as to help them issue cheaper. To the degree that the SMP enables deficit financing as buyers on the primary market sell the next day to the SMP (something that may well be going), the SMP undermines Art. 123 of the Treaty which forbids deficit monetization. Once the issuer cannot access primary markets, he has no reason to care about gyrations of secondary markets. A stringent application of the analogy would open the refinancing windows to sovereigns, which however is ruled out by Art. 123 of the Treaty to avoid the monetization of deficits.

(ii) Avoid systemic crises. The LOLR backstop is mainly justified by the systemic nature of banks. They are thought to be the backbone of the economy. During the crisis, backstops were extended to other entities that were thought to be systemic, such as automobile manufacturers. But how systemic are sovereign bond markets? Peripheral sovereign bonds have largely lost their benchmark function vis-a-vis local issuers, which trade "through" the sovereign curve. Again, no reason to worry about gyrations of secondary markets.

(iii) Liquidity support. The LOLR is providing liquidity, not solvency support. To that end, the central bank uses collateral rules and can (in many cases) provide further emergency liquidity to solvent institutions, with a supervisory body in charge of monitoring the debtor's solvency. In case of insolvency, a systemic but viable debtor could receive a capital injection from public coffers, but not from the central bank but the fiscal. How does this compare to the SMP? The ECB has no clearly communicated guidance about the solvency of the sovereign debtor. There is no institution supervising sovereigns in the euro area, or an explicit or implicit recapitalization assurance or collateral.

While the analogy in principal is appealing, it seems neither the conditions nor the instrument are defined the right way. Sticking to the analogy, the ECB could also buy corporate bonds and provide liquidity support to pension funds and other investors in the sovereign bond markets!

To make the SMP work in an analogy to the LOLR, the ECB (or another watchdog) would need to develop and apply a threshold of solvency, receive fiscal guarantees towards maintenance of the debtors' solvency, and (after changing Art. 123 of the Treaty) buy on primary markets or open the refinancing window to sovereigns. Shifting any losses to the fiscal would reshape the SMP into a pure liquidity operation of unlimited dimension, which the enhanced EFSF or ESM is lacking.